mariadb/sql/item_func.cc

6635 lines
168 KiB
C++
Raw Normal View History

/* Copyright (c) 2000, 2011, Oracle and/or its affiliates. All rights reserved.
2000-07-31 21:29:14 +02:00
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; version 2 of the License.
2000-07-31 21:29:14 +02:00
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Fix for BUG#11755168 '46895: test "outfile_loaddata" fails (reproducible)'. In sql_class.cc, 'row_count', of type 'ha_rows', was used as last argument for ER_TRUNCATED_WRONG_VALUE_FOR_FIELD which is "Incorrect %-.32s value: '%-.128s' for column '%.192s' at row %ld". So 'ha_rows' was used as 'long'. On SPARC32 Solaris builds, 'long' is 4 bytes and 'ha_rows' is 'longlong' i.e. 8 bytes. So the printf-like code was reading only the first 4 bytes. Because the CPU is big-endian, 1LL is 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x01 so the first four bytes yield 0. So the warning message had "row 0" instead of "row 1" in test outfile_loaddata.test: -Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 1 +Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 0 All error-messaging functions which internally invoke some printf-life function are potential candidate for such mistakes. One apparently easy way to catch such mistakes is to use ATTRIBUTE_FORMAT (from my_attribute.h). But this works only when call site has both: a) the format as a string literal b) the types of arguments. So: func(ER(ER_BLAH), 10); will silently not be checked, because ER(ER_BLAH) is not known at compile time (it is known at run-time, and depends on the chosen language). And func("%s", a va_list argument); has the same problem, as the *real* type of arguments is not known at this site at compile time (it's known in some caller). Moreover, func(ER(ER_BLAH)); though possibly correct (if ER(ER_BLAH) has no '%' markers), will not compile (gcc says "error: format not a string literal and no format arguments"). Consequences: 1) ATTRIBUTE_FORMAT is here added only to functions which in practice take "string literal" formats: "my_error_reporter" and "print_admin_msg". 2) it cannot be added to the other functions: my_error(), push_warning_printf(), Table_check_intact::report_error(), general_log_print(). To do a one-time check of functions listed in (2), the following "static code analysis" has been done: 1) replace my_error(ER_xxx, arguments for substitution in format) with the equivalent my_printf_error(ER_xxx,ER(ER_xxx), arguments for substitution in format), so that we have ER(ER_xxx) and the arguments *in the same call site* 2) add ATTRIBUTE_FORMAT to push_warning_printf(), Table_check_intact::report_error(), general_log_print() 3) replace ER(xxx) with the hard-coded English text found in errmsg.txt (like: ER(ER_UNKNOWN_ERROR) is replaced with "Unknown error"), so that a call site has the format as string literal 4) this way, ATTRIBUTE_FORMAT can effectively do its job 5) compile, fix errors detected by ATTRIBUTE_FORMAT 6) revert steps 1-2-3. The present patch has no compiler error when submitted again to the static code analysis above. It cannot catch all problems though: see Field::set_warning(), in which a call to push_warning_printf() has a variable error (thus, not replacable by a string literal); I checked set_warning() calls by hand though. See also WL 5883 for one proposal to avoid such bugs from appearing again in the future. The issues fixed in the patch are: a) mismatch in types (like 'int' passed to '%ld') b) more arguments passed than specified in the format. This patch resolves mismatches by changing the type/number of arguments, not by changing error messages of sql/share/errmsg.txt. The latter would be wrong, per the following old rule: errmsg.txt must be as stable as possible; no insertions or deletions of messages, no changes of type or number of printf-like format specifiers, are allowed, as long as the change impacts a message already released in a GA version. If this rule is not followed: - Connectors, which use error message numbers, will be confused (by insertions/deletions of messages) - using errmsg.sys of MySQL 5.1.n with mysqld of MySQL 5.1.(n+1) could produce wrong messages or crash; such usage can easily happen if installing 5.1.(n+1) while /etc/my.cnf still has --language=/path/to/5.1.n/xxx; or if copying mysqld from 5.1.(n+1) into a 5.1.n installation. When fixing b), I have verified that the superfluous arguments were not used in the format in the first 5.1 GA (5.1.30 'bteam@astra04-20081114162938-z8mctjp6st27uobm'). Had they been used, then passing them today, even if the message doesn't use them anymore, would have been necessary, as explained above.
2011-05-16 22:04:01 +02:00
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA */
2000-07-31 21:29:14 +02:00
2007-10-11 19:29:09 +02:00
/**
@file
@brief
This file defines all numerical functions
*/
2000-07-31 21:29:14 +02:00
#ifdef USE_PRAGMA_IMPLEMENTATION
2000-07-31 21:29:14 +02:00
#pragma implementation // gcc: Class implementation
#endif
#include "my_global.h" /* NO_EMBEDDED_ACCESS_CHECKS */
#include "sql_priv.h"
/*
It is necessary to include set_var.h instead of item.h because there
are dependencies on include order for set_var.h and item.h. This
will be resolved later.
*/
#include "sql_class.h" // set_var.h: THD
#include "set_var.h"
2003-08-07 19:42:26 +02:00
#include "slave.h" // for wait_for_master_pos
#include "sql_show.h" // append_identifier
#include "strfunc.h" // find_type
#include "sql_parse.h" // is_update_query
#include "sql_acl.h" // EXECUTE_ACL
#include "mysqld.h" // LOCK_uuid_generator
#include "rpl_mi.h"
2000-07-31 21:29:14 +02:00
#include <m_ctype.h>
#include <hash.h>
#include <time.h>
#include <ft_global.h>
#include <my_bit.h>
2000-07-31 21:29:14 +02:00
#include "sp_head.h"
#include "sp_rcontext.h"
#include "sp.h"
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
#include "set_var.h"
#include "debug_sync.h"
#include <mysql/plugin.h>
#include <mysql/service_thd_wait.h>
2003-08-07 19:42:26 +02:00
#ifdef NO_EMBEDDED_ACCESS_CHECKS
#define sp_restore_security_context(A,B) while (0) {}
#endif
bool check_reserved_words(LEX_STRING *name)
{
if (!my_strcasecmp(system_charset_info, name->str, "GLOBAL") ||
!my_strcasecmp(system_charset_info, name->str, "LOCAL") ||
!my_strcasecmp(system_charset_info, name->str, "SESSION"))
return TRUE;
return FALSE;
}
2007-10-11 19:29:09 +02:00
/**
@return
TRUE if item is a constant
*/
2000-07-31 21:29:14 +02:00
bool
eval_const_cond(COND *cond)
{
return ((Item_func*) cond)->val_int() ? TRUE : FALSE;
}
2003-08-07 19:42:26 +02:00
/**
Test if the sum of arguments overflows the ulonglong range.
*/
static inline bool test_if_sum_overflows_ull(ulonglong arg1, ulonglong arg2)
{
return ULONGLONG_MAX - arg1 < arg2;
}
void Item_func::set_arguments(List<Item> &list)
2000-07-31 21:29:14 +02:00
{
allowed_arg_cols= 1;
2000-07-31 21:29:14 +02:00
arg_count=list.elements;
args= tmp_arg; // If 2 arguments
if (arg_count <= 2 || (args=(Item**) sql_alloc(sizeof(Item*)*arg_count)))
2000-07-31 21:29:14 +02:00
{
List_iterator_fast<Item> li(list);
2000-07-31 21:29:14 +02:00
Item *item;
Item **save_args= args;
2000-07-31 21:29:14 +02:00
while ((item=li++))
{
*(save_args++)= item;
2000-07-31 21:29:14 +02:00
with_sum_func|=item->with_sum_func;
}
}
list.empty(); // Fields are used
}
Item_func::Item_func(List<Item> &list)
:allowed_arg_cols(1)
{
set_arguments(list);
}
Item_func::Item_func(THD *thd, Item_func *item)
:Item_result_field(thd, item),
allowed_arg_cols(item->allowed_arg_cols),
arg_count(item->arg_count),
used_tables_cache(item->used_tables_cache),
not_null_tables_cache(item->not_null_tables_cache),
const_item_cache(item->const_item_cache)
{
if (arg_count)
{
if (arg_count <=2)
args= tmp_arg;
else
{
if (!(args=(Item**) thd->alloc(sizeof(Item*)*arg_count)))
return;
}
memcpy((char*) args, (char*) item->args, sizeof(Item*)*arg_count);
}
}
/*
Resolve references to table column for a function and its argument
SYNOPSIS:
fix_fields()
thd Thread object
ref Pointer to where this object is used. This reference
is used if we want to replace this object with another
one (for example in the summary functions).
DESCRIPTION
Call fix_fields() for all arguments to the function. The main intention
is to allow all Item_field() objects to setup pointers to the table fields.
Sets as a side effect the following class variables:
maybe_null Set if any argument may return NULL
with_sum_func Set if any of the arguments contains a sum function
2004-11-16 21:27:31 +01:00
used_tables_cache Set to union of the tables used by arguments
str_value.charset If this is a string function, set this to the
character set for the first argument.
If any argument is binary, this is set to binary
If for any item any of the defaults are wrong, then this can
be fixed in the fix_length_and_dec() function that is called
after this one or by writing a specialized fix_fields() for the
item.
RETURN VALUES
FALSE ok
TRUE Got error. Stored with my_error().
*/
2000-07-31 21:29:14 +02:00
bool
Item_func::fix_fields(THD *thd, Item **ref)
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 0);
2000-07-31 21:29:14 +02:00
Item **arg,**arg_end;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar buff[STACK_BUFF_ALLOC]; // Max argument in function
used_tables_cache= not_null_tables_cache= 0;
2000-07-31 21:29:14 +02:00
const_item_cache=1;
/*
Use stack limit of STACK_MIN_SIZE * 2 since
on some platforms a recursive call to fix_fields
requires more than STACK_MIN_SIZE bytes (e.g. for
MIPS, it takes about 22kB to make one recursive
call to Item_func::fix_fields())
*/
if (check_stack_overrun(thd, STACK_MIN_SIZE * 2, buff))
return TRUE; // Fatal error if flag is set!
2000-07-31 21:29:14 +02:00
if (arg_count)
{ // Print purify happy
for (arg=args, arg_end=args+arg_count; arg != arg_end ; arg++)
{
Item *item;
2004-03-18 14:21:06 +01:00
/*
We can't yet set item to *arg as fix_fields may change *arg
We shouldn't call fix_fields() twice, so check 'fixed' field first
*/
if ((!(*arg)->fixed && (*arg)->fix_fields(thd, arg)))
return TRUE; /* purecov: inspected */
item= *arg;
if (allowed_arg_cols)
{
if (item->check_cols(allowed_arg_cols))
return 1;
}
else
{
/* we have to fetch allowed_arg_cols from first argument */
DBUG_ASSERT(arg == args); // it is first argument
allowed_arg_cols= item->cols();
DBUG_ASSERT(allowed_arg_cols); // Can't be 0 any more
}
2003-03-24 21:52:46 +01:00
if (item->maybe_null)
2000-07-31 21:29:14 +02:00
maybe_null=1;
2003-03-24 21:52:46 +01:00
with_sum_func= with_sum_func || item->with_sum_func;
used_tables_cache|= item->used_tables();
not_null_tables_cache|= item->not_null_tables();
const_item_cache&= item->const_item();
with_subselect|= item->with_subselect;
2000-07-31 21:29:14 +02:00
}
}
fix_length_and_dec();
if (thd->is_error()) // An error inside fix_length_and_dec occured
return TRUE;
fixed= 1;
return FALSE;
2000-07-31 21:29:14 +02:00
}
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
bool Item_func::walk(Item_processor processor, bool walk_subquery,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *argument)
{
if (arg_count)
{
Item **arg,**arg_end;
for (arg= args, arg_end= args+arg_count; arg != arg_end; arg++)
{
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
if ((*arg)->walk(processor, walk_subquery, argument))
return 1;
}
}
return (this->*processor)(argument);
}
void Item_func::traverse_cond(Cond_traverser traverser,
void *argument, traverse_order order)
2004-12-17 21:13:22 +01:00
{
if (arg_count)
{
Item **arg,**arg_end;
switch (order) {
case(PREFIX):
2005-02-11 22:05:24 +01:00
(*traverser)(this, argument);
for (arg= args, arg_end= args+arg_count; arg != arg_end; arg++)
{
(*arg)->traverse_cond(traverser, argument, order);
}
break;
case (POSTFIX):
for (arg= args, arg_end= args+arg_count; arg != arg_end; arg++)
{
(*arg)->traverse_cond(traverser, argument, order);
}
2005-02-11 22:05:24 +01:00
(*traverser)(this, argument);
2004-12-17 21:13:22 +01:00
}
}
else
(*traverser)(this, argument);
2004-12-17 21:13:22 +01:00
}
2007-10-11 19:29:09 +02:00
/**
Transform an Item_func object with a transformer callback function.
The function recursively applies the transform method to each
argument of the Item_func node.
If the call of the method for an argument item returns a new item
the old item is substituted for a new one.
After this the transformer is applied to the root node
of the Item_func object.
2007-10-11 19:29:09 +02:00
@param transformer the transformer callback function to be applied to
the nodes of the tree of the object
@param argument parameter to be passed to the transformer
@return
Item returned as the result of transformation of the root node
*/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
Item *Item_func::transform(Item_transformer transformer, uchar *argument)
{
DBUG_ASSERT(!current_thd->stmt_arena->is_stmt_prepare());
if (arg_count)
{
Item **arg,**arg_end;
for (arg= args, arg_end= args+arg_count; arg != arg_end; arg++)
{
Item *new_item= (*arg)->transform(transformer, argument);
if (!new_item)
return 0;
/*
THD::change_item_tree() should be called only if the tree was
really transformed, i.e. when a new item has been created.
Otherwise we'll be allocating a lot of unnecessary memory for
change records at each execution.
*/
if (*arg != new_item)
current_thd->change_item_tree(arg, new_item);
}
}
return (this->*transformer)(argument);
}
2007-10-11 19:29:09 +02:00
/**
Compile Item_func object with a processor and a transformer
callback functions.
First the function applies the analyzer to the root node of
the Item_func object. Then if the analizer succeeeds (returns TRUE)
the function recursively applies the compile method to each argument
of the Item_func node.
If the call of the method for an argument item returns a new item
the old item is substituted for a new one.
After this the transformer is applied to the root node
of the Item_func object.
2007-10-11 19:29:09 +02:00
@param analyzer the analyzer callback function to be applied to the
nodes of the tree of the object
@param[in,out] arg_p parameter to be passed to the processor
@param transformer the transformer callback function to be applied to the
nodes of the tree of the object
@param arg_t parameter to be passed to the transformer
@return
Item returned as the result of transformation of the root node
*/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
Item *Item_func::compile(Item_analyzer analyzer, uchar **arg_p,
Item_transformer transformer, uchar *arg_t)
{
if (!(this->*analyzer)(arg_p))
return 0;
if (arg_count)
{
Item **arg,**arg_end;
for (arg= args, arg_end= args+arg_count; arg != arg_end; arg++)
{
/*
The same parameter value of arg_p must be passed
to analyze any argument of the condition formula.
*/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *arg_v= *arg_p;
Item *new_item= (*arg)->compile(analyzer, &arg_v, transformer, arg_t);
if (new_item && *arg != new_item)
current_thd->change_item_tree(arg, new_item);
}
}
return (this->*transformer)(arg_t);
}
2007-10-11 19:29:09 +02:00
/**
See comments in Item_cmp_func::split_sum_func()
*/
void Item_func::split_sum_func(THD *thd, Item **ref_pointer_array,
List<Item> &fields)
2000-07-31 21:29:14 +02:00
{
Item **arg, **arg_end;
for (arg= args, arg_end= args+arg_count; arg != arg_end ; arg++)
(*arg)->split_sum_func2(thd, ref_pointer_array, fields, arg, TRUE);
2000-07-31 21:29:14 +02:00
}
void Item_func::update_used_tables()
{
used_tables_cache=0;
const_item_cache=1;
for (uint i=0 ; i < arg_count ; i++)
{
args[i]->update_used_tables();
used_tables_cache|=args[i]->used_tables();
const_item_cache&=args[i]->const_item();
}
}
table_map Item_func::used_tables() const
{
return used_tables_cache;
}
table_map Item_func::not_null_tables() const
{
return not_null_tables_cache;
}
void Item_func::print(String *str, enum_query_type query_type)
2000-07-31 21:29:14 +02:00
{
str->append(func_name());
str->append('(');
print_args(str, 0, query_type);
str->append(')');
}
void Item_func::print_args(String *str, uint from, enum_query_type query_type)
{
2003-11-03 11:28:36 +01:00
for (uint i=from ; i < arg_count ; i++)
2000-07-31 21:29:14 +02:00
{
2003-11-03 11:28:36 +01:00
if (i != from)
2000-07-31 21:29:14 +02:00
str->append(',');
args[i]->print(str, query_type);
2000-07-31 21:29:14 +02:00
}
}
void Item_func::print_op(String *str, enum_query_type query_type)
2000-07-31 21:29:14 +02:00
{
str->append('(');
for (uint i=0 ; i < arg_count-1 ; i++)
{
args[i]->print(str, query_type);
2000-07-31 21:29:14 +02:00
str->append(' ');
str->append(func_name());
str->append(' ');
}
args[arg_count-1]->print(str, query_type);
2000-07-31 21:29:14 +02:00
str->append(')');
}
bool Item_func::eq(const Item *item, bool binary_cmp) const
2000-07-31 21:29:14 +02:00
{
/* Assume we don't have rtti */
if (this == item)
return 1;
if (item->type() != FUNC_ITEM)
return 0;
Item_func *item_func=(Item_func*) item;
Item_func::Functype func_type;
if ((func_type= functype()) != item_func->functype() ||
arg_count != item_func->arg_count ||
(func_type != Item_func::FUNC_SP &&
func_name() != item_func->func_name()) ||
(func_type == Item_func::FUNC_SP &&
my_strcasecmp(system_charset_info, func_name(), item_func->func_name())))
2000-07-31 21:29:14 +02:00
return 0;
for (uint i=0; i < arg_count ; i++)
if (!args[i]->eq(item_func->args[i], binary_cmp))
2000-07-31 21:29:14 +02:00
return 0;
return 1;
}
Field *Item_func::tmp_table_field(TABLE *table)
{
2009-09-17 17:25:52 +02:00
Field *field= NULL;
2002-09-04 14:24:27 +02:00
switch (result_type()) {
case INT_RESULT:
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
if (max_char_length() > MY_INT32_NUM_DECIMAL_DIGITS)
field= new Field_longlong(max_char_length(), maybe_null, name,
unsigned_flag);
else
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
field= new Field_long(max_char_length(), maybe_null, name,
unsigned_flag);
break;
case REAL_RESULT:
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
field= new Field_double(max_char_length(), maybe_null, name, decimals);
break;
case STRING_RESULT:
return make_string_field(table);
break;
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
field= Field_new_decimal::create_from_item(this);
2005-02-08 23:50:45 +01:00
break;
case ROW_RESULT:
2003-01-31 11:07:07 +01:00
default:
2005-02-08 23:50:45 +01:00
// This case should never be chosen
2002-11-15 19:32:09 +01:00
DBUG_ASSERT(0);
field= 0;
2002-11-15 19:32:09 +01:00
break;
}
if (field)
field->init(table);
return field;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
bool Item_func::is_expensive_processor(uchar *arg)
{
return is_expensive();
}
2005-02-08 23:50:45 +01:00
my_decimal *Item_func::val_decimal(my_decimal *decimal_value)
{
DBUG_ASSERT(fixed);
longlong nr= val_int();
if (null_value)
return 0; /* purecov: inspected */
int2my_decimal(E_DEC_FATAL_ERROR, nr, unsigned_flag, decimal_value);
2005-02-08 23:50:45 +01:00
return decimal_value;
}
2000-07-31 21:29:14 +02:00
String *Item_real_func::val_str(String *str)
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double nr= val_real();
2000-07-31 21:29:14 +02:00
if (null_value)
return 0; /* purecov: inspected */
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_real(nr, decimals, collation.collation);
2000-07-31 21:29:14 +02:00
return str;
}
my_decimal *Item_real_func::val_decimal(my_decimal *decimal_value)
{
DBUG_ASSERT(fixed);
double nr= val_real();
if (null_value)
return 0; /* purecov: inspected */
double2my_decimal(E_DEC_FATAL_ERROR, nr, decimal_value);
return decimal_value;
}
2005-02-08 23:50:45 +01:00
void Item_func::fix_num_length_and_dec()
2000-07-31 21:29:14 +02:00
{
uint fl_length= 0;
2000-07-31 21:29:14 +02:00
decimals=0;
2005-02-08 23:50:45 +01:00
for (uint i=0 ; i < arg_count ; i++)
2000-07-31 21:29:14 +02:00
{
set_if_bigger(decimals,args[i]->decimals);
set_if_bigger(fl_length, args[i]->max_length);
}
2000-07-31 21:29:14 +02:00
max_length=float_length(decimals);
if (fl_length > max_length)
{
decimals= NOT_FIXED_DEC;
max_length= float_length(NOT_FIXED_DEC);
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
}
void Item_func_numhybrid::fix_num_length_and_dec()
{}
2007-10-11 19:29:09 +02:00
/**
2005-02-08 23:50:45 +01:00
Set max_length/decimals of function if function is fixed point and
2007-10-11 19:29:09 +02:00
result length/precision depends on argument ones.
2005-02-08 23:50:45 +01:00
*/
void Item_func::count_decimal_length()
{
int max_int_part= 0;
2005-02-08 23:50:45 +01:00
decimals= 0;
unsigned_flag= 1;
2005-02-08 23:50:45 +01:00
for (uint i=0 ; i < arg_count ; i++)
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
set_if_bigger(decimals, args[i]->decimals);
set_if_bigger(max_int_part, args[i]->decimal_int_part());
set_if_smaller(unsigned_flag, args[i]->unsigned_flag);
2000-07-31 21:29:14 +02:00
}
int precision= min(max_int_part + decimals, DECIMAL_MAX_PRECISION);
Bug#54916 GROUP_CONCAT + IFNULL truncates output Problem: a few functions did not calculate their max_length correctly. This is an after-fix for WL#2649 Number-to-string conversions". Fix: changing the buggy functions to calculate max_length using fix_char_length() introduced in WL#2649, instead of setting max_length directly mysql-test/include/ctype_numconv.inc Adding new tests mysql-test/r/ctype_binary.result Adding new tests mysql-test/r/ctype_cp1251.result Adding new tests mysql-test/r/ctype_latin1.result Adding new tests mysql-test/r/ctype_ucs.result Adding new tests mysql-test/r/ctype_utf8.result Adding new tests mysql-test/t/ctype_utf8.test Including ctype_numconv sql/item.h - Introducing new method fix_char_length_ulonglong(), for the cases when length is potentially greater than UINT_MAX32. This method removes a few instances of duplicate code, e.g. in item_strfunc.cc. - Setting collation in Item_copy properly. This change fixes wrong metadata on client side in some cases, when "binary" instead of the real character set was reported. sql/item_cmpfunc.cc - Using fix_char_length() and max_char_length() methods, instead of direct access to max_length, to calculate item length properly. - Moving count_only_length() in COALESCE after agg_arg_charsets_for_string_result(). The old order was incorrect and led to wrong length calucation in case of multi-byte character sets. sql/item_func.cc Fixing that count_only_length() didn't work properly for multi-byte character sets. Using fix_char_length() and max_char_length() instead of direct access to max_length. sql/item_strfunc.cc - Using fix_char_length(), fix_char_length_ulonglong(), max_char_length() instead of direct access to max_length. - Removing wierd condition: "if (collation.collation->mbmaxlen > 0)", which is never FALSE.
2010-08-19 13:55:35 +02:00
fix_char_length(my_decimal_precision_to_length_no_truncation(precision,
decimals,
unsigned_flag));
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/**
Set max_length of if it is maximum length of its arguments.
2005-02-08 23:50:45 +01:00
*/
void Item_func::count_only_length()
2000-07-31 21:29:14 +02:00
{
Bug#54916 GROUP_CONCAT + IFNULL truncates output Problem: a few functions did not calculate their max_length correctly. This is an after-fix for WL#2649 Number-to-string conversions". Fix: changing the buggy functions to calculate max_length using fix_char_length() introduced in WL#2649, instead of setting max_length directly mysql-test/include/ctype_numconv.inc Adding new tests mysql-test/r/ctype_binary.result Adding new tests mysql-test/r/ctype_cp1251.result Adding new tests mysql-test/r/ctype_latin1.result Adding new tests mysql-test/r/ctype_ucs.result Adding new tests mysql-test/r/ctype_utf8.result Adding new tests mysql-test/t/ctype_utf8.test Including ctype_numconv sql/item.h - Introducing new method fix_char_length_ulonglong(), for the cases when length is potentially greater than UINT_MAX32. This method removes a few instances of duplicate code, e.g. in item_strfunc.cc. - Setting collation in Item_copy properly. This change fixes wrong metadata on client side in some cases, when "binary" instead of the real character set was reported. sql/item_cmpfunc.cc - Using fix_char_length() and max_char_length() methods, instead of direct access to max_length, to calculate item length properly. - Moving count_only_length() in COALESCE after agg_arg_charsets_for_string_result(). The old order was incorrect and led to wrong length calucation in case of multi-byte character sets. sql/item_func.cc Fixing that count_only_length() didn't work properly for multi-byte character sets. Using fix_char_length() and max_char_length() instead of direct access to max_length. sql/item_strfunc.cc - Using fix_char_length(), fix_char_length_ulonglong(), max_char_length() instead of direct access to max_length. - Removing wierd condition: "if (collation.collation->mbmaxlen > 0)", which is never FALSE.
2010-08-19 13:55:35 +02:00
uint32 char_length= 0;
unsigned_flag= 0;
2000-07-31 21:29:14 +02:00
for (uint i=0 ; i < arg_count ; i++)
{
Bug#54916 GROUP_CONCAT + IFNULL truncates output Problem: a few functions did not calculate their max_length correctly. This is an after-fix for WL#2649 Number-to-string conversions". Fix: changing the buggy functions to calculate max_length using fix_char_length() introduced in WL#2649, instead of setting max_length directly mysql-test/include/ctype_numconv.inc Adding new tests mysql-test/r/ctype_binary.result Adding new tests mysql-test/r/ctype_cp1251.result Adding new tests mysql-test/r/ctype_latin1.result Adding new tests mysql-test/r/ctype_ucs.result Adding new tests mysql-test/r/ctype_utf8.result Adding new tests mysql-test/t/ctype_utf8.test Including ctype_numconv sql/item.h - Introducing new method fix_char_length_ulonglong(), for the cases when length is potentially greater than UINT_MAX32. This method removes a few instances of duplicate code, e.g. in item_strfunc.cc. - Setting collation in Item_copy properly. This change fixes wrong metadata on client side in some cases, when "binary" instead of the real character set was reported. sql/item_cmpfunc.cc - Using fix_char_length() and max_char_length() methods, instead of direct access to max_length, to calculate item length properly. - Moving count_only_length() in COALESCE after agg_arg_charsets_for_string_result(). The old order was incorrect and led to wrong length calucation in case of multi-byte character sets. sql/item_func.cc Fixing that count_only_length() didn't work properly for multi-byte character sets. Using fix_char_length() and max_char_length() instead of direct access to max_length. sql/item_strfunc.cc - Using fix_char_length(), fix_char_length_ulonglong(), max_char_length() instead of direct access to max_length. - Removing wierd condition: "if (collation.collation->mbmaxlen > 0)", which is never FALSE.
2010-08-19 13:55:35 +02:00
set_if_bigger(char_length, args[i]->max_char_length());
set_if_bigger(unsigned_flag, args[i]->unsigned_flag);
}
Bug#54916 GROUP_CONCAT + IFNULL truncates output Problem: a few functions did not calculate their max_length correctly. This is an after-fix for WL#2649 Number-to-string conversions". Fix: changing the buggy functions to calculate max_length using fix_char_length() introduced in WL#2649, instead of setting max_length directly mysql-test/include/ctype_numconv.inc Adding new tests mysql-test/r/ctype_binary.result Adding new tests mysql-test/r/ctype_cp1251.result Adding new tests mysql-test/r/ctype_latin1.result Adding new tests mysql-test/r/ctype_ucs.result Adding new tests mysql-test/r/ctype_utf8.result Adding new tests mysql-test/t/ctype_utf8.test Including ctype_numconv sql/item.h - Introducing new method fix_char_length_ulonglong(), for the cases when length is potentially greater than UINT_MAX32. This method removes a few instances of duplicate code, e.g. in item_strfunc.cc. - Setting collation in Item_copy properly. This change fixes wrong metadata on client side in some cases, when "binary" instead of the real character set was reported. sql/item_cmpfunc.cc - Using fix_char_length() and max_char_length() methods, instead of direct access to max_length, to calculate item length properly. - Moving count_only_length() in COALESCE after agg_arg_charsets_for_string_result(). The old order was incorrect and led to wrong length calucation in case of multi-byte character sets. sql/item_func.cc Fixing that count_only_length() didn't work properly for multi-byte character sets. Using fix_char_length() and max_char_length() instead of direct access to max_length. sql/item_strfunc.cc - Using fix_char_length(), fix_char_length_ulonglong(), max_char_length() instead of direct access to max_length. - Removing wierd condition: "if (collation.collation->mbmaxlen > 0)", which is never FALSE.
2010-08-19 13:55:35 +02:00
fix_char_length(char_length);
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
2007-10-11 19:29:09 +02:00
/**
2005-02-08 23:50:45 +01:00
Set max_length/decimals of function if function is floating point and
2007-10-11 19:29:09 +02:00
result length/precision depends on argument ones.
2005-02-08 23:50:45 +01:00
*/
void Item_func::count_real_length()
{
uint32 length= 0;
decimals= 0;
max_length= 0;
for (uint i=0 ; i < arg_count ; i++)
{
if (decimals != NOT_FIXED_DEC)
{
set_if_bigger(decimals, args[i]->decimals);
set_if_bigger(length, (args[i]->max_length - args[i]->decimals));
}
set_if_bigger(max_length, args[i]->max_length);
}
if (decimals != NOT_FIXED_DEC)
{
max_length= length;
length+= decimals;
if (length < max_length) // If previous operation gave overflow
max_length= UINT_MAX32;
else
max_length= length;
}
}
void Item_func::signal_divide_by_null()
{
THD *thd= current_thd;
if (thd->variables.sql_mode & MODE_ERROR_FOR_DIVISION_BY_ZERO)
push_warning(thd, MYSQL_ERROR::WARN_LEVEL_WARN, ER_DIVISION_BY_ZERO,
ER(ER_DIVISION_BY_ZERO));
null_value= 1;
}
Item *Item_func::get_tmp_table_item(THD *thd)
{
if (!with_sum_func && !const_item())
return new Item_field(result_field);
return copy_or_same(thd);
}
double Item_int_func::val_real()
{
DBUG_ASSERT(fixed == 1);
return unsigned_flag ? (double) ((ulonglong) val_int()) : (double) val_int();
}
2000-07-31 21:29:14 +02:00
String *Item_int_func::val_str(String *str)
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
longlong nr=val_int();
if (null_value)
return 0;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_int(nr, unsigned_flag, collation.collation);
2000-07-31 21:29:14 +02:00
return str;
}
2005-02-08 23:50:45 +01:00
void Item_func_connection_id::fix_length_and_dec()
{
Item_int_func::fix_length_and_dec();
max_length= 10;
}
bool Item_func_connection_id::fix_fields(THD *thd, Item **ref)
{
if (Item_int_func::fix_fields(thd, ref))
return TRUE;
thd->thread_specific_used= TRUE;
value= thd->variables.pseudo_thread_id;
return FALSE;
}
2007-10-11 19:29:09 +02:00
/**
Check arguments here to determine result's type for a numeric
function of two arguments.
*/
2000-07-31 21:29:14 +02:00
void Item_num_op::find_num_type(void)
{
2005-02-08 23:50:45 +01:00
DBUG_ENTER("Item_num_op::find_num_type");
DBUG_PRINT("info", ("name %s", func_name()));
DBUG_ASSERT(arg_count == 2);
Item_result r0= args[0]->result_type();
Item_result r1= args[1]->result_type();
if (r0 == REAL_RESULT || r1 == REAL_RESULT ||
r0 == STRING_RESULT || r1 ==STRING_RESULT)
{
2005-02-08 23:50:45 +01:00
count_real_length();
max_length= float_length(decimals);
hybrid_type= REAL_RESULT;
}
else if (r0 == DECIMAL_RESULT || r1 == DECIMAL_RESULT)
{
hybrid_type= DECIMAL_RESULT;
result_precision();
}
else
{
DBUG_ASSERT(r0 == INT_RESULT && r1 == INT_RESULT);
2005-02-08 23:50:45 +01:00
decimals= 0;
2000-07-31 21:29:14 +02:00
hybrid_type=INT_RESULT;
2005-02-08 23:50:45 +01:00
result_precision();
}
2005-02-08 23:50:45 +01:00
DBUG_PRINT("info", ("Type: %s",
(hybrid_type == REAL_RESULT ? "REAL_RESULT" :
hybrid_type == DECIMAL_RESULT ? "DECIMAL_RESULT" :
hybrid_type == INT_RESULT ? "INT_RESULT" :
"--ILLEGAL!!!--")));
DBUG_VOID_RETURN;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
2007-10-11 19:29:09 +02:00
/**
Set result type for a numeric function of one argument
(can be also used by a numeric function of many arguments, if the result
type depends only on the first argument)
2005-02-08 23:50:45 +01:00
*/
2005-02-08 23:50:45 +01:00
void Item_func_num1::find_num_type()
{
DBUG_ENTER("Item_func_num1::find_num_type");
DBUG_PRINT("info", ("name %s", func_name()));
switch (hybrid_type= args[0]->result_type()) {
2005-02-08 23:50:45 +01:00
case INT_RESULT:
unsigned_flag= args[0]->unsigned_flag;
2005-02-08 23:50:45 +01:00
break;
case STRING_RESULT:
case REAL_RESULT:
hybrid_type= REAL_RESULT;
max_length= float_length(decimals);
break;
case DECIMAL_RESULT:
break;
default:
DBUG_ASSERT(0);
}
2005-02-08 23:50:45 +01:00
DBUG_PRINT("info", ("Type: %s",
(hybrid_type == REAL_RESULT ? "REAL_RESULT" :
hybrid_type == DECIMAL_RESULT ? "DECIMAL_RESULT" :
hybrid_type == INT_RESULT ? "INT_RESULT" :
"--ILLEGAL!!!--")));
DBUG_VOID_RETURN;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
void Item_func_num1::fix_num_length_and_dec()
{
decimals= args[0]->decimals;
max_length= args[0]->max_length;
}
void Item_func_numhybrid::fix_length_and_dec()
{
fix_num_length_and_dec();
find_num_type();
}
String *Item_func_numhybrid::val_str(String *str)
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
switch (hybrid_type) {
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
my_decimal decimal_value, *val;
if (!(val= decimal_op(&decimal_value)))
return 0; // null is set
2005-02-08 23:50:45 +01:00
my_decimal_round(E_DEC_FATAL_ERROR, val, decimals, FALSE, val);
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_charset(collation.collation);
2005-02-08 23:50:45 +01:00
my_decimal2string(E_DEC_FATAL_ERROR, val, 0, 0, 0, str);
break;
}
case INT_RESULT:
{
longlong nr= int_op();
2000-07-31 21:29:14 +02:00
if (null_value)
return 0; /* purecov: inspected */
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_int(nr, unsigned_flag, collation.collation);
2005-02-08 23:50:45 +01:00
break;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
case REAL_RESULT:
2000-07-31 21:29:14 +02:00
{
double nr= real_op();
2000-07-31 21:29:14 +02:00
if (null_value)
return 0; /* purecov: inspected */
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_real(nr, decimals, collation.collation);
2005-02-08 23:50:45 +01:00
break;
}
case STRING_RESULT:
return str_op(&str_value);
2005-02-08 23:50:45 +01:00
default:
DBUG_ASSERT(0);
2000-07-31 21:29:14 +02:00
}
return str;
}
2005-02-08 23:50:45 +01:00
double Item_func_numhybrid::val_real()
{
DBUG_ASSERT(fixed == 1);
switch (hybrid_type) {
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
my_decimal decimal_value, *val;
double result;
if (!(val= decimal_op(&decimal_value)))
return 0.0; // null is set
2005-02-08 23:50:45 +01:00
my_decimal2double(E_DEC_FATAL_ERROR, val, &result);
return result;
}
case INT_RESULT:
{
longlong result= int_op();
return unsigned_flag ? (double) ((ulonglong) result) : (double) result;
}
2005-02-08 23:50:45 +01:00
case REAL_RESULT:
return real_op();
case STRING_RESULT:
{
char *end_not_used;
int err_not_used;
String *res= str_op(&str_value);
return (res ? my_strntod(res->charset(), (char*) res->ptr(), res->length(),
&end_not_used, &err_not_used) : 0.0);
}
2005-02-08 23:50:45 +01:00
default:
DBUG_ASSERT(0);
}
return 0.0;
}
longlong Item_func_numhybrid::val_int()
{
DBUG_ASSERT(fixed == 1);
switch (hybrid_type) {
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
my_decimal decimal_value, *val;
if (!(val= decimal_op(&decimal_value)))
return 0; // null is set
2005-02-08 23:50:45 +01:00
longlong result;
my_decimal2int(E_DEC_FATAL_ERROR, val, unsigned_flag, &result);
return result;
}
case INT_RESULT:
return int_op();
case REAL_RESULT:
return (longlong) rint(real_op());
case STRING_RESULT:
{
int err_not_used;
String *res;
if (!(res= str_op(&str_value)))
return 0;
char *end= (char*) res->ptr() + res->length();
CHARSET_INFO *cs= res->charset();
return (*(cs->cset->strtoll10))(cs, res->ptr(), &end, &err_not_used);
}
2005-02-08 23:50:45 +01:00
default:
DBUG_ASSERT(0);
}
return 0;
}
my_decimal *Item_func_numhybrid::val_decimal(my_decimal *decimal_value)
{
my_decimal *val= decimal_value;
DBUG_ASSERT(fixed == 1);
switch (hybrid_type) {
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
val= decimal_op(decimal_value);
break;
case INT_RESULT:
{
longlong result= int_op();
int2my_decimal(E_DEC_FATAL_ERROR, result, unsigned_flag, decimal_value);
break;
}
case REAL_RESULT:
{
double result= (double)real_op();
2005-02-08 23:50:45 +01:00
double2my_decimal(E_DEC_FATAL_ERROR, result, decimal_value);
break;
}
case STRING_RESULT:
{
String *res;
if (!(res= str_op(&str_value)))
return NULL;
str2my_decimal(E_DEC_FATAL_ERROR, (char*) res->ptr(),
res->length(), res->charset(), decimal_value);
break;
}
2005-02-08 23:50:45 +01:00
case ROW_RESULT:
default:
DBUG_ASSERT(0);
}
return val;
}
void Item_func_signed::print(String *str, enum_query_type query_type)
{
str->append(STRING_WITH_LEN("cast("));
args[0]->print(str, query_type);
str->append(STRING_WITH_LEN(" as signed)"));
}
longlong Item_func_signed::val_int_from_str(int *error)
{
char buff[MAX_FIELD_WIDTH], *end, *start;
uint32 length;
String tmp(buff,sizeof(buff), &my_charset_bin), *res;
longlong value;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
CHARSET_INFO *cs;
/*
For a string result, we must first get the string and then convert it
to a longlong
*/
if (!(res= args[0]->val_str(&tmp)))
{
null_value= 1;
*error= 0;
return 0;
}
null_value= 0;
start= (char *)res->ptr();
length= res->length();
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
cs= res->charset();
end= start + length;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
value= cs->cset->strtoll10(cs, start, &end, error);
if (*error > 0 || end != start+ length)
{
char err_buff[128];
String err_tmp(err_buff,(uint32) sizeof(err_buff), system_charset_info);
err_tmp.copy(start, length, system_charset_info);
push_warning_printf(current_thd, MYSQL_ERROR::WARN_LEVEL_WARN,
ER_TRUNCATED_WRONG_VALUE,
ER(ER_TRUNCATED_WRONG_VALUE), "INTEGER",
err_tmp.c_ptr());
}
return value;
}
longlong Item_func_signed::val_int()
{
longlong value;
int error;
if (args[0]->cast_to_int_type() != STRING_RESULT ||
args[0]->result_as_longlong())
{
value= args[0]->val_int();
null_value= args[0]->null_value;
return value;
}
value= val_int_from_str(&error);
if (value < 0 && error == 0)
{
push_warning(current_thd, MYSQL_ERROR::WARN_LEVEL_WARN, ER_UNKNOWN_ERROR,
"Cast to signed converted positive out-of-range integer to "
"it's negative complement");
}
return value;
}
void Item_func_unsigned::print(String *str, enum_query_type query_type)
{
str->append(STRING_WITH_LEN("cast("));
args[0]->print(str, query_type);
str->append(STRING_WITH_LEN(" as unsigned)"));
}
longlong Item_func_unsigned::val_int()
{
longlong value;
int error;
if (args[0]->cast_to_int_type() == DECIMAL_RESULT)
{
my_decimal tmp, *dec= args[0]->val_decimal(&tmp);
if (!(null_value= args[0]->null_value))
my_decimal2int(E_DEC_FATAL_ERROR, dec, 1, &value);
else
value= 0;
return value;
}
else if (args[0]->cast_to_int_type() != STRING_RESULT ||
args[0]->result_as_longlong())
{
value= args[0]->val_int();
null_value= args[0]->null_value;
return value;
}
value= val_int_from_str(&error);
if (error < 0)
push_warning(current_thd, MYSQL_ERROR::WARN_LEVEL_WARN, ER_UNKNOWN_ERROR,
"Cast to unsigned converted negative integer to it's "
"positive complement");
return value;
}
2005-02-08 23:50:45 +01:00
String *Item_decimal_typecast::val_str(String *str)
{
my_decimal tmp_buf, *tmp= val_decimal(&tmp_buf);
if (null_value)
return NULL;
my_decimal2string(E_DEC_FATAL_ERROR, tmp, 0, 0, 0, str);
2005-02-08 23:50:45 +01:00
return str;
}
double Item_decimal_typecast::val_real()
{
my_decimal tmp_buf, *tmp= val_decimal(&tmp_buf);
double res;
if (null_value)
return 0.0;
2005-02-08 23:50:45 +01:00
my_decimal2double(E_DEC_FATAL_ERROR, tmp, &res);
return res;
}
longlong Item_decimal_typecast::val_int()
{
my_decimal tmp_buf, *tmp= val_decimal(&tmp_buf);
longlong res;
if (null_value)
return 0;
2005-02-08 23:50:45 +01:00
my_decimal2int(E_DEC_FATAL_ERROR, tmp, unsigned_flag, &res);
return res;
}
my_decimal *Item_decimal_typecast::val_decimal(my_decimal *dec)
{
my_decimal tmp_buf, *tmp= args[0]->val_decimal(&tmp_buf);
bool sign;
2007-05-11 17:56:22 +02:00
uint precision;
if ((null_value= args[0]->null_value))
return NULL;
2005-02-08 23:50:45 +01:00
my_decimal_round(E_DEC_FATAL_ERROR, tmp, decimals, FALSE, dec);
sign= dec->sign();
if (unsigned_flag)
{
if (sign)
{
my_decimal_set_zero(dec);
goto err;
}
}
2007-05-11 17:56:22 +02:00
precision= my_decimal_length_to_precision(max_length,
decimals, unsigned_flag);
if (precision - decimals < (uint) my_decimal_intg(dec))
{
2007-05-11 17:56:22 +02:00
max_my_decimal(dec, precision, decimals);
dec->sign(sign);
goto err;
}
return dec;
err:
push_warning_printf(current_thd, MYSQL_ERROR::WARN_LEVEL_WARN,
ER_WARN_DATA_OUT_OF_RANGE,
ER(ER_WARN_DATA_OUT_OF_RANGE),
Fix for BUG#11755168 '46895: test "outfile_loaddata" fails (reproducible)'. In sql_class.cc, 'row_count', of type 'ha_rows', was used as last argument for ER_TRUNCATED_WRONG_VALUE_FOR_FIELD which is "Incorrect %-.32s value: '%-.128s' for column '%.192s' at row %ld". So 'ha_rows' was used as 'long'. On SPARC32 Solaris builds, 'long' is 4 bytes and 'ha_rows' is 'longlong' i.e. 8 bytes. So the printf-like code was reading only the first 4 bytes. Because the CPU is big-endian, 1LL is 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x01 so the first four bytes yield 0. So the warning message had "row 0" instead of "row 1" in test outfile_loaddata.test: -Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 1 +Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 0 All error-messaging functions which internally invoke some printf-life function are potential candidate for such mistakes. One apparently easy way to catch such mistakes is to use ATTRIBUTE_FORMAT (from my_attribute.h). But this works only when call site has both: a) the format as a string literal b) the types of arguments. So: func(ER(ER_BLAH), 10); will silently not be checked, because ER(ER_BLAH) is not known at compile time (it is known at run-time, and depends on the chosen language). And func("%s", a va_list argument); has the same problem, as the *real* type of arguments is not known at this site at compile time (it's known in some caller). Moreover, func(ER(ER_BLAH)); though possibly correct (if ER(ER_BLAH) has no '%' markers), will not compile (gcc says "error: format not a string literal and no format arguments"). Consequences: 1) ATTRIBUTE_FORMAT is here added only to functions which in practice take "string literal" formats: "my_error_reporter" and "print_admin_msg". 2) it cannot be added to the other functions: my_error(), push_warning_printf(), Table_check_intact::report_error(), general_log_print(). To do a one-time check of functions listed in (2), the following "static code analysis" has been done: 1) replace my_error(ER_xxx, arguments for substitution in format) with the equivalent my_printf_error(ER_xxx,ER(ER_xxx), arguments for substitution in format), so that we have ER(ER_xxx) and the arguments *in the same call site* 2) add ATTRIBUTE_FORMAT to push_warning_printf(), Table_check_intact::report_error(), general_log_print() 3) replace ER(xxx) with the hard-coded English text found in errmsg.txt (like: ER(ER_UNKNOWN_ERROR) is replaced with "Unknown error"), so that a call site has the format as string literal 4) this way, ATTRIBUTE_FORMAT can effectively do its job 5) compile, fix errors detected by ATTRIBUTE_FORMAT 6) revert steps 1-2-3. The present patch has no compiler error when submitted again to the static code analysis above. It cannot catch all problems though: see Field::set_warning(), in which a call to push_warning_printf() has a variable error (thus, not replacable by a string literal); I checked set_warning() calls by hand though. See also WL 5883 for one proposal to avoid such bugs from appearing again in the future. The issues fixed in the patch are: a) mismatch in types (like 'int' passed to '%ld') b) more arguments passed than specified in the format. This patch resolves mismatches by changing the type/number of arguments, not by changing error messages of sql/share/errmsg.txt. The latter would be wrong, per the following old rule: errmsg.txt must be as stable as possible; no insertions or deletions of messages, no changes of type or number of printf-like format specifiers, are allowed, as long as the change impacts a message already released in a GA version. If this rule is not followed: - Connectors, which use error message numbers, will be confused (by insertions/deletions of messages) - using errmsg.sys of MySQL 5.1.n with mysqld of MySQL 5.1.(n+1) could produce wrong messages or crash; such usage can easily happen if installing 5.1.(n+1) while /etc/my.cnf still has --language=/path/to/5.1.n/xxx; or if copying mysqld from 5.1.(n+1) into a 5.1.n installation. When fixing b), I have verified that the superfluous arguments were not used in the format in the first 5.1 GA (5.1.30 'bteam@astra04-20081114162938-z8mctjp6st27uobm'). Had they been used, then passing them today, even if the message doesn't use them anymore, would have been necessary, as explained above.
2011-05-16 22:04:01 +02:00
name, 1L);
2005-02-08 23:50:45 +01:00
return dec;
}
void Item_decimal_typecast::print(String *str, enum_query_type query_type)
{
char len_buf[20*3 + 1];
char *end;
uint precision= my_decimal_length_to_precision(max_length, decimals,
unsigned_flag);
str->append(STRING_WITH_LEN("cast("));
args[0]->print(str, query_type);
str->append(STRING_WITH_LEN(" as decimal("));
end=int10_to_str(precision, len_buf,10);
str->append(len_buf, (uint32) (end - len_buf));
str->append(',');
end=int10_to_str(decimals, len_buf,10);
str->append(len_buf, (uint32) (end - len_buf));
str->append(')');
str->append(')');
}
2005-02-08 23:50:45 +01:00
double Item_func_plus::real_op()
2000-07-31 21:29:14 +02:00
{
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real() + args[1]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value || args[1]->null_value))
return 0.0;
return check_float_overflow(value);
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
longlong Item_func_plus::int_op()
2000-07-31 21:29:14 +02:00
{
longlong val0= args[0]->val_int();
longlong val1= args[1]->val_int();
longlong res= val0 + val1;
bool res_unsigned= FALSE;
if ((null_value= args[0]->null_value || args[1]->null_value))
2005-02-08 23:50:45 +01:00
return 0;
/*
First check whether the result can be represented as a
(bool unsigned_flag, longlong value) pair, then check if it is compatible
with this Item's unsigned_flag by calling check_integer_overflow().
*/
if (args[0]->unsigned_flag)
{
if (args[1]->unsigned_flag || val1 >= 0)
{
if (test_if_sum_overflows_ull((ulonglong) val0, (ulonglong) val1))
goto err;
res_unsigned= TRUE;
}
else
{
/* val1 is negative */
if ((ulonglong) val0 > (ulonglong) LONGLONG_MAX)
res_unsigned= TRUE;
}
}
else
{
if (args[1]->unsigned_flag)
{
if (val0 >= 0)
{
if (test_if_sum_overflows_ull((ulonglong) val0, (ulonglong) val1))
goto err;
res_unsigned= TRUE;
}
else
{
if ((ulonglong) val1 > (ulonglong) LONGLONG_MAX)
res_unsigned= TRUE;
}
}
else
{
if (val0 >=0 && val1 >= 0)
res_unsigned= TRUE;
else if (val0 < 0 && val1 < 0 && res >= 0)
goto err;
}
}
return check_integer_overflow(res, res_unsigned);
err:
return raise_integer_overflow();
2005-02-08 23:50:45 +01:00
}
2007-10-11 19:29:09 +02:00
/**
Calculate plus of two decimals.
2007-10-11 19:29:09 +02:00
@param decimal_value Buffer that can be used to store result
2007-10-11 19:29:09 +02:00
@retval
0 Value was NULL; In this case null_value is set
@retval
\# Value of operation as a decimal
*/
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_plus::decimal_op(my_decimal *decimal_value)
{
my_decimal value1, *val1;
my_decimal value2, *val2;
val1= args[0]->val_decimal(&value1);
2005-02-08 23:50:45 +01:00
if ((null_value= args[0]->null_value))
return 0;
val2= args[1]->val_decimal(&value2);
if (!(null_value= (args[1]->null_value ||
check_decimal_overflow(my_decimal_add(E_DEC_FATAL_ERROR &
~E_DEC_OVERFLOW,
decimal_value,
val1, val2)) > 3)))
return decimal_value;
return 0;
2005-02-08 23:50:45 +01:00
}
2007-10-11 19:29:09 +02:00
/**
2005-02-08 23:50:45 +01:00
Set precision of results for additive operations (+ and -)
*/
void Item_func_additive_op::result_precision()
{
decimals= max(args[0]->decimals, args[1]->decimals);
int arg1_int= args[0]->decimal_precision() - args[0]->decimals;
int arg2_int= args[1]->decimal_precision() - args[1]->decimals;
int precision= max(arg1_int, arg2_int) + 1 + decimals;
/* Integer operations keep unsigned_flag if one of arguments is unsigned */
if (result_type() == INT_RESULT)
unsigned_flag= args[0]->unsigned_flag | args[1]->unsigned_flag;
else
unsigned_flag= args[0]->unsigned_flag & args[1]->unsigned_flag;
max_length= my_decimal_precision_to_length_no_truncation(precision, decimals,
unsigned_flag);
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/**
The following function is here to allow the user to force
subtraction of UNSIGNED BIGINT to return negative values.
*/
void Item_func_minus::fix_length_and_dec()
{
Item_num_op::fix_length_and_dec();
if (unsigned_flag &&
(current_thd->variables.sql_mode & MODE_NO_UNSIGNED_SUBTRACTION))
unsigned_flag=0;
}
2005-02-08 23:50:45 +01:00
double Item_func_minus::real_op()
2000-07-31 21:29:14 +02:00
{
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real() - args[1]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value || args[1]->null_value))
return 0.0;
return check_float_overflow(value);
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
longlong Item_func_minus::int_op()
2000-07-31 21:29:14 +02:00
{
longlong val0= args[0]->val_int();
longlong val1= args[1]->val_int();
longlong res= val0 - val1;
bool res_unsigned= FALSE;
if ((null_value= args[0]->null_value || args[1]->null_value))
2005-02-08 23:50:45 +01:00
return 0;
/*
First check whether the result can be represented as a
(bool unsigned_flag, longlong value) pair, then check if it is compatible
with this Item's unsigned_flag by calling check_integer_overflow().
*/
if (args[0]->unsigned_flag)
{
if (args[1]->unsigned_flag)
{
if ((ulonglong) val0 < (ulonglong) val1)
{
if (res >= 0)
goto err;
}
else
res_unsigned= TRUE;
}
else
{
if (val1 >= 0)
{
if ((ulonglong) val0 > (ulonglong) val1)
res_unsigned= TRUE;
}
else
{
if (test_if_sum_overflows_ull((ulonglong) val0, (ulonglong) -val1))
goto err;
res_unsigned= TRUE;
}
}
}
else
{
if (args[1]->unsigned_flag)
{
if ((ulonglong) (val0 - LONGLONG_MIN) < (ulonglong) val1)
goto err;
}
else
{
if (val0 > 0 && val1 < 0)
res_unsigned= TRUE;
else if (val0 < 0 && val1 > 0 && res >= 0)
goto err;
}
}
return check_integer_overflow(res, res_unsigned);
err:
return raise_integer_overflow();
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/**
See Item_func_plus::decimal_op for comments.
*/
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_minus::decimal_op(my_decimal *decimal_value)
{
my_decimal value1, *val1;
my_decimal value2, *val2=
val1= args[0]->val_decimal(&value1);
2005-02-08 23:50:45 +01:00
if ((null_value= args[0]->null_value))
return 0;
val2= args[1]->val_decimal(&value2);
if (!(null_value= (args[1]->null_value ||
(check_decimal_overflow(my_decimal_sub(E_DEC_FATAL_ERROR &
~E_DEC_OVERFLOW,
decimal_value, val1,
val2)) > 3))))
return decimal_value;
return 0;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
double Item_func_mul::real_op()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real() * args[1]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value || args[1]->null_value))
2005-02-08 23:50:45 +01:00
return 0.0;
return check_float_overflow(value);
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
longlong Item_func_mul::int_op()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
longlong a= args[0]->val_int();
longlong b= args[1]->val_int();
longlong res;
ulonglong res0, res1;
ulong a0, a1, b0, b1;
bool res_unsigned= FALSE;
bool a_negative= FALSE, b_negative= FALSE;
if ((null_value= args[0]->null_value || args[1]->null_value))
2005-02-08 23:50:45 +01:00
return 0;
/*
First check whether the result can be represented as a
(bool unsigned_flag, longlong value) pair, then check if it is compatible
with this Item's unsigned_flag by calling check_integer_overflow().
Let a = a1 * 2^32 + a0 and b = b1 * 2^32 + b0. Then
a * b = (a1 * 2^32 + a0) * (b1 * 2^32 + b0) = a1 * b1 * 2^64 +
+ (a1 * b0 + a0 * b1) * 2^32 + a0 * b0;
We can determine if the above sum overflows the ulonglong range by
sequentially checking the following conditions:
1. If both a1 and b1 are non-zero.
2. Otherwise, if (a1 * b0 + a0 * b1) is greater than ULONG_MAX.
3. Otherwise, if (a1 * b0 + a0 * b1) * 2^32 + a0 * b0 is greater than
ULONGLONG_MAX.
Since we also have to take the unsigned_flag for a and b into account,
it is easier to first work with absolute values and set the
correct sign later.
*/
if (!args[0]->unsigned_flag && a < 0)
{
a_negative= TRUE;
a= -a;
}
if (!args[1]->unsigned_flag && b < 0)
{
b_negative= TRUE;
b= -b;
}
a0= 0xFFFFFFFFUL & a;
a1= ((ulonglong) a) >> 32;
b0= 0xFFFFFFFFUL & b;
b1= ((ulonglong) b) >> 32;
if (a1 && b1)
goto err;
res1= (ulonglong) a1 * b0 + (ulonglong) a0 * b1;
if (res1 > 0xFFFFFFFFUL)
goto err;
res1= res1 << 32;
res0= (ulonglong) a0 * b0;
if (test_if_sum_overflows_ull(res1, res0))
goto err;
res= res1 + res0;
if (a_negative != b_negative)
{
if ((ulonglong) res > (ulonglong) LONGLONG_MIN + 1)
goto err;
res= -res;
}
else
res_unsigned= TRUE;
return check_integer_overflow(res, res_unsigned);
err:
return raise_integer_overflow();
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/** See Item_func_plus::decimal_op for comments. */
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_mul::decimal_op(my_decimal *decimal_value)
{
my_decimal value1, *val1;
my_decimal value2, *val2;
val1= args[0]->val_decimal(&value1);
2005-02-08 23:50:45 +01:00
if ((null_value= args[0]->null_value))
return 0;
val2= args[1]->val_decimal(&value2);
if (!(null_value= (args[1]->null_value ||
(check_decimal_overflow(my_decimal_mul(E_DEC_FATAL_ERROR &
~E_DEC_OVERFLOW,
decimal_value, val1,
val2)) > 3))))
return decimal_value;
return 0;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
void Item_func_mul::result_precision()
{
/* Integer operations keep unsigned_flag if one of arguments is unsigned */
if (result_type() == INT_RESULT)
unsigned_flag= args[0]->unsigned_flag | args[1]->unsigned_flag;
else
unsigned_flag= args[0]->unsigned_flag & args[1]->unsigned_flag;
decimals= min(args[0]->decimals + args[1]->decimals, DECIMAL_MAX_SCALE);
uint est_prec = args[0]->decimal_precision() + args[1]->decimal_precision();
uint precision= min(est_prec, DECIMAL_MAX_PRECISION);
max_length= my_decimal_precision_to_length_no_truncation(precision, decimals,
unsigned_flag);
2005-02-08 23:50:45 +01:00
}
double Item_func_div::real_op()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
double val2= args[1]->val_real();
if ((null_value= args[0]->null_value || args[1]->null_value))
return 0.0;
if (val2 == 0.0)
{
signal_divide_by_null();
2000-07-31 21:29:14 +02:00
return 0.0;
}
return check_float_overflow(value/val2);
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_div::decimal_op(my_decimal *decimal_value)
2000-07-31 21:29:14 +02:00
{
my_decimal value1, *val1;
my_decimal value2, *val2;
2005-10-26 08:54:15 +02:00
int err;
val1= args[0]->val_decimal(&value1);
2005-02-08 23:50:45 +01:00
if ((null_value= args[0]->null_value))
return 0;
val2= args[1]->val_decimal(&value2);
2005-02-08 23:50:45 +01:00
if ((null_value= args[1]->null_value))
return 0;
if ((err= check_decimal_overflow(my_decimal_div(E_DEC_FATAL_ERROR &
~E_DEC_OVERFLOW &
~E_DEC_DIV_ZERO,
decimal_value,
val1, val2,
prec_increment))) > 3)
2005-10-26 08:54:15 +02:00
{
if (err == E_DEC_DIV_ZERO)
signal_divide_by_null();
null_value= 1;
2005-02-08 23:50:45 +01:00
return 0;
}
2005-10-26 08:54:15 +02:00
return decimal_value;
2005-02-08 23:50:45 +01:00
}
void Item_func_div::result_precision()
{
uint precision=min(args[0]->decimal_precision() +
args[1]->decimals + prec_increment,
DECIMAL_MAX_PRECISION);
/* Integer operations keep unsigned_flag if one of arguments is unsigned */
if (result_type() == INT_RESULT)
unsigned_flag= args[0]->unsigned_flag | args[1]->unsigned_flag;
else
unsigned_flag= args[0]->unsigned_flag & args[1]->unsigned_flag;
decimals= min(args[0]->decimals + prec_increment, DECIMAL_MAX_SCALE);
max_length= my_decimal_precision_to_length_no_truncation(precision, decimals,
unsigned_flag);
2000-07-31 21:29:14 +02:00
}
2000-07-31 21:29:14 +02:00
void Item_func_div::fix_length_and_dec()
{
2005-02-08 23:50:45 +01:00
DBUG_ENTER("Item_func_div::fix_length_and_dec");
prec_increment= current_thd->variables.div_precincrement;
2005-05-13 11:29:30 +02:00
Item_num_op::fix_length_and_dec();
switch(hybrid_type) {
2005-02-08 23:50:45 +01:00
case REAL_RESULT:
{
decimals=max(args[0]->decimals,args[1]->decimals)+prec_increment;
2005-02-08 23:50:45 +01:00
set_if_smaller(decimals, NOT_FIXED_DEC);
uint tmp=float_length(decimals);
if (decimals == NOT_FIXED_DEC)
max_length= tmp;
else
{
max_length=args[0]->max_length - args[0]->decimals + decimals;
set_if_smaller(max_length,tmp);
}
2005-02-08 23:50:45 +01:00
break;
}
case INT_RESULT:
hybrid_type= DECIMAL_RESULT;
DBUG_PRINT("info", ("Type changed: DECIMAL_RESULT"));
result_precision();
break;
case DECIMAL_RESULT:
result_precision();
break;
default:
DBUG_ASSERT(0);
}
maybe_null= 1; // devision by zero
DBUG_VOID_RETURN;
2000-07-31 21:29:14 +02:00
}
/* Integer division */
longlong Item_func_int_div::val_int()
{
DBUG_ASSERT(fixed == 1);
/*
Perform division using DECIMAL math if either of the operands has a
non-integer type
*/
if (args[0]->result_type() != INT_RESULT ||
args[1]->result_type() != INT_RESULT)
{
my_decimal tmp;
my_decimal *val0p= args[0]->val_decimal(&tmp);
if ((null_value= args[0]->null_value))
return 0;
my_decimal val0= *val0p;
my_decimal *val1p= args[1]->val_decimal(&tmp);
if ((null_value= args[1]->null_value))
return 0;
my_decimal val1= *val1p;
int err;
if ((err= my_decimal_div(E_DEC_FATAL_ERROR & ~E_DEC_DIV_ZERO, &tmp,
&val0, &val1, 0)) > 3)
{
if (err == E_DEC_DIV_ZERO)
signal_divide_by_null();
return 0;
}
my_decimal truncated;
const bool do_truncate= true;
if (my_decimal_round(E_DEC_FATAL_ERROR, &tmp, 0, do_truncate, &truncated))
DBUG_ASSERT(false);
longlong res;
if (my_decimal2int(E_DEC_FATAL_ERROR, &truncated, unsigned_flag, &res) &
E_DEC_OVERFLOW)
raise_integer_overflow();
return res;
}
longlong val0=args[0]->val_int();
longlong val1=args[1]->val_int();
bool val0_negative, val1_negative, res_negative;
ulonglong uval0, uval1, res;
if ((null_value= (args[0]->null_value || args[1]->null_value)))
return 0;
if (val1 == 0)
{
signal_divide_by_null();
return 0;
}
val0_negative= !args[0]->unsigned_flag && val0 < 0;
val1_negative= !args[1]->unsigned_flag && val1 < 0;
res_negative= val0_negative != val1_negative;
uval0= (ulonglong) (val0_negative ? -val0 : val0);
uval1= (ulonglong) (val1_negative ? -val1 : val1);
res= uval0 / uval1;
if (res_negative)
{
if (res > (ulonglong) LONGLONG_MAX)
return raise_integer_overflow();
res= (ulonglong) (-(longlong) res);
}
return check_integer_overflow(res, !res_negative);
}
void Item_func_int_div::fix_length_and_dec()
{
Item_result argtype= args[0]->result_type();
/* use precision ony for the data type it is applicable for and valid */
max_length=args[0]->max_length -
(argtype == DECIMAL_RESULT || argtype == INT_RESULT ?
args[0]->decimals : 0);
maybe_null=1;
2005-02-08 23:50:45 +01:00
unsigned_flag=args[0]->unsigned_flag | args[1]->unsigned_flag;
}
2005-02-08 23:50:45 +01:00
longlong Item_func_mod::int_op()
{
DBUG_ASSERT(fixed == 1);
longlong val0= args[0]->val_int();
longlong val1= args[1]->val_int();
bool val0_negative, val1_negative;
ulonglong uval0, uval1;
ulonglong res;
2005-02-08 23:50:45 +01:00
if ((null_value= args[0]->null_value || args[1]->null_value))
return 0; /* purecov: inspected */
if (val1 == 0)
2005-02-08 23:50:45 +01:00
{
signal_divide_by_null();
return 0;
}
/*
'%' is calculated by integer division internally. Since dividing
LONGLONG_MIN by -1 generates SIGFPE, we calculate using unsigned values and
then adjust the sign appropriately.
*/
val0_negative= !args[0]->unsigned_flag && val0 < 0;
val1_negative= !args[1]->unsigned_flag && val1 < 0;
uval0= (ulonglong) (val0_negative ? -val0 : val0);
uval1= (ulonglong) (val1_negative ? -val1 : val1);
res= uval0 % uval1;
return check_integer_overflow(val0_negative ? -(longlong) res : res,
!val0_negative);
2005-02-08 23:50:45 +01:00
}
double Item_func_mod::real_op()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
double val2= args[1]->val_real();
if ((null_value= args[0]->null_value || args[1]->null_value))
2000-07-31 21:29:14 +02:00
return 0.0; /* purecov: inspected */
if (val2 == 0.0)
{
signal_divide_by_null();
return 0.0;
}
2000-07-31 21:29:14 +02:00
return fmod(value,val2);
}
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_mod::decimal_op(my_decimal *decimal_value)
2000-07-31 21:29:14 +02:00
{
my_decimal value1, *val1;
my_decimal value2, *val2;
val1= args[0]->val_decimal(&value1);
2005-02-08 23:50:45 +01:00
if ((null_value= args[0]->null_value))
return 0;
val2= args[1]->val_decimal(&value2);
2005-02-08 23:50:45 +01:00
if ((null_value= args[1]->null_value))
return 0;
switch (my_decimal_mod(E_DEC_FATAL_ERROR & ~E_DEC_DIV_ZERO, decimal_value,
val1, val2)) {
2005-02-08 23:50:45 +01:00
case E_DEC_TRUNCATED:
case E_DEC_OK:
return decimal_value;
case E_DEC_DIV_ZERO:
signal_divide_by_null();
2005-02-08 23:50:45 +01:00
default:
null_value= 1;
return 0;
}
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
void Item_func_mod::result_precision()
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
decimals= max(args[0]->decimals, args[1]->decimals);
max_length= max(args[0]->max_length, args[1]->max_length);
2000-07-31 21:29:14 +02:00
}
void Item_func_mod::fix_length_and_dec()
{
Item_num_op::fix_length_and_dec();
maybe_null= 1;
unsigned_flag= args[0]->unsigned_flag;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
double Item_func_neg::real_op()
2000-07-31 21:29:14 +02:00
{
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2005-02-08 23:50:45 +01:00
null_value= args[0]->null_value;
2000-07-31 21:29:14 +02:00
return -value;
}
2003-08-07 19:42:26 +02:00
2005-02-08 23:50:45 +01:00
longlong Item_func_neg::int_op()
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
longlong value= args[0]->val_int();
if ((null_value= args[0]->null_value))
return 0;
if (args[0]->unsigned_flag &&
(ulonglong) value > (ulonglong) LONGLONG_MAX + 1)
return raise_integer_overflow();
return check_integer_overflow(-value, !args[0]->unsigned_flag && value < 0);
2000-07-31 21:29:14 +02:00
}
2003-08-07 19:42:26 +02:00
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_neg::decimal_op(my_decimal *decimal_value)
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
my_decimal val, *value= args[0]->val_decimal(&val);
if (!(null_value= args[0]->null_value))
{
my_decimal2decimal(value, decimal_value);
my_decimal_neg(decimal_value);
return decimal_value;
2005-02-08 23:50:45 +01:00
}
return 0;
2005-02-08 23:50:45 +01:00
}
2005-02-08 23:50:45 +01:00
void Item_func_neg::fix_num_length_and_dec()
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
decimals= args[0]->decimals;
/* 1 add because sign can appear */
max_length= args[0]->max_length + 1;
2000-07-31 21:29:14 +02:00
}
2003-08-07 19:42:26 +02:00
void Item_func_neg::fix_length_and_dec()
2000-07-31 21:29:14 +02:00
{
DBUG_ENTER("Item_func_neg::fix_length_and_dec");
2005-02-08 23:50:45 +01:00
Item_func_num1::fix_length_and_dec();
/*
If this is in integer context keep the context as integer if possible
(This is how multiplication and other integer functions works)
Use val() to get value as arg_type doesn't mean that item is
Item_int or Item_real due to existence of Item_param.
*/
if (hybrid_type == INT_RESULT && args[0]->const_item())
2003-08-07 19:42:26 +02:00
{
longlong val= args[0]->val_int();
if ((ulonglong) val >= (ulonglong) LONGLONG_MIN &&
((ulonglong) val != (ulonglong) LONGLONG_MIN ||
args[0]->type() != INT_ITEM))
{
/*
Ensure that result is converted to DECIMAL, as longlong can't hold
the negated number
*/
hybrid_type= DECIMAL_RESULT;
DBUG_PRINT("info", ("Type changed: DECIMAL_RESULT"));
}
2003-08-07 19:42:26 +02:00
}
unsigned_flag= 0;
2005-02-08 23:50:45 +01:00
DBUG_VOID_RETURN;
2000-07-31 21:29:14 +02:00
}
2003-08-07 19:42:26 +02:00
2005-02-08 23:50:45 +01:00
double Item_func_abs::real_op()
2000-07-31 21:29:14 +02:00
{
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2005-02-08 23:50:45 +01:00
null_value= args[0]->null_value;
2000-07-31 21:29:14 +02:00
return fabs(value);
}
2003-08-07 19:42:26 +02:00
2005-02-08 23:50:45 +01:00
longlong Item_func_abs::int_op()
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
longlong value= args[0]->val_int();
if ((null_value= args[0]->null_value))
return 0;
if (unsigned_flag)
return value;
/* -LONGLONG_MIN = LONGLONG_MAX + 1 => outside of signed longlong range */
if (value == LONGLONG_MIN)
return raise_integer_overflow();
return (value >= 0) ? value : -value;
2000-07-31 21:29:14 +02:00
}
2003-08-07 19:42:26 +02:00
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_abs::decimal_op(my_decimal *decimal_value)
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
my_decimal val, *value= args[0]->val_decimal(&val);
if (!(null_value= args[0]->null_value))
2003-08-07 19:42:26 +02:00
{
2005-02-08 23:50:45 +01:00
my_decimal2decimal(value, decimal_value);
if (decimal_value->sign())
my_decimal_neg(decimal_value);
return decimal_value;
2003-08-07 19:42:26 +02:00
}
return 0;
2005-02-08 23:50:45 +01:00
}
void Item_func_abs::fix_length_and_dec()
{
Item_func_num1::fix_length_and_dec();
unsigned_flag= args[0]->unsigned_flag;
2000-07-31 21:29:14 +02:00
}
2003-08-07 19:42:26 +02:00
2007-10-11 19:29:09 +02:00
/** Gateway to natural LOG function. */
2004-11-11 19:39:35 +01:00
double Item_func_ln::val_real()
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
if ((null_value= args[0]->null_value))
return 0.0;
if (value <= 0.0)
{
signal_divide_by_null();
return 0.0;
}
return log(value);
}
2007-10-11 19:29:09 +02:00
/**
Extended but so slower LOG function.
We have to check if all values are > zero and first one is not one
as these are the cases then result is not a number.
*/
2004-11-11 19:39:35 +01:00
double Item_func_log::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
if ((null_value= args[0]->null_value))
return 0.0;
if (value <= 0.0)
{
signal_divide_by_null();
return 0.0;
}
if (arg_count == 2)
{
2004-11-11 19:39:35 +01:00
double value2= args[1]->val_real();
if ((null_value= args[1]->null_value))
return 0.0;
if (value2 <= 0.0 || value == 1.0)
{
signal_divide_by_null();
return 0.0;
}
return log(value2) / log(value);
}
2000-07-31 21:29:14 +02:00
return log(value);
}
2004-11-11 19:39:35 +01:00
double Item_func_log2::val_real()
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
if ((null_value=args[0]->null_value))
return 0.0;
if (value <= 0.0)
{
signal_divide_by_null();
return 0.0;
}
return log(value) / M_LN2;
}
2004-11-11 19:39:35 +01:00
double Item_func_log10::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
if ((null_value= args[0]->null_value))
return 0.0;
if (value <= 0.0)
{
signal_divide_by_null();
return 0.0;
}
2000-07-31 21:29:14 +02:00
return log10(value);
}
2004-11-11 19:39:35 +01:00
double Item_func_exp::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value))
return 0.0; /* purecov: inspected */
return check_float_overflow(exp(value));
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_sqrt::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=(args[0]->null_value || value < 0)))
return 0.0; /* purecov: inspected */
return sqrt(value);
}
2004-11-11 19:39:35 +01:00
double Item_func_pow::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
double val2= args[1]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=(args[0]->null_value || args[1]->null_value)))
return 0.0; /* purecov: inspected */
return check_float_overflow(pow(value,val2));
2000-07-31 21:29:14 +02:00
}
// Trigonometric functions
2004-11-11 19:39:35 +01:00
double Item_func_acos::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
/* One can use this to defer SELECT processing. */
DEBUG_SYNC(current_thd, "before_acos_function");
2004-02-03 15:21:48 +01:00
// the volatile's for BUG #2338 to calm optimizer down (because of gcc's bug)
2004-11-11 19:39:35 +01:00
volatile double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=(args[0]->null_value || (value < -1.0 || value > 1.0))))
return 0.0;
return acos(value);
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_asin::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-02-03 15:21:48 +01:00
// the volatile's for BUG #2338 to calm optimizer down (because of gcc's bug)
2004-11-11 19:39:35 +01:00
volatile double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=(args[0]->null_value || (value < -1.0 || value > 1.0))))
return 0.0;
return asin(value);
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_atan::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value))
return 0.0;
if (arg_count == 2)
{
2004-11-11 19:39:35 +01:00
double val2= args[1]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[1]->null_value))
return 0.0;
return check_float_overflow(atan2(value,val2));
2000-07-31 21:29:14 +02:00
}
return atan(value);
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_cos::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value))
return 0.0;
return cos(value);
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_sin::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value))
return 0.0;
return sin(value);
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_tan::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value))
return 0.0;
return check_float_overflow(tan(value));
}
double Item_func_cot::val_real()
{
DBUG_ASSERT(fixed == 1);
double value= args[0]->val_real();
if ((null_value=args[0]->null_value))
return 0.0;
return check_float_overflow(1.0 / tan(value));
2000-07-31 21:29:14 +02:00
}
// Shift-functions, same as << and >> in C/C++
longlong Item_func_shift_left::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
uint shift;
ulonglong res= ((ulonglong) args[0]->val_int() <<
(shift=(uint) args[1]->val_int()));
if (args[0]->null_value || args[1]->null_value)
{
null_value=1;
return 0;
}
null_value=0;
return (shift < sizeof(longlong)*8 ? (longlong) res : LL(0));
}
longlong Item_func_shift_right::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
uint shift;
ulonglong res= (ulonglong) args[0]->val_int() >>
(shift=(uint) args[1]->val_int());
if (args[0]->null_value || args[1]->null_value)
{
null_value=1;
return 0;
}
null_value=0;
return (shift < sizeof(longlong)*8 ? (longlong) res : LL(0));
}
longlong Item_func_bit_neg::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
ulonglong res= (ulonglong) args[0]->val_int();
if ((null_value=args[0]->null_value))
return 0;
return ~res;
}
// Conversion functions
void Item_func_integer::fix_length_and_dec()
{
max_length=args[0]->max_length - args[0]->decimals+1;
uint tmp=float_length(decimals);
set_if_smaller(max_length,tmp);
decimals=0;
}
2005-02-08 23:50:45 +01:00
void Item_func_int_val::fix_num_length_and_dec()
2000-07-31 21:29:14 +02:00
{
ulonglong tmp_max_length= (ulonglong ) args[0]->max_length -
(args[0]->decimals ? args[0]->decimals + 1 : 0) + 2;
2011-03-28 10:52:47 +02:00
max_length= tmp_max_length > (ulonglong) 4294967295U ?
(uint32) 4294967295U : (uint32) tmp_max_length;
2005-02-08 23:50:45 +01:00
uint tmp= float_length(decimals);
set_if_smaller(max_length,tmp);
decimals= 0;
}
void Item_func_int_val::find_num_type()
{
DBUG_ENTER("Item_func_int_val::find_num_type");
DBUG_PRINT("info", ("name %s", func_name()));
switch(hybrid_type= args[0]->result_type())
{
case STRING_RESULT:
case REAL_RESULT:
hybrid_type= REAL_RESULT;
max_length= float_length(decimals);
break;
case INT_RESULT:
case DECIMAL_RESULT:
/*
-2 because in most high position can't be used any digit for longlong
and one position for increasing value during operation
*/
if ((args[0]->max_length - args[0]->decimals) >=
(DECIMAL_LONGLONG_DIGITS - 2))
{
hybrid_type= DECIMAL_RESULT;
}
else
{
unsigned_flag= args[0]->unsigned_flag;
hybrid_type= INT_RESULT;
}
break;
default:
DBUG_ASSERT(0);
}
DBUG_PRINT("info", ("Type: %s",
(hybrid_type == REAL_RESULT ? "REAL_RESULT" :
hybrid_type == DECIMAL_RESULT ? "DECIMAL_RESULT" :
hybrid_type == INT_RESULT ? "INT_RESULT" :
"--ILLEGAL!!!--")));
DBUG_VOID_RETURN;
}
longlong Item_func_ceiling::int_op()
{
longlong result;
switch (args[0]->result_type()) {
case INT_RESULT:
result= args[0]->val_int();
null_value= args[0]->null_value;
break;
case DECIMAL_RESULT:
{
my_decimal dec_buf, *dec;
if ((dec= Item_func_ceiling::decimal_op(&dec_buf)))
my_decimal2int(E_DEC_FATAL_ERROR, dec, unsigned_flag, &result);
else
result= 0;
break;
}
default:
result= (longlong)Item_func_ceiling::real_op();
};
return result;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
double Item_func_ceiling::real_op()
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
/*
the volatile's for BUG #3051 to calm optimizer down (because of gcc's
bug)
*/
2004-11-11 19:39:35 +01:00
volatile double value= args[0]->val_real();
2005-02-08 23:50:45 +01:00
null_value= args[0]->null_value;
return ceil(value);
}
my_decimal *Item_func_ceiling::decimal_op(my_decimal *decimal_value)
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
my_decimal val, *value= args[0]->val_decimal(&val);
if (!(null_value= (args[0]->null_value ||
my_decimal_ceiling(E_DEC_FATAL_ERROR, value,
decimal_value) > 1)))
return decimal_value;
return 0;
2005-02-08 23:50:45 +01:00
}
longlong Item_func_floor::int_op()
{
longlong result;
switch (args[0]->result_type()) {
case INT_RESULT:
result= args[0]->val_int();
null_value= args[0]->null_value;
break;
case DECIMAL_RESULT:
{
my_decimal dec_buf, *dec;
if ((dec= Item_func_floor::decimal_op(&dec_buf)))
my_decimal2int(E_DEC_FATAL_ERROR, dec, unsigned_flag, &result);
else
result= 0;
break;
}
default:
result= (longlong)Item_func_floor::real_op();
};
return result;
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
double Item_func_floor::real_op()
2000-07-31 21:29:14 +02:00
{
2005-02-08 23:50:45 +01:00
/*
the volatile's for BUG #3051 to calm optimizer down (because of gcc's
bug)
*/
volatile double value= args[0]->val_real();
null_value= args[0]->null_value;
return floor(value);
}
my_decimal *Item_func_floor::decimal_op(my_decimal *decimal_value)
{
my_decimal val, *value= args[0]->val_decimal(&val);
if (!(null_value= (args[0]->null_value ||
my_decimal_floor(E_DEC_FATAL_ERROR, value,
decimal_value) > 1)))
return decimal_value;
return 0;
2005-02-08 23:50:45 +01:00
}
void Item_func_round::fix_length_and_dec()
2005-02-08 23:50:45 +01:00
{
int decimals_to_set;
longlong val1;
bool val1_unsigned;
unsigned_flag= args[0]->unsigned_flag;
if (!args[1]->const_item())
2000-07-31 21:29:14 +02:00
{
decimals= args[0]->decimals;
max_length= float_length(decimals);
if (args[0]->result_type() == DECIMAL_RESULT)
{
max_length++;
hybrid_type= DECIMAL_RESULT;
}
else
hybrid_type= REAL_RESULT;
return;
}
val1= args[1]->val_int();
if ((null_value= args[1]->is_null()))
return;
val1_unsigned= args[1]->unsigned_flag;
if (val1 < 0)
decimals_to_set= val1_unsigned ? INT_MAX : 0;
else
decimals_to_set= (val1 > INT_MAX) ? INT_MAX : (int) val1;
if (args[0]->decimals == NOT_FIXED_DEC)
{
decimals= min(decimals_to_set, NOT_FIXED_DEC);
max_length= float_length(decimals);
hybrid_type= REAL_RESULT;
return;
}
switch (args[0]->result_type()) {
case REAL_RESULT:
case STRING_RESULT:
hybrid_type= REAL_RESULT;
decimals= min(decimals_to_set, NOT_FIXED_DEC);
max_length= float_length(decimals);
break;
case INT_RESULT:
if ((!decimals_to_set && truncate) || (args[0]->decimal_precision() < DECIMAL_LONGLONG_DIGITS))
{
int length_can_increase= test(!truncate && (val1 < 0) && !val1_unsigned);
max_length= args[0]->max_length + length_can_increase;
/* Here we can keep INT_RESULT */
hybrid_type= INT_RESULT;
decimals= 0;
break;
}
/* fall through */
case DECIMAL_RESULT:
{
hybrid_type= DECIMAL_RESULT;
decimals_to_set= min(DECIMAL_MAX_SCALE, decimals_to_set);
int decimals_delta= args[0]->decimals - decimals_to_set;
int precision= args[0]->decimal_precision();
int length_increase= ((decimals_delta <= 0) || truncate) ? 0:1;
precision-= decimals_delta - length_increase;
decimals= min(decimals_to_set, DECIMAL_MAX_SCALE);
max_length= my_decimal_precision_to_length_no_truncation(precision,
decimals,
unsigned_flag);
break;
}
default:
DBUG_ASSERT(0); /* This result type isn't handled */
2000-07-31 21:29:14 +02:00
}
}
double my_double_round(double value, longlong dec, bool dec_unsigned,
bool truncate)
2000-07-31 21:29:14 +02:00
{
2003-03-10 11:00:19 +01:00
double tmp;
bool dec_negative= (dec < 0) && !dec_unsigned;
ulonglong abs_dec= dec_negative ? -dec : dec;
2003-03-10 11:00:19 +01:00
/*
tmp2 is here to avoid return the value with 80 bit precision
This will fix that the test round(0.1,1) = round(0.1,1) is true
Tagging with volatile is no guarantee, it may still be optimized away...
2003-03-10 11:00:19 +01:00
*/
volatile double tmp2;
2000-07-31 21:29:14 +02:00
2003-03-10 11:00:19 +01:00
tmp=(abs_dec < array_elements(log_10) ?
log_10[abs_dec] : pow(10.0,(double) abs_dec));
2000-07-31 21:29:14 +02:00
// Pre-compute these, to avoid optimizing away e.g. 'floor(v/tmp) * tmp'.
volatile double value_div_tmp= value / tmp;
volatile double value_mul_tmp= value * tmp;
if (dec_negative && my_isinf(tmp))
tmp2= 0.0;
else if (!dec_negative && my_isinf(value_mul_tmp))
tmp2= value;
else if (truncate)
{
if (value >= 0.0)
tmp2= dec < 0 ? floor(value_div_tmp) * tmp : floor(value_mul_tmp) / tmp;
else
tmp2= dec < 0 ? ceil(value_div_tmp) * tmp : ceil(value_mul_tmp) / tmp;
}
2003-03-10 11:00:19 +01:00
else
tmp2=dec < 0 ? rint(value_div_tmp) * tmp : rint(value_mul_tmp) / tmp;
2003-03-10 11:00:19 +01:00
return tmp2;
2000-07-31 21:29:14 +02:00
}
double Item_func_round::real_op()
{
double value= args[0]->val_real();
if (!(null_value= args[0]->null_value || args[1]->null_value))
return my_double_round(value, args[1]->val_int(), args[1]->unsigned_flag,
truncate);
return 0.0;
}
/*
Rounds a given value to a power of 10 specified as the 'to' argument,
avoiding overflows when the value is close to the ulonglong range boundary.
*/
static inline ulonglong my_unsigned_round(ulonglong value, ulonglong to)
{
ulonglong tmp= value / to * to;
return (value - tmp < (to >> 1)) ? tmp : tmp + to;
}
2005-02-08 23:50:45 +01:00
longlong Item_func_round::int_op()
{
longlong value= args[0]->val_int();
longlong dec= args[1]->val_int();
2005-02-08 23:50:45 +01:00
decimals= 0;
ulonglong abs_dec;
2005-02-08 23:50:45 +01:00
if ((null_value= args[0]->null_value || args[1]->null_value))
return 0;
if ((dec >= 0) || args[1]->unsigned_flag)
2005-02-08 23:50:45 +01:00
return value; // integer have not digits after point
abs_dec= -dec;
longlong tmp;
if(abs_dec >= array_elements(log_10_int))
return 0;
tmp= log_10_int[abs_dec];
2005-02-08 23:50:45 +01:00
if (truncate)
value= (unsigned_flag) ?
((ulonglong) value / tmp) * tmp : (value / tmp) * tmp;
2005-02-08 23:50:45 +01:00
else
value= (unsigned_flag || value >= 0) ?
my_unsigned_round((ulonglong) value, tmp) :
-(longlong) my_unsigned_round((ulonglong) -value, tmp);
return value;
2005-02-08 23:50:45 +01:00
}
my_decimal *Item_func_round::decimal_op(my_decimal *decimal_value)
{
my_decimal val, *value= args[0]->val_decimal(&val);
longlong dec= args[1]->val_int();
if (dec >= 0 || args[1]->unsigned_flag)
dec= min((ulonglong) dec, decimals);
else if (dec < INT_MIN)
dec= INT_MIN;
if (!(null_value= (args[0]->null_value || args[1]->null_value ||
my_decimal_round(E_DEC_FATAL_ERROR, value, (int) dec,
truncate, decimal_value) > 1)))
return decimal_value;
return 0;
2005-02-08 23:50:45 +01:00
}
void Item_func_rand::seed_random(Item *arg)
{
/*
TODO: do not do reinit 'rand' for every execute of PS/SP if
args[0] is a constant.
*/
uint32 tmp= (uint32) arg->val_int();
randominit(rand, (uint32) (tmp*0x10001L+55555555L),
(uint32) (tmp*0x10000001L));
}
bool Item_func_rand::fix_fields(THD *thd,Item **ref)
2000-07-31 21:29:14 +02:00
{
if (Item_real_func::fix_fields(thd, ref))
return TRUE;
used_tables_cache|= RAND_TABLE_BIT;
2000-07-31 21:29:14 +02:00
if (arg_count)
{ // Only use argument once in query
/*
Allocate rand structure once: we must use thd->stmt_arena
to create rand in proper mem_root if it's a prepared statement or
stored procedure.
No need to send a Rand log event if seed was given eg: RAND(seed),
as it will be replicated in the query as such.
*/
if (!rand && !(rand= (struct rand_struct*)
thd->stmt_arena->alloc(sizeof(*rand))))
return TRUE;
2000-07-31 21:29:14 +02:00
}
else
2002-10-25 00:48:34 +02:00
{
/*
Save the seed only the first time RAND() is used in the query
Once events are forwarded rather than recreated,
the following can be skipped if inside the slave thread
*/
if (!thd->rand_used)
{
thd->rand_used= 1;
thd->rand_saved_seed1= thd->rand.seed1;
thd->rand_saved_seed2= thd->rand.seed2;
}
rand= &thd->rand;
2002-10-25 00:48:34 +02:00
}
return FALSE;
}
2004-03-16 19:19:36 +01:00
void Item_func_rand::update_used_tables()
{
Item_real_func::update_used_tables();
used_tables_cache|= RAND_TABLE_BIT;
}
2004-11-11 19:39:35 +01:00
double Item_func_rand::val_real()
{
DBUG_ASSERT(fixed == 1);
if (arg_count)
{
if (!args[0]->const_item())
seed_random(args[0]);
else if (first_eval)
{
/*
Constantness of args[0] may be set during JOIN::optimize(), if arg[0]
is a field item of "constant" table. Thus, we have to evaluate
seed_random() for constant arg there but not at the fix_fields method.
*/
first_eval= FALSE;
seed_random(args[0]);
}
}
return my_rnd(rand);
2000-07-31 21:29:14 +02:00
}
longlong Item_func_sign::val_int()
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
null_value=args[0]->null_value;
return value < 0.0 ? -1 : (value > 0 ? 1 : 0);
}
2004-11-11 19:39:35 +01:00
double Item_func_units::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double value= args[0]->val_real();
2000-07-31 21:29:14 +02:00
if ((null_value=args[0]->null_value))
return 0;
return check_float_overflow(value * mul + add);
2000-07-31 21:29:14 +02:00
}
void Item_func_min_max::fix_length_and_dec()
{
int max_int_part=0;
bool datetime_found= FALSE;
2000-07-31 21:29:14 +02:00
decimals=0;
max_length=0;
maybe_null=0;
2000-07-31 21:29:14 +02:00
cmp_type=args[0]->result_type();
2000-07-31 21:29:14 +02:00
for (uint i=0 ; i < arg_count ; i++)
{
2005-02-08 23:50:45 +01:00
set_if_bigger(max_length, args[i]->max_length);
set_if_bigger(decimals, args[i]->decimals);
set_if_bigger(max_int_part, args[i]->decimal_int_part());
if (args[i]->maybe_null)
maybe_null=1;
2000-07-31 21:29:14 +02:00
cmp_type=item_cmp_type(cmp_type,args[i]->result_type());
if (args[i]->result_type() != ROW_RESULT && args[i]->is_datetime())
{
datetime_found= TRUE;
if (!datetime_item || args[i]->field_type() == MYSQL_TYPE_DATETIME)
datetime_item= args[i];
}
2000-07-31 21:29:14 +02:00
}
2003-07-04 17:19:07 +02:00
if (cmp_type == STRING_RESULT)
{
agg_arg_charsets_for_string_result_with_comparison(collation,
args, arg_count);
if (datetime_found)
{
thd= current_thd;
compare_as_dates= TRUE;
}
}
else if ((cmp_type == DECIMAL_RESULT) || (cmp_type == INT_RESULT))
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
{
collation.set_numeric();
fix_char_length(my_decimal_precision_to_length_no_truncation(max_int_part +
decimals,
decimals,
unsigned_flag));
}
Bug #54461: crash with longblob and union or update with subquery Queries may crash, if 1) the GREATEST or the LEAST function has a mixed list of numeric and LONGBLOB arguments and 2) the result of such a function goes through an intermediate temporary table. An Item that references a LONGBLOB field has max_length of UINT_MAX32 == (2^32 - 1). The current implementation of GREATEST/LEAST returns REAL result for a mixed list of numeric and string arguments (that contradicts with the current documentation, this contradiction was discussed and it was decided to update the documentation). The max_length of such a function call was calculated as a maximum of argument max_length values (i.e. UINT_MAX32). That max_length value of UINT_MAX32 was used as a length for the intermediate temporary table Field_double to hold GREATEST/LEAST function result. The Field_double::val_str() method call on that field allocates a String value. Since an allocation of String reserves an additional byte for a zero-termination, the size of String buffer was set to (UINT_MAX32 + 1), that caused an integer overflow: actually, an empty buffer of size 0 was allocated. An initialization of the "first" byte of that zero-size buffer with '\0' caused a crash. The Item_func_min_max::fix_length_and_dec() has been modified to calculate max_length for the REAL result like we do it for arithmetical operators. ****** Bug #54461: crash with longblob and union or update with subquery Queries may crash, if 1) the GREATEST or the LEAST function has a mixed list of numeric and LONGBLOB arguments and 2) the result of such a function goes through an intermediate temporary table. An Item that references a LONGBLOB field has max_length of UINT_MAX32 == (2^32 - 1). The current implementation of GREATEST/LEAST returns REAL result for a mixed list of numeric and string arguments (that contradicts with the current documentation, this contradiction was discussed and it was decided to update the documentation). The max_length of such a function call was calculated as a maximum of argument max_length values (i.e. UINT_MAX32). That max_length value of UINT_MAX32 was used as a length for the intermediate temporary table Field_double to hold GREATEST/LEAST function result. The Field_double::val_str() method call on that field allocates a String value. Since an allocation of String reserves an additional byte for a zero-termination, the size of String buffer was set to (UINT_MAX32 + 1), that caused an integer overflow: actually, an empty buffer of size 0 was allocated. An initialization of the "first" byte of that zero-size buffer with '\0' caused a crash. The Item_func_min_max::fix_length_and_dec() has been modified to calculate max_length for the REAL result like we do it for arithmetical operators.
2010-08-01 20:12:36 +02:00
else if (cmp_type == REAL_RESULT)
2010-08-20 13:22:46 +02:00
fix_char_length(float_length(decimals));
cached_field_type= agg_field_type(args, arg_count);
2000-07-31 21:29:14 +02:00
}
/*
Compare item arguments in the DATETIME context.
SYNOPSIS
cmp_datetimes()
value [out] found least/greatest DATE/DATETIME value
DESCRIPTION
Compare item arguments as DATETIME values and return the index of the
least/greatest argument in the arguments array.
The correct integer DATE/DATETIME value of the found argument is
stored to the value pointer, if latter is provided.
RETURN
0 If one of arguments is NULL or there was a execution error
# index of the least/greatest argument
*/
uint Item_func_min_max::cmp_datetimes(ulonglong *value)
{
longlong UNINIT_VAR(min_max);
uint min_max_idx= 0;
for (uint i=0; i < arg_count ; i++)
{
Item **arg= args + i;
bool is_null;
longlong res= get_datetime_value(thd, &arg, 0, datetime_item, &is_null);
/* Check if we need to stop (because of error or KILL) and stop the loop */
if (thd->is_error())
{
null_value= 1;
return 0;
}
if ((null_value= args[i]->null_value))
return 0;
if (i == 0 || (res < min_max ? cmp_sign : -cmp_sign) > 0)
{
min_max= res;
min_max_idx= i;
}
}
if (value)
{
*value= min_max;
if (datetime_item->field_type() == MYSQL_TYPE_DATE)
*value/= 1000000L;
}
return min_max_idx;
}
2000-07-31 21:29:14 +02:00
String *Item_func_min_max::val_str(String *str)
{
DBUG_ASSERT(fixed == 1);
if (compare_as_dates)
{
String *str_res;
uint min_max_idx= cmp_datetimes(NULL);
if (null_value)
return 0;
str_res= args[min_max_idx]->val_str(str);
if (args[min_max_idx]->null_value)
{
// check if the call to val_str() above returns a NULL value
null_value= 1;
return NULL;
}
str_res->set_charset(collation.collation);
return str_res;
}
2000-07-31 21:29:14 +02:00
switch (cmp_type) {
case INT_RESULT:
{
longlong nr=val_int();
if (null_value)
return 0;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_int(nr, unsigned_flag, collation.collation);
2000-07-31 21:29:14 +02:00
return str;
}
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
my_decimal dec_buf, *dec_val= val_decimal(&dec_buf);
if (null_value)
return 0;
my_decimal2string(E_DEC_FATAL_ERROR, dec_val, 0, 0, 0, str);
return str;
}
2000-07-31 21:29:14 +02:00
case REAL_RESULT:
{
2004-11-11 19:39:35 +01:00
double nr= val_real();
2000-07-31 21:29:14 +02:00
if (null_value)
return 0; /* purecov: inspected */
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_real(nr, decimals, collation.collation);
2000-07-31 21:29:14 +02:00
return str;
}
case STRING_RESULT:
{
String *UNINIT_VAR(res);
2000-07-31 21:29:14 +02:00
for (uint i=0; i < arg_count ; i++)
{
if (i == 0)
2000-07-31 21:29:14 +02:00
res=args[i]->val_str(str);
else
{
String *res2;
res2= args[i]->val_str(res == str ? &tmp_value : str);
if (res2)
{
int cmp= sortcmp(res,res2,collation.collation);
2000-07-31 21:29:14 +02:00
if ((cmp_sign < 0 ? cmp : -cmp) < 0)
res=res2;
}
}
if ((null_value= args[i]->null_value))
return 0;
2000-07-31 21:29:14 +02:00
}
res->set_charset(collation.collation);
2000-07-31 21:29:14 +02:00
return res;
}
case ROW_RESULT:
2003-01-31 11:07:07 +01:00
default:
2005-02-08 23:50:45 +01:00
// This case should never be chosen
2002-11-15 19:32:09 +01:00
DBUG_ASSERT(0);
return 0;
2000-07-31 21:29:14 +02:00
}
return 0; // Keep compiler happy
}
2004-11-11 19:39:35 +01:00
double Item_func_min_max::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
double value=0.0;
if (compare_as_dates)
{
2007-05-07 16:24:46 +02:00
ulonglong result= 0;
(void)cmp_datetimes(&result);
return (double)result;
}
2000-07-31 21:29:14 +02:00
for (uint i=0; i < arg_count ; i++)
{
if (i == 0)
2004-11-11 19:39:35 +01:00
value= args[i]->val_real();
2000-07-31 21:29:14 +02:00
else
{
2004-11-11 19:39:35 +01:00
double tmp= args[i]->val_real();
2000-07-31 21:29:14 +02:00
if (!args[i]->null_value && (tmp < value ? cmp_sign : -cmp_sign) > 0)
value=tmp;
}
if ((null_value= args[i]->null_value))
break;
2000-07-31 21:29:14 +02:00
}
return value;
}
longlong Item_func_min_max::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
longlong value=0;
if (compare_as_dates)
{
2007-05-08 08:42:40 +02:00
ulonglong result= 0;
(void)cmp_datetimes(&result);
return (longlong)result;
}
2000-07-31 21:29:14 +02:00
for (uint i=0; i < arg_count ; i++)
{
if (i == 0)
value=args[i]->val_int();
2000-07-31 21:29:14 +02:00
else
{
longlong tmp=args[i]->val_int();
if (!args[i]->null_value && (tmp < value ? cmp_sign : -cmp_sign) > 0)
value=tmp;
2000-07-31 21:29:14 +02:00
}
if ((null_value= args[i]->null_value))
break;
2000-07-31 21:29:14 +02:00
}
return value;
}
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_min_max::val_decimal(my_decimal *dec)
{
DBUG_ASSERT(fixed == 1);
my_decimal tmp_buf, *tmp, *UNINIT_VAR(res);
if (compare_as_dates)
{
2007-05-08 08:42:40 +02:00
ulonglong value= 0;
(void)cmp_datetimes(&value);
ulonglong2decimal(value, dec);
return dec;
}
2005-02-08 23:50:45 +01:00
for (uint i=0; i < arg_count ; i++)
{
if (i == 0)
2005-02-08 23:50:45 +01:00
res= args[i]->val_decimal(dec);
else
{
tmp= args[i]->val_decimal(&tmp_buf); // Zero if NULL
if (tmp && (my_decimal_cmp(tmp, res) * cmp_sign) < 0)
2005-02-08 23:50:45 +01:00
{
if (tmp == &tmp_buf)
{
/* Move value out of tmp_buf as this will be reused on next loop */
2005-02-08 23:50:45 +01:00
my_decimal2decimal(tmp, dec);
res= dec;
}
else
res= tmp;
}
}
if ((null_value= args[i]->null_value))
{
res= 0;
break;
}
2005-02-08 23:50:45 +01:00
}
return res;
}
2000-07-31 21:29:14 +02:00
longlong Item_func_length::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *res=args[0]->val_str(&value);
if (!res)
{
null_value=1;
return 0; /* purecov: inspected */
}
null_value=0;
return (longlong) res->length();
}
2003-08-07 19:42:26 +02:00
2000-07-31 21:29:14 +02:00
longlong Item_func_char_length::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *res=args[0]->val_str(&value);
if (!res)
{
null_value=1;
return 0; /* purecov: inspected */
}
null_value=0;
return (longlong) res->numchars();
2000-07-31 21:29:14 +02:00
}
2003-08-07 19:42:26 +02:00
longlong Item_func_coercibility::val_int()
{
DBUG_ASSERT(fixed == 1);
null_value= 0;
return (longlong) args[0]->collation.derivation;
}
2000-07-31 21:29:14 +02:00
2003-08-07 19:42:26 +02:00
void Item_func_locate::fix_length_and_dec()
{
max_length= MY_INT32_NUM_DECIMAL_DIGITS;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
agg_arg_charsets_for_comparison(cmp_collation, args, 2);
}
2003-08-07 19:42:26 +02:00
2000-07-31 21:29:14 +02:00
longlong Item_func_locate::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *a=args[0]->val_str(&value1);
String *b=args[1]->val_str(&value2);
if (!a || !b)
{
null_value=1;
return 0; /* purecov: inspected */
}
null_value=0;
/* must be longlong to avoid truncation */
longlong start= 0;
longlong start0= 0;
my_match_t match;
2000-07-31 21:29:14 +02:00
if (arg_count == 3)
{
start0= start= args[2]->val_int() - 1;
if ((start < 0) || (start > a->length()))
return 0;
/* start is now sufficiently valid to pass to charpos function */
start= a->charpos((int) start);
if (start + b->length() > a->length())
2000-07-31 21:29:14 +02:00
return 0;
}
2000-07-31 21:29:14 +02:00
if (!b->length()) // Found empty string at start
return start + 1;
if (!cmp_collation.collation->coll->instr(cmp_collation.collation,
a->ptr()+start,
(uint) (a->length()-start),
b->ptr(), b->length(),
&match, 1))
return 0;
return (longlong) match.mb_len + start0 + 1;
2000-07-31 21:29:14 +02:00
}
void Item_func_locate::print(String *str, enum_query_type query_type)
{
str->append(STRING_WITH_LEN("locate("));
args[1]->print(str, query_type);
str->append(',');
args[0]->print(str, query_type);
if (arg_count == 3)
{
str->append(',');
args[2]->print(str, query_type);
}
str->append(')');
}
2000-07-31 21:29:14 +02:00
longlong Item_func_field::val_int()
{
DBUG_ASSERT(fixed == 1);
if (cmp_type == STRING_RESULT)
{
String *field;
if (!(field= args[0]->val_str(&value)))
return 0;
for (uint i=1 ; i < arg_count ; i++)
{
String *tmp_value=args[i]->val_str(&tmp);
if (tmp_value && !sortcmp(field,tmp_value,cmp_collation.collation))
return (longlong) (i);
}
}
else if (cmp_type == INT_RESULT)
{
longlong val= args[0]->val_int();
2005-06-02 15:45:20 +02:00
if (args[0]->null_value)
return 0;
for (uint i=1; i < arg_count ; i++)
{
if (val == args[i]->val_int() && !args[i]->null_value)
return (longlong) (i);
}
}
2005-02-08 23:50:45 +01:00
else if (cmp_type == DECIMAL_RESULT)
{
my_decimal dec_arg_buf, *dec_arg,
dec_buf, *dec= args[0]->val_decimal(&dec_buf);
2005-06-02 15:45:20 +02:00
if (args[0]->null_value)
return 0;
2005-02-08 23:50:45 +01:00
for (uint i=1; i < arg_count; i++)
{
dec_arg= args[i]->val_decimal(&dec_arg_buf);
2005-06-06 23:31:53 +02:00
if (!args[i]->null_value && !my_decimal_cmp(dec_arg, dec))
2005-02-08 23:50:45 +01:00
return (longlong) (i);
}
}
else
2000-07-31 21:29:14 +02:00
{
2004-11-11 19:39:35 +01:00
double val= args[0]->val_real();
2005-06-02 15:45:20 +02:00
if (args[0]->null_value)
return 0;
for (uint i=1; i < arg_count ; i++)
{
2005-06-07 04:43:59 +02:00
if (val == args[i]->val_real() && !args[i]->null_value)
return (longlong) (i);
}
2000-07-31 21:29:14 +02:00
}
return 0;
}
2003-08-07 19:42:26 +02:00
void Item_func_field::fix_length_and_dec()
{
maybe_null=0; max_length=3;
cmp_type= args[0]->result_type();
for (uint i=1; i < arg_count ; i++)
cmp_type= item_cmp_type(cmp_type, args[i]->result_type());
if (cmp_type == STRING_RESULT)
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
agg_arg_charsets_for_comparison(cmp_collation, args, arg_count);
}
2000-07-31 21:29:14 +02:00
2000-07-31 21:29:14 +02:00
longlong Item_func_ascii::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *res=args[0]->val_str(&value);
if (!res)
{
null_value=1;
return 0;
}
null_value=0;
return (longlong) (res->length() ? (uchar) (*res)[0] : (uchar) 0);
}
longlong Item_func_ord::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *res=args[0]->val_str(&value);
if (!res)
{
null_value=1;
return 0;
}
null_value=0;
if (!res->length()) return 0;
#ifdef USE_MB
if (use_mb(res->charset()))
2000-07-31 21:29:14 +02:00
{
register const char *str=res->ptr();
register uint32 n=0, l=my_ismbchar(res->charset(),str,str+res->length());
if (!l)
return (longlong)((uchar) *str);
2000-07-31 21:29:14 +02:00
while (l--)
n=(n<<8)|(uint32)((uchar) *str++);
return (longlong) n;
}
#endif
return (longlong) ((uchar) (*res)[0]);
}
/* Search after a string in a string of strings separated by ',' */
/* Returns number of found type >= 1 or 0 if not found */
/* This optimizes searching in enums to bit testing! */
void Item_func_find_in_set::fix_length_and_dec()
{
decimals=0;
max_length=3; // 1-999
if (args[0]->const_item() && args[1]->type() == FIELD_ITEM)
{
Field *field= ((Item_field*) args[1])->field;
if (field->real_type() == MYSQL_TYPE_SET)
2000-07-31 21:29:14 +02:00
{
String *find=args[0]->val_str(&value);
if (find)
{
// find is not NULL pointer so args[0] is not a null-value
DBUG_ASSERT(!args[0]->null_value);
enum_value= find_type(((Field_enum*) field)->typelib,find->ptr(),
find->length(), 0);
2000-07-31 21:29:14 +02:00
enum_bit=0;
if (enum_value)
enum_bit=LL(1) << (enum_value-1);
}
}
}
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
agg_arg_charsets_for_comparison(cmp_collation, args, 2);
2000-07-31 21:29:14 +02:00
}
static const char separator=',';
longlong Item_func_find_in_set::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
if (enum_value)
{
// enum_value is set iff args[0]->const_item() in fix_length_and_dec().
DBUG_ASSERT(args[0]->const_item());
ulonglong tmp= (ulonglong) args[1]->val_int();
null_value= args[1]->null_value;
/*
No need to check args[0]->null_value since enum_value is set iff
args[0] is a non-null const item. Note: no DBUG_ASSERT on
args[0]->null_value here because args[0] may have been replaced
by an Item_cache on which val_int() has not been called. See
BUG#11766317
*/
if (!null_value)
2000-07-31 21:29:14 +02:00
{
if (tmp & enum_bit)
return enum_value;
2000-07-31 21:29:14 +02:00
}
return 0L;
}
String *find=args[0]->val_str(&value);
String *buffer=args[1]->val_str(&value2);
if (!find || !buffer)
{
null_value=1;
return 0; /* purecov: inspected */
}
null_value=0;
int diff;
if ((diff=buffer->length() - find->length()) >= 0)
{
my_wc_t wc= 0;
CHARSET_INFO *cs= cmp_collation.collation;
const char *str_begin= buffer->ptr();
const char *str_end= buffer->ptr();
const char *real_end= str_end+buffer->length();
const uchar *find_str= (const uchar *) find->ptr();
uint find_str_len= find->length();
int position= 0;
while (1)
2000-07-31 21:29:14 +02:00
{
int symbol_len;
if ((symbol_len= cs->cset->mb_wc(cs, &wc, (uchar*) str_end,
(uchar*) real_end)) > 0)
2000-07-31 21:29:14 +02:00
{
const char *substr_end= str_end + symbol_len;
bool is_last_item= (substr_end == real_end);
bool is_separator= (wc == (my_wc_t) separator);
if (is_separator || is_last_item)
{
position++;
if (is_last_item && !is_separator)
str_end= substr_end;
if (!my_strnncoll(cs, (const uchar *) str_begin,
(uint) (str_end - str_begin),
find_str, find_str_len))
return (longlong) position;
else
str_begin= substr_end;
}
str_end= substr_end;
2000-07-31 21:29:14 +02:00
}
else if (str_end - str_begin == 0 &&
find_str_len == 0 &&
wc == (my_wc_t) separator)
return (longlong) ++position;
else
return LL(0);
}
2000-07-31 21:29:14 +02:00
}
return 0;
}
longlong Item_func_bit_count::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
ulonglong value= (ulonglong) args[0]->val_int();
if ((null_value= args[0]->null_value))
2000-07-31 21:29:14 +02:00
return 0; /* purecov: inspected */
2003-05-04 18:43:07 +02:00
return (longlong) my_count_bits(value);
2000-07-31 21:29:14 +02:00
}
/****************************************************************************
** Functions to handle dynamic loadable functions
** Original source by: Alexis Mikhailov <root@medinf.chuvashia.su>
** Rewritten by monty.
2000-07-31 21:29:14 +02:00
****************************************************************************/
#ifdef HAVE_DLOPEN
void udf_handler::cleanup()
2000-07-31 21:29:14 +02:00
{
if (!not_original)
2000-07-31 21:29:14 +02:00
{
if (initialized)
2000-07-31 21:29:14 +02:00
{
if (u_d->func_deinit != NULL)
{
Udf_func_deinit deinit= u_d->func_deinit;
(*deinit)(&initid);
}
free_udf(u_d);
initialized= FALSE;
2000-07-31 21:29:14 +02:00
}
if (buffers) // Because of bug in ecc
delete [] buffers;
buffers= 0;
2000-07-31 21:29:14 +02:00
}
}
bool
udf_handler::fix_fields(THD *thd, Item_result_field *func,
2000-07-31 21:29:14 +02:00
uint arg_count, Item **arguments)
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar buff[STACK_BUFF_ALLOC]; // Max argument in function
2000-07-31 21:29:14 +02:00
DBUG_ENTER("Item_udf_func::fix_fields");
if (check_stack_overrun(thd, STACK_MIN_SIZE, buff))
DBUG_RETURN(TRUE); // Fatal error flag is set!
udf_func *tmp_udf=find_udf(u_d->name.str,(uint) u_d->name.length,1);
2000-07-31 21:29:14 +02:00
if (!tmp_udf)
{
Fix for BUG#11755168 '46895: test "outfile_loaddata" fails (reproducible)'. In sql_class.cc, 'row_count', of type 'ha_rows', was used as last argument for ER_TRUNCATED_WRONG_VALUE_FOR_FIELD which is "Incorrect %-.32s value: '%-.128s' for column '%.192s' at row %ld". So 'ha_rows' was used as 'long'. On SPARC32 Solaris builds, 'long' is 4 bytes and 'ha_rows' is 'longlong' i.e. 8 bytes. So the printf-like code was reading only the first 4 bytes. Because the CPU is big-endian, 1LL is 0x00 0x00 0x00 0x00 0x00 0x00 0x00 0x01 so the first four bytes yield 0. So the warning message had "row 0" instead of "row 1" in test outfile_loaddata.test: -Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 1 +Warning 1366 Incorrect string value: '\xE1\xE2\xF7' for column 'b' at row 0 All error-messaging functions which internally invoke some printf-life function are potential candidate for such mistakes. One apparently easy way to catch such mistakes is to use ATTRIBUTE_FORMAT (from my_attribute.h). But this works only when call site has both: a) the format as a string literal b) the types of arguments. So: func(ER(ER_BLAH), 10); will silently not be checked, because ER(ER_BLAH) is not known at compile time (it is known at run-time, and depends on the chosen language). And func("%s", a va_list argument); has the same problem, as the *real* type of arguments is not known at this site at compile time (it's known in some caller). Moreover, func(ER(ER_BLAH)); though possibly correct (if ER(ER_BLAH) has no '%' markers), will not compile (gcc says "error: format not a string literal and no format arguments"). Consequences: 1) ATTRIBUTE_FORMAT is here added only to functions which in practice take "string literal" formats: "my_error_reporter" and "print_admin_msg". 2) it cannot be added to the other functions: my_error(), push_warning_printf(), Table_check_intact::report_error(), general_log_print(). To do a one-time check of functions listed in (2), the following "static code analysis" has been done: 1) replace my_error(ER_xxx, arguments for substitution in format) with the equivalent my_printf_error(ER_xxx,ER(ER_xxx), arguments for substitution in format), so that we have ER(ER_xxx) and the arguments *in the same call site* 2) add ATTRIBUTE_FORMAT to push_warning_printf(), Table_check_intact::report_error(), general_log_print() 3) replace ER(xxx) with the hard-coded English text found in errmsg.txt (like: ER(ER_UNKNOWN_ERROR) is replaced with "Unknown error"), so that a call site has the format as string literal 4) this way, ATTRIBUTE_FORMAT can effectively do its job 5) compile, fix errors detected by ATTRIBUTE_FORMAT 6) revert steps 1-2-3. The present patch has no compiler error when submitted again to the static code analysis above. It cannot catch all problems though: see Field::set_warning(), in which a call to push_warning_printf() has a variable error (thus, not replacable by a string literal); I checked set_warning() calls by hand though. See also WL 5883 for one proposal to avoid such bugs from appearing again in the future. The issues fixed in the patch are: a) mismatch in types (like 'int' passed to '%ld') b) more arguments passed than specified in the format. This patch resolves mismatches by changing the type/number of arguments, not by changing error messages of sql/share/errmsg.txt. The latter would be wrong, per the following old rule: errmsg.txt must be as stable as possible; no insertions or deletions of messages, no changes of type or number of printf-like format specifiers, are allowed, as long as the change impacts a message already released in a GA version. If this rule is not followed: - Connectors, which use error message numbers, will be confused (by insertions/deletions of messages) - using errmsg.sys of MySQL 5.1.n with mysqld of MySQL 5.1.(n+1) could produce wrong messages or crash; such usage can easily happen if installing 5.1.(n+1) while /etc/my.cnf still has --language=/path/to/5.1.n/xxx; or if copying mysqld from 5.1.(n+1) into a 5.1.n installation. When fixing b), I have verified that the superfluous arguments were not used in the format in the first 5.1 GA (5.1.30 'bteam@astra04-20081114162938-z8mctjp6st27uobm'). Had they been used, then passing them today, even if the message doesn't use them anymore, would have been necessary, as explained above.
2011-05-16 22:04:01 +02:00
my_error(ER_CANT_FIND_UDF, MYF(0), u_d->name.str);
DBUG_RETURN(TRUE);
2000-07-31 21:29:14 +02:00
}
u_d=tmp_udf;
args=arguments;
/* Fix all arguments */
func->maybe_null=0;
2000-07-31 21:29:14 +02:00
used_tables_cache=0;
const_item_cache=1;
if ((f_args.arg_count=arg_count))
{
if (!(f_args.arg_type= (Item_result*)
sql_alloc(f_args.arg_count*sizeof(Item_result))))
{
free_udf(u_d);
DBUG_RETURN(TRUE);
2000-07-31 21:29:14 +02:00
}
uint i;
Item **arg,**arg_end;
for (i=0, arg=arguments, arg_end=arguments+arg_count;
arg != arg_end ;
arg++,i++)
{
2005-02-08 23:50:45 +01:00
if (!(*arg)->fixed &&
(*arg)->fix_fields(thd, arg))
2004-03-19 11:35:00 +01:00
DBUG_RETURN(1);
// we can't assign 'item' before, because fix_fields() can change arg
2003-05-19 15:35:49 +02:00
Item *item= *arg;
if (item->check_cols(1))
DBUG_RETURN(TRUE);
/*
TODO: We should think about this. It is not always
right way just to set an UDF result to return my_charset_bin
if one argument has binary sorting order.
The result collation should be calculated according to arguments
derivations in some cases and should not in other cases.
Moreover, some arguments can represent a numeric input
which doesn't effect the result character set and collation.
There is no a general rule for UDF. Everything depends on
2005-02-08 23:50:45 +01:00
the particular user defined function.
*/
if (item->collation.collation->state & MY_CS_BINSORT)
func->collation.set(&my_charset_bin);
2003-03-24 21:52:46 +01:00
if (item->maybe_null)
2000-07-31 21:29:14 +02:00
func->maybe_null=1;
2003-03-24 21:52:46 +01:00
func->with_sum_func= func->with_sum_func || item->with_sum_func;
used_tables_cache|=item->used_tables();
const_item_cache&=item->const_item();
f_args.arg_type[i]=item->result_type();
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
//TODO: why all following memory is not allocated with 1 call of sql_alloc?
2000-07-31 21:29:14 +02:00
if (!(buffers=new String[arg_count]) ||
!(f_args.args= (char**) sql_alloc(arg_count * sizeof(char *))) ||
!(f_args.lengths= (ulong*) sql_alloc(arg_count * sizeof(long))) ||
!(f_args.maybe_null= (char*) sql_alloc(arg_count * sizeof(char))) ||
!(num_buffer= (char*) sql_alloc(arg_count *
ALIGN_SIZE(sizeof(double)))) ||
!(f_args.attributes= (char**) sql_alloc(arg_count * sizeof(char *))) ||
!(f_args.attribute_lengths= (ulong*) sql_alloc(arg_count *
sizeof(long))))
2000-07-31 21:29:14 +02:00
{
free_udf(u_d);
DBUG_RETURN(TRUE);
2000-07-31 21:29:14 +02:00
}
}
func->fix_length_and_dec();
initid.max_length=func->max_length;
initid.maybe_null=func->maybe_null;
initid.const_item=const_item_cache;
initid.decimals=func->decimals;
initid.ptr=0;
if (u_d->func_init)
{
char init_msg_buff[MYSQL_ERRMSG_SIZE];
2000-07-31 21:29:14 +02:00
char *to=num_buffer;
for (uint i=0; i < arg_count; i++)
{
/*
For a constant argument i, args->args[i] points to the argument value.
For non-constant, args->args[i] is NULL.
*/
f_args.args[i]= NULL; /* Non-const unless updated below. */
f_args.lengths[i]= arguments[i]->max_length;
f_args.maybe_null[i]= (char) arguments[i]->maybe_null;
f_args.attributes[i]= arguments[i]->name;
f_args.attribute_lengths[i]= arguments[i]->name_length;
2000-07-31 21:29:14 +02:00
if (arguments[i]->const_item())
2000-07-31 21:29:14 +02:00
{
switch (arguments[i]->result_type())
{
case STRING_RESULT:
case DECIMAL_RESULT:
{
String *res= arguments[i]->val_str(&buffers[i]);
if (arguments[i]->null_value)
continue;
f_args.args[i]= (char*) res->c_ptr_safe();
f_args.lengths[i]= res->length();
break;
}
case INT_RESULT:
*((longlong*) to)= arguments[i]->val_int();
if (arguments[i]->null_value)
continue;
f_args.args[i]= to;
to+= ALIGN_SIZE(sizeof(longlong));
break;
case REAL_RESULT:
*((double*) to)= arguments[i]->val_real();
if (arguments[i]->null_value)
continue;
f_args.args[i]= to;
to+= ALIGN_SIZE(sizeof(double));
break;
case ROW_RESULT:
default:
// This case should never be chosen
DBUG_ASSERT(0);
break;
}
2000-07-31 21:29:14 +02:00
}
}
Udf_func_init init= u_d->func_init;
if ((error=(uchar) init(&initid, &f_args, init_msg_buff)))
2000-07-31 21:29:14 +02:00
{
my_error(ER_CANT_INITIALIZE_UDF, MYF(0),
u_d->name.str, init_msg_buff);
2000-07-31 21:29:14 +02:00
free_udf(u_d);
DBUG_RETURN(TRUE);
2000-07-31 21:29:14 +02:00
}
func->max_length=min(initid.max_length,MAX_BLOB_WIDTH);
func->maybe_null=initid.maybe_null;
const_item_cache=initid.const_item;
/*
Keep used_tables_cache in sync with const_item_cache.
See the comment in Item_udf_func::update_used tables.
*/
if (!const_item_cache && !used_tables_cache)
used_tables_cache= RAND_TABLE_BIT;
func->decimals=min(initid.decimals,NOT_FIXED_DEC);
2000-07-31 21:29:14 +02:00
}
initialized=1;
if (error)
{
my_error(ER_CANT_INITIALIZE_UDF, MYF(0),
u_d->name.str, ER(ER_UNKNOWN_ERROR));
DBUG_RETURN(TRUE);
2000-07-31 21:29:14 +02:00
}
DBUG_RETURN(FALSE);
2000-07-31 21:29:14 +02:00
}
bool udf_handler::get_arguments()
{
if (error)
return 1; // Got an error earlier
char *to= num_buffer;
uint str_count=0;
for (uint i=0; i < f_args.arg_count; i++)
{
f_args.args[i]=0;
switch (f_args.arg_type[i]) {
case STRING_RESULT:
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
2000-07-31 21:29:14 +02:00
{
String *res=args[i]->val_str(&buffers[str_count++]);
if (!(args[i]->null_value))
{
f_args.args[i]= (char*) res->ptr();
f_args.lengths[i]= res->length();
break;
}
}
case INT_RESULT:
*((longlong*) to) = args[i]->val_int();
if (!args[i]->null_value)
{
f_args.args[i]=to;
to+= ALIGN_SIZE(sizeof(longlong));
}
break;
case REAL_RESULT:
2004-11-11 19:39:35 +01:00
*((double*) to)= args[i]->val_real();
2000-07-31 21:29:14 +02:00
if (!args[i]->null_value)
{
f_args.args[i]=to;
to+= ALIGN_SIZE(sizeof(double));
}
break;
case ROW_RESULT:
2003-01-31 11:07:07 +01:00
default:
2005-02-08 23:50:45 +01:00
// This case should never be chosen
2002-11-15 19:32:09 +01:00
DBUG_ASSERT(0);
break;
2000-07-31 21:29:14 +02:00
}
}
return 0;
}
2007-10-11 19:29:09 +02:00
/**
@return
(String*)NULL in case of NULL values
*/
2000-07-31 21:29:14 +02:00
String *udf_handler::val_str(String *str,String *save_str)
{
uchar is_null_tmp=0;
2000-07-31 21:29:14 +02:00
ulong res_length;
DBUG_ENTER("udf_handler::val_str");
2000-07-31 21:29:14 +02:00
if (get_arguments())
DBUG_RETURN(0);
2000-07-31 21:29:14 +02:00
char * (*func)(UDF_INIT *, UDF_ARGS *, char *, ulong *, uchar *, uchar *)=
(char* (*)(UDF_INIT *, UDF_ARGS *, char *, ulong *, uchar *, uchar *))
u_d->func;
if ((res_length=str->alloced_length()) < MAX_FIELD_WIDTH)
{ // This happens VERY seldom
if (str->alloc(MAX_FIELD_WIDTH))
{
error=1;
DBUG_RETURN(0);
2000-07-31 21:29:14 +02:00
}
}
char *res=func(&initid, &f_args, (char*) str->ptr(), &res_length,
&is_null_tmp, &error);
DBUG_PRINT("info", ("udf func returned, res_length: %lu", res_length));
if (is_null_tmp || !res || error) // The !res is for safety
2000-07-31 21:29:14 +02:00
{
DBUG_PRINT("info", ("Null or error"));
DBUG_RETURN(0);
2000-07-31 21:29:14 +02:00
}
if (res == str->ptr())
{
str->length(res_length);
DBUG_PRINT("exit", ("str: %*.s", (int) str->length(), str->ptr()));
DBUG_RETURN(str);
2000-07-31 21:29:14 +02:00
}
save_str->set(res, res_length, str->charset());
DBUG_PRINT("exit", ("save_str: %s", save_str->ptr()));
DBUG_RETURN(save_str);
2000-07-31 21:29:14 +02:00
}
/*
For the moment, UDF functions are returning DECIMAL values as strings
*/
2005-02-08 23:50:45 +01:00
my_decimal *udf_handler::val_decimal(my_bool *null_value, my_decimal *dec_buf)
{
char buf[DECIMAL_MAX_STR_LENGTH+1], *end;
ulong res_length= DECIMAL_MAX_STR_LENGTH;
if (get_arguments())
{
*null_value=1;
return 0;
}
char *(*func)(UDF_INIT *, UDF_ARGS *, char *, ulong *, uchar *, uchar *)=
(char* (*)(UDF_INIT *, UDF_ARGS *, char *, ulong *, uchar *, uchar *))
u_d->func;
char *res= func(&initid, &f_args, buf, &res_length, &is_null, &error);
if (is_null || error)
{
*null_value= 1;
return 0;
}
end= res+ res_length;
str2my_decimal(E_DEC_FATAL_ERROR, res, dec_buf, &end);
2005-02-08 23:50:45 +01:00
return dec_buf;
}
2000-07-31 21:29:14 +02:00
void Item_udf_func::cleanup()
{
udf.cleanup();
Item_func::cleanup();
}
2000-07-31 21:29:14 +02:00
void Item_udf_func::print(String *str, enum_query_type query_type)
{
str->append(func_name());
str->append('(');
for (uint i=0 ; i < arg_count ; i++)
{
if (i != 0)
str->append(',');
args[i]->print_item_w_name(str, query_type);
}
str->append(')');
}
2004-11-11 19:39:35 +01:00
double Item_func_udf_float::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
DBUG_ENTER("Item_func_udf_float::val");
DBUG_PRINT("info",("result_type: %d arg_count: %d",
args[0]->result_type(), arg_count));
DBUG_RETURN(udf.val(&null_value));
}
String *Item_func_udf_float::val_str(String *str)
{
DBUG_ASSERT(fixed == 1);
2004-11-11 19:39:35 +01:00
double nr= val_real();
2000-07-31 21:29:14 +02:00
if (null_value)
return 0; /* purecov: inspected */
str->set_real(nr,decimals,&my_charset_bin);
2000-07-31 21:29:14 +02:00
return str;
}
longlong Item_func_udf_int::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
DBUG_ENTER("Item_func_udf_int::val_int");
DBUG_RETURN(udf.val_int(&null_value));
}
String *Item_func_udf_int::val_str(String *str)
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
longlong nr=val_int();
if (null_value)
return 0;
str->set_int(nr, unsigned_flag, &my_charset_bin);
2000-07-31 21:29:14 +02:00
return str;
}
2005-02-08 23:50:45 +01:00
longlong Item_func_udf_decimal::val_int()
{
my_decimal dec_buf, *dec= udf.val_decimal(&null_value, &dec_buf);
longlong result;
2005-02-08 23:50:45 +01:00
if (null_value)
return 0;
my_decimal2int(E_DEC_FATAL_ERROR, dec, unsigned_flag, &result);
return result;
}
double Item_func_udf_decimal::val_real()
{
my_decimal dec_buf, *dec= udf.val_decimal(&null_value, &dec_buf);
double result;
2005-02-08 23:50:45 +01:00
if (null_value)
return 0.0;
my_decimal2double(E_DEC_FATAL_ERROR, dec, &result);
return result;
}
my_decimal *Item_func_udf_decimal::val_decimal(my_decimal *dec_buf)
{
DBUG_ASSERT(fixed == 1);
DBUG_ENTER("Item_func_udf_decimal::val_decimal");
DBUG_PRINT("info",("result_type: %d arg_count: %d",
args[0]->result_type(), arg_count));
DBUG_RETURN(udf.val_decimal(&null_value, dec_buf));
}
String *Item_func_udf_decimal::val_str(String *str)
{
my_decimal dec_buf, *dec= udf.val_decimal(&null_value, &dec_buf);
if (null_value)
return 0;
if (str->length() < DECIMAL_MAX_STR_LENGTH)
str->length(DECIMAL_MAX_STR_LENGTH);
my_decimal_round(E_DEC_FATAL_ERROR, dec, decimals, FALSE, &dec_buf);
my_decimal2string(E_DEC_FATAL_ERROR, &dec_buf, 0, 0, '0', str);
return str;
}
void Item_func_udf_decimal::fix_length_and_dec()
{
fix_num_length_and_dec();
}
2000-07-31 21:29:14 +02:00
/* Default max_length is max argument length */
void Item_func_udf_str::fix_length_and_dec()
{
DBUG_ENTER("Item_func_udf_str::fix_length_and_dec");
max_length=0;
for (uint i = 0; i < arg_count; i++)
set_if_bigger(max_length,args[i]->max_length);
DBUG_VOID_RETURN;
}
String *Item_func_udf_str::val_str(String *str)
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *res=udf.val_str(str,&str_value);
null_value = !res;
return res;
}
2007-10-11 19:29:09 +02:00
/**
@note
This has to come last in the udf_handler methods, or C for AIX
version 6.0.0.0 fails to compile with debugging enabled. (Yes, really.)
*/
udf_handler::~udf_handler()
{
/* Everything should be properly cleaned up by this moment. */
DBUG_ASSERT(not_original || !(initialized || buffers));
}
2000-07-31 21:29:14 +02:00
#else
bool udf_handler::get_arguments() { return 0; }
#endif /* HAVE_DLOPEN */
/*
** User level locks
*/
mysql_mutex_t LOCK_user_locks;
2000-07-31 21:29:14 +02:00
static HASH hash_user_locks;
class User_level_lock
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *key;
size_t key_length;
2000-07-31 21:29:14 +02:00
public:
int count;
bool locked;
mysql_cond_t cond;
Fixed compiler warnings Fixed compile-pentium64 scripts Fixed wrong estimate of update_with_key_prefix in sql-bench Merge bk-internal.mysql.com:/home/bk/mysql-5.1 into mysql.com:/home/my/mysql-5.1 Fixed unsafe define of uint4korr() Fixed that --extern works with mysql-test-run.pl Small trivial cleanups This also fixes a bug in counting number of rows that are updated when we have many simultanous queries Move all connection handling and command exectuion main loop from sql_parse.cc to sql_connection.cc Split handle_one_connection() into reusable sub functions. Split create_new_thread() into reusable sub functions. Added thread_scheduler; Preliminary interface code for future thread_handling code. Use 'my_thread_id' for internal thread id's Make thr_alarm_kill() to depend on thread_id instead of thread Make thr_abort_locks_for_thread() depend on thread_id instead of thread In store_globals(), set my_thread_var->id to be thd->thread_id. Use my_thread_var->id as basis for my_thread_name() The above changes makes the connection we have between THD and threads more soft. Added a lot of DBUG_PRINT() and DBUG_ASSERT() functions Fixed compiler warnings Fixed core dumps when running with --debug Removed setting of signal masks (was never used) Made event code call pthread_exit() (portability fix) Fixed that event code doesn't call DBUG_xxx functions before my_thread_init() is called. Made handling of thread_id and thd->variables.pseudo_thread_id uniform. Removed one common 'not freed memory' warning from mysqltest Fixed a couple of usage of not initialized warnings (unlikely cases) Suppress compiler warnings from bdb and (for the moment) warnings from ndb
2007-02-23 12:13:55 +01:00
my_thread_id thread_id;
void set_thread(THD *thd) { thread_id= thd->thread_id; }
2000-07-31 21:29:14 +02:00
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
User_level_lock(const uchar *key_arg,uint length, ulong id)
:key_length(length),count(1),locked(1), thread_id(id)
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
key= (uchar*) my_memdup(key_arg,length,MYF(0));
mysql_cond_init(key_user_level_lock_cond, &cond, NULL);
2000-07-31 21:29:14 +02:00
if (key)
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_hash_insert(&hash_user_locks,(uchar*) this))
2000-07-31 21:29:14 +02:00
{
Bug#34043: Server loops excessively in _checkchunk() when safemalloc is enabled Essentially, the problem is that safemalloc is excruciatingly slow as it checks all allocated blocks for overrun at each memory management primitive, yielding a almost exponential slowdown for the memory management functions (malloc, realloc, free). The overrun check basically consists of verifying some bytes of a block for certain magic keys, which catches some simple forms of overrun. Another minor problem is violation of aliasing rules and that its own internal list of blocks is prone to corruption. Another issue with safemalloc is rather the maintenance cost as the tool has a significant impact on the server code. Given the magnitude of memory debuggers available nowadays, especially those that are provided with the platform malloc implementation, maintenance of a in-house and largely obsolete memory debugger becomes a burden that is not worth the effort due to its slowness and lack of support for detecting more common forms of heap corruption. Since there are third-party tools that can provide the same functionality at a lower or comparable performance cost, the solution is to simply remove safemalloc. Third-party tools can provide the same functionality at a lower or comparable performance cost. The removal of safemalloc also allows a simplification of the malloc wrappers, removing quite a bit of kludge: redefinition of my_malloc, my_free and the removal of the unused second argument of my_free. Since free() always check whether the supplied pointer is null, redudant checks are also removed. Also, this patch adds unit testing for my_malloc and moves my_realloc implementation into the same file as the other memory allocation primitives.
2010-07-08 23:20:08 +02:00
my_free(key);
2000-07-31 21:29:14 +02:00
key=0;
}
}
}
~User_level_lock()
2000-07-31 21:29:14 +02:00
{
if (key)
{
my_hash_delete(&hash_user_locks,(uchar*) this);
Bug#34043: Server loops excessively in _checkchunk() when safemalloc is enabled Essentially, the problem is that safemalloc is excruciatingly slow as it checks all allocated blocks for overrun at each memory management primitive, yielding a almost exponential slowdown for the memory management functions (malloc, realloc, free). The overrun check basically consists of verifying some bytes of a block for certain magic keys, which catches some simple forms of overrun. Another minor problem is violation of aliasing rules and that its own internal list of blocks is prone to corruption. Another issue with safemalloc is rather the maintenance cost as the tool has a significant impact on the server code. Given the magnitude of memory debuggers available nowadays, especially those that are provided with the platform malloc implementation, maintenance of a in-house and largely obsolete memory debugger becomes a burden that is not worth the effort due to its slowness and lack of support for detecting more common forms of heap corruption. Since there are third-party tools that can provide the same functionality at a lower or comparable performance cost, the solution is to simply remove safemalloc. Third-party tools can provide the same functionality at a lower or comparable performance cost. The removal of safemalloc also allows a simplification of the malloc wrappers, removing quite a bit of kludge: redefinition of my_malloc, my_free and the removal of the unused second argument of my_free. Since free() always check whether the supplied pointer is null, redudant checks are also removed. Also, this patch adds unit testing for my_malloc and moves my_realloc implementation into the same file as the other memory allocation primitives.
2010-07-08 23:20:08 +02:00
my_free(key);
2000-07-31 21:29:14 +02:00
}
mysql_cond_destroy(&cond);
2000-07-31 21:29:14 +02:00
}
inline bool initialized() { return key != 0; }
friend void item_user_lock_release(User_level_lock *ull);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
friend uchar *ull_get_key(const User_level_lock *ull, size_t *length,
my_bool not_used);
2000-07-31 21:29:14 +02:00
};
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uchar *ull_get_key(const User_level_lock *ull, size_t *length,
my_bool not_used __attribute__((unused)))
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
*length= ull->key_length;
return ull->key;
2000-07-31 21:29:14 +02:00
}
#ifdef HAVE_PSI_INTERFACE
static PSI_mutex_key key_LOCK_user_locks;
static PSI_mutex_info all_user_mutexes[]=
{
{ &key_LOCK_user_locks, "LOCK_user_locks", PSI_FLAG_GLOBAL}
};
static void init_user_lock_psi_keys(void)
{
const char* category= "sql";
int count;
if (PSI_server == NULL)
return;
count= array_elements(all_user_mutexes);
PSI_server->register_mutex(category, all_user_mutexes, count);
}
#endif
static bool item_user_lock_inited= 0;
2000-07-31 21:29:14 +02:00
void item_user_lock_init(void)
{
#ifdef HAVE_PSI_INTERFACE
init_user_lock_psi_keys();
#endif
mysql_mutex_init(key_LOCK_user_locks, &LOCK_user_locks, MY_MUTEX_INIT_SLOW);
my_hash_init(&hash_user_locks,system_charset_info,
16,0,0,(my_hash_get_key) ull_get_key,NULL,0);
item_user_lock_inited= 1;
2000-07-31 21:29:14 +02:00
}
void item_user_lock_free(void)
{
if (item_user_lock_inited)
{
item_user_lock_inited= 0;
my_hash_free(&hash_user_locks);
mysql_mutex_destroy(&LOCK_user_locks);
}
2000-07-31 21:29:14 +02:00
}
void item_user_lock_release(User_level_lock *ull)
2000-07-31 21:29:14 +02:00
{
ull->locked=0;
ull->thread_id= 0;
2000-07-31 21:29:14 +02:00
if (--ull->count)
mysql_cond_signal(&ull->cond);
2000-07-31 21:29:14 +02:00
else
delete ull;
}
2007-10-11 19:29:09 +02:00
/**
Wait until we are at or past the given position in the master binlog
on the slave.
*/
longlong Item_master_pos_wait::val_int()
{
DBUG_ASSERT(fixed == 1);
THD* thd = current_thd;
String *log_name = args[0]->val_str(&value);
int event_count= 0;
2001-01-22 03:46:32 +01:00
null_value=0;
if (thd->slave_thread || !log_name || !log_name->length())
{
null_value = 1;
return 0;
}
#ifdef HAVE_REPLICATION
2003-02-04 20:52:14 +01:00
longlong pos = (ulong)args[1]->val_int();
longlong timeout = (arg_count==3) ? args[2]->val_int() : 0 ;
if ((event_count = active_mi->rli.wait_for_pos(thd, log_name, pos, timeout)) == -2)
2001-01-22 03:46:32 +01:00
{
null_value = 1;
event_count=0;
}
#endif
return event_count;
}
/**
Enables a session to wait on a condition until a timeout or a network
disconnect occurs.
@remark The connection is polled every m_interrupt_interval nanoseconds.
*/
class Interruptible_wait
{
THD *m_thd;
struct timespec m_abs_timeout;
static const ulonglong m_interrupt_interval;
public:
Interruptible_wait(THD *thd)
: m_thd(thd) {}
~Interruptible_wait() {}
public:
/**
Set the absolute timeout.
@param timeout The amount of time in nanoseconds to wait
*/
void set_timeout(ulonglong timeout)
{
/*
Calculate the absolute system time at the start so it can
be controlled in slices. It relies on the fact that once
the absolute time passes, the timed wait call will fail
automatically with a timeout error.
*/
set_timespec_nsec(m_abs_timeout, timeout);
}
/** The timed wait. */
int wait(mysql_cond_t *, mysql_mutex_t *);
};
/** Time to wait before polling the connection status. */
const ulonglong Interruptible_wait::m_interrupt_interval= 5 * ULL(1000000000);
/**
Wait for a given condition to be signaled.
@param cond The condition variable to wait on.
@param mutex The associated mutex.
@remark The absolute timeout is preserved across calls.
@retval return value from mysql_cond_timedwait
*/
int Interruptible_wait::wait(mysql_cond_t *cond, mysql_mutex_t *mutex)
{
int error;
struct timespec timeout;
while (1)
{
/* Wait for a fixed interval. */
set_timespec_nsec(timeout, m_interrupt_interval);
/* But only if not past the absolute timeout. */
if (cmp_timespec(timeout, m_abs_timeout) > 0)
timeout= m_abs_timeout;
error= mysql_cond_timedwait(cond, mutex, &timeout);
if (error == ETIMEDOUT || error == ETIME)
{
/* Return error if timed out or connection is broken. */
if (!cmp_timespec(timeout, m_abs_timeout) || !m_thd->is_connected())
break;
}
/* Otherwise, propagate status to the caller. */
else
break;
}
return error;
}
2007-10-11 19:29:09 +02:00
/**
Get a user level lock. If the thread has an old lock this is first released.
@retval
1 : Got lock
@retval
0 : Timeout
@retval
NULL : Error
2000-07-31 21:29:14 +02:00
*/
longlong Item_func_get_lock::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *res=args[0]->val_str(&value);
ulonglong timeout= args[1]->val_int();
2000-07-31 21:29:14 +02:00
THD *thd=current_thd;
User_level_lock *ull;
int error;
Interruptible_wait timed_cond(thd);
DBUG_ENTER("Item_func_get_lock::val_int");
2000-07-31 21:29:14 +02:00
/*
In slave thread no need to get locks, everything is serialized. Anyway
there is no way to make GET_LOCK() work on slave like it did on master
(i.e. make it return exactly the same value) because we don't have the
same other concurrent threads environment. No matter what we return here,
it's not guaranteed to be same as on master.
*/
if (thd->slave_thread)
DBUG_RETURN(1);
mysql_mutex_lock(&LOCK_user_locks);
2000-07-31 21:29:14 +02:00
if (!res || !res->length())
{
mysql_mutex_unlock(&LOCK_user_locks);
2000-07-31 21:29:14 +02:00
null_value=1;
DBUG_RETURN(0);
2000-07-31 21:29:14 +02:00
}
DBUG_PRINT("info", ("lock %.*s, thd=%ld", res->length(), res->ptr(),
(long) thd->real_id));
2000-07-31 21:29:14 +02:00
null_value=0;
if (thd->ull)
{
item_user_lock_release(thd->ull);
thd->ull=0;
}
if (!(ull= ((User_level_lock *) my_hash_search(&hash_user_locks,
(uchar*) res->ptr(),
(size_t) res->length()))))
2000-07-31 21:29:14 +02:00
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
ull= new User_level_lock((uchar*) res->ptr(), (size_t) res->length(),
thd->thread_id);
2000-07-31 21:29:14 +02:00
if (!ull || !ull->initialized())
{
delete ull;
mysql_mutex_unlock(&LOCK_user_locks);
2000-07-31 21:29:14 +02:00
null_value=1; // Probably out of memory
DBUG_RETURN(0);
2000-07-31 21:29:14 +02:00
}
ull->set_thread(thd);
2000-07-31 21:29:14 +02:00
thd->ull=ull;
mysql_mutex_unlock(&LOCK_user_locks);
DBUG_PRINT("info", ("made new lock"));
DBUG_RETURN(1); // Got new lock
2000-07-31 21:29:14 +02:00
}
ull->count++;
DBUG_PRINT("info", ("ull->count=%d", ull->count));
2000-07-31 21:29:14 +02:00
/*
Structure is now initialized. Try to get the lock.
Set up control struct to allow others to abort locks.
*/
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length.
2007-02-22 16:03:08 +01:00
thd_proc_info(thd, "User lock");
2000-07-31 21:29:14 +02:00
thd->mysys_var->current_mutex= &LOCK_user_locks;
thd->mysys_var->current_cond= &ull->cond;
timed_cond.set_timeout(timeout * ULL(1000000000));
error= 0;
2011-03-04 13:12:31 +01:00
thd_wait_begin(thd, THD_WAIT_USER_LOCK);
while (ull->locked && !thd->killed)
{
DBUG_PRINT("info", ("waiting on lock"));
error= timed_cond.wait(&ull->cond, &LOCK_user_locks);
if (error == ETIMEDOUT || error == ETIME)
{
DBUG_PRINT("info", ("lock wait timeout"));
break;
}
error= 0;
}
2011-03-04 13:12:31 +01:00
thd_wait_end(thd);
2000-07-31 21:29:14 +02:00
if (ull->locked)
{
if (!--ull->count)
{
DBUG_ASSERT(0);
2000-07-31 21:29:14 +02:00
delete ull; // Should never happen
}
if (!error) // Killed (thd->killed != 0)
2000-07-31 21:29:14 +02:00
{
error=1;
null_value=1; // Return NULL
}
}
else // We got the lock
2000-07-31 21:29:14 +02:00
{
ull->locked=1;
ull->set_thread(thd);
ull->thread_id= thd->thread_id;
2000-07-31 21:29:14 +02:00
thd->ull=ull;
error=0;
DBUG_PRINT("info", ("got the lock"));
2000-07-31 21:29:14 +02:00
}
mysql_mutex_unlock(&LOCK_user_locks);
2000-07-31 21:29:14 +02:00
mysql_mutex_lock(&thd->mysys_var->mutex);
Prevent bugs by making DBUG_* expressions syntactically equivalent to a single statement. --- Bug#24795: SHOW PROFILE Profiling is only partially functional on some architectures. Where there is no getrusage() system call, presently Null values are returned where it would be required. Notably, Windows needs some love applied to make it as useful. Syntax this adds: SHOW PROFILES SHOW PROFILE [types] [FOR QUERY n] [OFFSET n] [LIMIT n] where "n" is an integer and "types" is zero or many (comma-separated) of "CPU" "MEMORY" (not presently supported) "BLOCK IO" "CONTEXT SWITCHES" "PAGE FAULTS" "IPC" "SWAPS" "SOURCE" "ALL" It also adds a session variable (boolean) "profiling", set to "no" by default, and (integer) profiling_history_size, set to 15 by default. This patch abstracts setting THDs' "proc_info" behind a macro that can be used as a hook into the profiling code when profiling support is compiled in. All future code in this line should use that mechanism for setting thd->proc_info. --- Tests are now set to omit the statistics. --- Adds an Information_schema table, "profiling" for access to "show profile" data. --- Merge zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community-3--bug24795 into zippy.cornsilk.net:/home/cmiller/work/mysql/mysql-5.0-community --- Fix merge problems. --- Fixed one bug in the query_source being NULL. Updated test results. --- Include more thorough profiling tests. Improve support for prepared statements. Use session-specific query IDs, starting at zero. --- Selecting from I_S.profiling is no longer quashed in profiling, as requested by Giuseppe. Limit the size of captured query text. No longer log queries that are zero length.
2007-02-22 16:03:08 +01:00
thd_proc_info(thd, 0);
2000-07-31 21:29:14 +02:00
thd->mysys_var->current_mutex= 0;
thd->mysys_var->current_cond= 0;
mysql_mutex_unlock(&thd->mysys_var->mutex);
2000-07-31 21:29:14 +02:00
DBUG_RETURN(!error ? 1 : 0);
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/**
Release a user level lock.
2007-10-11 19:29:09 +02:00
@return
- 1 if lock released
- 0 if lock wasn't held
- (SQL) NULL if no such lock
2000-07-31 21:29:14 +02:00
*/
longlong Item_func_release_lock::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
String *res=args[0]->val_str(&value);
User_level_lock *ull;
2000-07-31 21:29:14 +02:00
longlong result;
THD *thd=current_thd;
DBUG_ENTER("Item_func_release_lock::val_int");
2000-07-31 21:29:14 +02:00
if (!res || !res->length())
{
null_value=1;
DBUG_RETURN(0);
2000-07-31 21:29:14 +02:00
}
DBUG_PRINT("info", ("lock %.*s", res->length(), res->ptr()));
2000-07-31 21:29:14 +02:00
null_value=0;
result=0;
mysql_mutex_lock(&LOCK_user_locks);
if (!(ull= ((User_level_lock*) my_hash_search(&hash_user_locks,
(const uchar*) res->ptr(),
(size_t) res->length()))))
2000-07-31 21:29:14 +02:00
{
null_value=1;
}
else
{
2007-06-12 17:41:56 +02:00
DBUG_PRINT("info", ("ull->locked=%d ull->thread=%lu thd=%lu",
(int) ull->locked,
2007-06-12 17:41:56 +02:00
(long)ull->thread_id,
(long)thd->thread_id));
Fixed compiler warnings Fixed compile-pentium64 scripts Fixed wrong estimate of update_with_key_prefix in sql-bench Merge bk-internal.mysql.com:/home/bk/mysql-5.1 into mysql.com:/home/my/mysql-5.1 Fixed unsafe define of uint4korr() Fixed that --extern works with mysql-test-run.pl Small trivial cleanups This also fixes a bug in counting number of rows that are updated when we have many simultanous queries Move all connection handling and command exectuion main loop from sql_parse.cc to sql_connection.cc Split handle_one_connection() into reusable sub functions. Split create_new_thread() into reusable sub functions. Added thread_scheduler; Preliminary interface code for future thread_handling code. Use 'my_thread_id' for internal thread id's Make thr_alarm_kill() to depend on thread_id instead of thread Make thr_abort_locks_for_thread() depend on thread_id instead of thread In store_globals(), set my_thread_var->id to be thd->thread_id. Use my_thread_var->id as basis for my_thread_name() The above changes makes the connection we have between THD and threads more soft. Added a lot of DBUG_PRINT() and DBUG_ASSERT() functions Fixed compiler warnings Fixed core dumps when running with --debug Removed setting of signal masks (was never used) Made event code call pthread_exit() (portability fix) Fixed that event code doesn't call DBUG_xxx functions before my_thread_init() is called. Made handling of thread_id and thd->variables.pseudo_thread_id uniform. Removed one common 'not freed memory' warning from mysqltest Fixed a couple of usage of not initialized warnings (unlikely cases) Suppress compiler warnings from bdb and (for the moment) warnings from ndb
2007-02-23 12:13:55 +01:00
if (ull->locked && current_thd->thread_id == ull->thread_id)
2000-07-31 21:29:14 +02:00
{
DBUG_PRINT("info", ("release lock"));
2000-07-31 21:29:14 +02:00
result=1; // Release is ok
item_user_lock_release(ull);
thd->ull=0;
2000-07-31 21:29:14 +02:00
}
}
mysql_mutex_unlock(&LOCK_user_locks);
DBUG_RETURN(result);
2000-07-31 21:29:14 +02:00
}
longlong Item_func_last_insert_id::val_int()
2000-07-31 21:29:14 +02:00
{
THD *thd= current_thd;
DBUG_ASSERT(fixed == 1);
if (arg_count)
{
longlong value= args[0]->val_int();
null_value= args[0]->null_value;
WL#3146 "less locking in auto_increment": this is a cleanup patch for our current auto_increment handling: new names for auto_increment variables in THD, new methods to manipulate them (see sql_class.h), some move into handler::, causing less backup/restore work when executing substatements. This makes the logic hopefully clearer, less work is is needed in mysql_insert(). By cleaning up, using different variables for different purposes (instead of one for 3 things...), we fix those bugs, which someone may want to fix in 5.0 too: BUG#20339 "stored procedure using LAST_INSERT_ID() does not replicate statement-based" BUG#20341 "stored function inserting into one auto_increment puts bad data in slave" BUG#19243 "wrong LAST_INSERT_ID() after ON DUPLICATE KEY UPDATE" (now if a row is updated, LAST_INSERT_ID() will return its id) and re-fixes: BUG#6880 "LAST_INSERT_ID() value changes during multi-row INSERT" (already fixed differently by Ramil in 4.1) Test of documented behaviour of mysql_insert_id() (there was no test). The behaviour changes introduced are: - LAST_INSERT_ID() now returns "the first autogenerated auto_increment value successfully inserted", instead of "the first autogenerated auto_increment value if any row was successfully inserted", see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() returns the id of the updated row if ON DUPLICATE KEY UPDATE, see auto_increment.test. Same for mysql_insert_id(), see mysql_client_test.c. - LAST_INSERT_ID() does not change if no autogenerated value was successfully inserted (it used to then be 0), see auto_increment.test. - if in INSERT SELECT no autogenerated value was successfully inserted, mysql_insert_id() now returns the id of the last inserted row (it already did this for INSERT VALUES), see mysql_client_test.c. - if INSERT SELECT uses LAST_INSERT_ID(X), mysql_insert_id() now returns X (it already did this for INSERT VALUES), see mysql_client_test.c. - NDB now behaves like other engines wrt SET INSERT_ID: with INSERT IGNORE, the id passed in SET INSERT_ID is re-used until a row succeeds; SET INSERT_ID influences not only the first row now. Additionally, when unlocking a table we check that the thread is not keeping a next_insert_id (as the table is unlocked that id is potentially out-of-date); forgetting about this next_insert_id is done in a new handler::ha_release_auto_increment(). Finally we prepare for engines capable of reserving finite-length intervals of auto_increment values: we store such intervals in THD. The next step (to be done by the replication team in 5.1) is to read those intervals from THD and actually store them in the statement-based binary log. NDB will be a good engine to test that.
2006-07-09 17:52:19 +02:00
/*
LAST_INSERT_ID(X) must affect the client's mysql_insert_id() as
documented in the manual. We don't want to touch
first_successful_insert_id_in_cur_stmt because it would make
LAST_INSERT_ID(X) take precedence over an generated auto_increment
value for this row.
*/
thd->arg_of_last_insert_id_function= TRUE;
thd->first_successful_insert_id_in_prev_stmt= value;
return value;
}
return
static_cast<longlong>(thd->read_first_successful_insert_id_in_prev_stmt());
2000-07-31 21:29:14 +02:00
}
bool Item_func_last_insert_id::fix_fields(THD *thd, Item **ref)
{
thd->lex->uncacheable(UNCACHEABLE_SIDEEFFECT);
return Item_int_func::fix_fields(thd, ref);
}
2000-07-31 21:29:14 +02:00
/* This function is just used to test speed of different functions */
longlong Item_func_benchmark::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
char buff[MAX_FIELD_WIDTH];
2003-01-29 14:31:20 +01:00
String tmp(buff,sizeof(buff), &my_charset_bin);
2007-03-02 04:04:49 +01:00
my_decimal tmp_decimal;
2000-07-31 21:29:14 +02:00
THD *thd=current_thd;
ulonglong loop_count;
2000-07-31 21:29:14 +02:00
loop_count= (ulonglong) args[0]->val_int();
if (args[0]->null_value ||
(!args[0]->unsigned_flag && (((longlong) loop_count) < 0)))
{
if (!args[0]->null_value)
{
char buff[22];
llstr(((longlong) loop_count), buff);
push_warning_printf(current_thd, MYSQL_ERROR::WARN_LEVEL_WARN,
ER_WRONG_VALUE_FOR_TYPE, ER(ER_WRONG_VALUE_FOR_TYPE),
"count", buff, "benchmark");
}
null_value= 1;
return 0;
}
null_value=0;
for (ulonglong loop=0 ; loop < loop_count && !thd->killed; loop++)
2000-07-31 21:29:14 +02:00
{
switch (args[1]->result_type()) {
2000-07-31 21:29:14 +02:00
case REAL_RESULT:
(void) args[1]->val_real();
2000-07-31 21:29:14 +02:00
break;
case INT_RESULT:
(void) args[1]->val_int();
2000-07-31 21:29:14 +02:00
break;
case STRING_RESULT:
(void) args[1]->val_str(&tmp);
2000-07-31 21:29:14 +02:00
break;
2007-03-02 04:04:49 +01:00
case DECIMAL_RESULT:
(void) args[1]->val_decimal(&tmp_decimal);
break;
case ROW_RESULT:
2003-01-31 11:07:07 +01:00
default:
2005-02-08 23:50:45 +01:00
// This case should never be chosen
2002-11-15 19:32:09 +01:00
DBUG_ASSERT(0);
return 0;
2000-07-31 21:29:14 +02:00
}
}
return 0;
}
void Item_func_benchmark::print(String *str, enum_query_type query_type)
{
str->append(STRING_WITH_LEN("benchmark("));
args[0]->print(str, query_type);
str->append(',');
args[1]->print(str, query_type);
str->append(')');
}
2007-10-11 19:29:09 +02:00
/** This function is just used to create tests with time gaps. */
longlong Item_func_sleep::val_int()
{
THD *thd= current_thd;
Interruptible_wait timed_cond(thd);
mysql_cond_t cond;
double timeout;
int error;
DBUG_ASSERT(fixed == 1);
timeout= args[0]->val_real();
/*
On 64-bit OSX mysql_cond_timedwait() waits forever
if passed abstime time has already been exceeded by
the system time.
When given a very short timeout (< 10 mcs) just return
immediately.
We assume that the lines between this test and the call
to mysql_cond_timedwait() will be executed in less than 0.00001 sec.
*/
if (timeout < 0.00001)
return 0;
timed_cond.set_timeout((ulonglong) (timeout * 1000000000.0));
mysql_cond_init(key_item_func_sleep_cond, &cond, NULL);
mysql_mutex_lock(&LOCK_user_locks);
thd_proc_info(thd, "User sleep");
thd->mysys_var->current_mutex= &LOCK_user_locks;
thd->mysys_var->current_cond= &cond;
error= 0;
2011-03-04 13:12:31 +01:00
thd_wait_begin(thd, THD_WAIT_SLEEP);
while (!thd->killed)
{
error= timed_cond.wait(&cond, &LOCK_user_locks);
if (error == ETIMEDOUT || error == ETIME)
break;
error= 0;
}
2011-03-04 13:12:31 +01:00
thd_wait_end(thd);
thd_proc_info(thd, 0);
mysql_mutex_unlock(&LOCK_user_locks);
mysql_mutex_lock(&thd->mysys_var->mutex);
thd->mysys_var->current_mutex= 0;
thd->mysys_var->current_cond= 0;
mysql_mutex_unlock(&thd->mysys_var->mutex);
mysql_cond_destroy(&cond);
return test(!error); // Return 1 killed
}
2000-07-31 21:29:14 +02:00
#define extra_size sizeof(double)
static user_var_entry *get_variable(HASH *hash, LEX_STRING &name,
bool create_if_not_exists)
{
user_var_entry *entry;
if (!(entry = (user_var_entry*) my_hash_search(hash, (uchar*) name.str,
name.length)) &&
2000-07-31 21:29:14 +02:00
create_if_not_exists)
{
uint size=ALIGN_SIZE(sizeof(user_var_entry))+name.length+1+extra_size;
if (!my_hash_inited(hash))
2000-07-31 21:29:14 +02:00
return 0;
if (!(entry = (user_var_entry*) my_malloc(size,MYF(MY_WME | ME_FATALERROR))))
2000-07-31 21:29:14 +02:00
return 0;
entry->name.str=(char*) entry+ ALIGN_SIZE(sizeof(user_var_entry))+
extra_size;
entry->name.length=name.length;
entry->value=0;
entry->length=0;
entry->update_query_id=0;
entry->collation.set(NULL, DERIVATION_IMPLICIT, 0);
2006-09-13 19:25:33 +02:00
entry->unsigned_flag= 0;
/*
If we are here, we were called from a SET or a query which sets a
variable. Imagine it is this:
INSERT INTO t SELECT @a:=10, @a:=@a+1.
Then when we have a Item_func_get_user_var (because of the @a+1) so we
think we have to write the value of @a to the binlog. But before that,
we have a Item_func_set_user_var to create @a (@a:=10), in this we mark
the variable as "already logged" (line below) so that it won't be logged
by Item_func_get_user_var (because that's not necessary).
*/
2003-01-30 18:39:54 +01:00
entry->used_query_id=current_thd->query_id;
2000-07-31 21:29:14 +02:00
entry->type=STRING_RESULT;
memcpy(entry->name.str, name.str, name.length+1);
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (my_hash_insert(hash,(uchar*) entry))
2000-07-31 21:29:14 +02:00
{
Bug#34043: Server loops excessively in _checkchunk() when safemalloc is enabled Essentially, the problem is that safemalloc is excruciatingly slow as it checks all allocated blocks for overrun at each memory management primitive, yielding a almost exponential slowdown for the memory management functions (malloc, realloc, free). The overrun check basically consists of verifying some bytes of a block for certain magic keys, which catches some simple forms of overrun. Another minor problem is violation of aliasing rules and that its own internal list of blocks is prone to corruption. Another issue with safemalloc is rather the maintenance cost as the tool has a significant impact on the server code. Given the magnitude of memory debuggers available nowadays, especially those that are provided with the platform malloc implementation, maintenance of a in-house and largely obsolete memory debugger becomes a burden that is not worth the effort due to its slowness and lack of support for detecting more common forms of heap corruption. Since there are third-party tools that can provide the same functionality at a lower or comparable performance cost, the solution is to simply remove safemalloc. Third-party tools can provide the same functionality at a lower or comparable performance cost. The removal of safemalloc also allows a simplification of the malloc wrappers, removing quite a bit of kludge: redefinition of my_malloc, my_free and the removal of the unused second argument of my_free. Since free() always check whether the supplied pointer is null, redudant checks are also removed. Also, this patch adds unit testing for my_malloc and moves my_realloc implementation into the same file as the other memory allocation primitives.
2010-07-08 23:20:08 +02:00
my_free(entry);
2000-07-31 21:29:14 +02:00
return 0;
}
}
return entry;
}
void Item_func_set_user_var::cleanup()
{
Item_func::cleanup();
entry= NULL;
}
bool Item_func_set_user_var::set_entry(THD *thd, bool create_if_not_exists)
{
if (entry && thd->thread_id == entry_thread_id)
goto end; // update entry->update_query_id for PS
if (!(entry= get_variable(&thd->user_vars, name, create_if_not_exists)))
{
entry_thread_id= 0;
return TRUE;
}
entry_thread_id= thd->thread_id;
Bug#14466617 - INVALID WRITES AND/OR CRASH WITH USER VARIABLES Analysis: ------------- After executing the query, new value of the user defined variables are set in the function "select_dumpvar::send_data". "select_dumpvar::send_data" first calls function "Item_func_set_user_var::save_item_result()". This function checks the nullness of the Item_field passed as parameter to it and saves it. The nullness of item is stored with arg[0]'s null_value flag. Then "select_dumpvar::send_data" calls "Item_func_set_user_var::update()" which notices null result that was saved and calls "Item_func_set_user_var:: update_hash". But here null_value is not set and args[0] is different from that given to function "Item_func_set_user_var:: set_item_result()". This causes "Item_func_set_user_var:: update_hash" function to believe that its getting non-null value. "user_var_entry::length" set to 0 and hence "user_var_entry::value" is made to point to extra_area allocated in "user_var_entry". And "Item_func_set_user_var::update_hash" tries to write at memory beyond extra_area for result type DECIMAL. Because of this invalid write issue is reported by Valgrind. Before this bug was introduced, we avoided this problem by creating "Item_func_set_user_var" object with the same Item_field as arg[0] and as parameter to Item_func_set_user_var::save_item_result(). But now they are refering to different args[0]. Because of this null_value flag set in parameter Item_field in function "Item_func_set_user_var::save_item_result()" is not reflected in "Item_func_set_user_var" object. Fix: ------------ This issue is reported on versions 5.5.24. Issue does not exists in 5.5.23, 5.1, 5.6 and trunk. This issue was introduced by revid:georgi.kodinov@oracle.com-20120309130449-82e3bs5v3et1x0ef (fix for bug #12408412), which was pushed into 5.5 and later releases. This patch has later been reversed in 5.6 and trunk by revid:norvald.ryeng@oracle.com-20121010135242-xj34gg73h04hrmyh (fix for bug #14664077). Backported this patch in 5.5 also to fix this issue.
2012-11-07 14:38:33 +01:00
end:
Bug#14466617 - INVALID WRITES AND/OR CRASH WITH USER VARIABLES Analysis: ------------- After executing the query, new value of the user defined variables are set in the function "select_dumpvar::send_data". "select_dumpvar::send_data" first calls function "Item_func_set_user_var::save_item_result()". This function checks the nullness of the Item_field passed as parameter to it and saves it. The nullness of item is stored with arg[0]'s null_value flag. Then "select_dumpvar::send_data" calls "Item_func_set_user_var::update()" which notices null result that was saved and calls "Item_func_set_user_var:: update_hash". But here null_value is not set and args[0] is different from that given to function "Item_func_set_user_var:: set_item_result()". This causes "Item_func_set_user_var:: update_hash" function to believe that its getting non-null value. "user_var_entry::length" set to 0 and hence "user_var_entry::value" is made to point to extra_area allocated in "user_var_entry". And "Item_func_set_user_var::update_hash" tries to write at memory beyond extra_area for result type DECIMAL. Because of this invalid write issue is reported by Valgrind. Before this bug was introduced, we avoided this problem by creating "Item_func_set_user_var" object with the same Item_field as arg[0] and as parameter to Item_func_set_user_var::save_item_result(). But now they are refering to different args[0]. Because of this null_value flag set in parameter Item_field in function "Item_func_set_user_var::save_item_result()" is not reflected in "Item_func_set_user_var" object. Fix: ------------ This issue is reported on versions 5.5.24. Issue does not exists in 5.5.23, 5.1, 5.6 and trunk. This issue was introduced by revid:georgi.kodinov@oracle.com-20120309130449-82e3bs5v3et1x0ef (fix for bug #12408412), which was pushed into 5.5 and later releases. This patch has later been reversed in 5.6 and trunk by revid:norvald.ryeng@oracle.com-20121010135242-xj34gg73h04hrmyh (fix for bug #14664077). Backported this patch in 5.5 also to fix this issue.
2012-11-07 14:38:33 +01:00
/*
Remember the last query which updated it, this way a query can later know
if this variable is a constant item in the query (it is if update_query_id
is different from query_id).
If this object has delayed setting of non-constness, we delay this
until Item_func_set-user_var::save_item_result()
*/
if (!delayed_non_constness)
entry->update_query_id= thd->query_id;
return FALSE;
}
/*
When a user variable is updated (in a SET command or a query like
SELECT @a:= ).
*/
bool Item_func_set_user_var::fix_fields(THD *thd, Item **ref)
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 0);
/* fix_fields will call Item_func_set_user_var::fix_length_and_dec */
if (Item_func::fix_fields(thd, ref) || set_entry(thd, TRUE))
return TRUE;
/*
As it is wrong and confusing to associate any
character set with NULL, @a should be latin2
after this query sequence:
SET @a=_latin2'string';
SET @a=NULL;
I.e. the second query should not change the charset
to the current default value, but should keep the
original value assigned during the first query.
In order to do it, we don't copy charset
from the argument if the argument is NULL
and the variable has previously been initialized.
*/
null_item= (args[0]->type() == NULL_ITEM);
if (!entry->collation.collation || !null_item)
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
entry->collation.set(args[0]->collation.derivation == DERIVATION_NUMERIC ?
default_charset() : args[0]->collation.collation,
DERIVATION_IMPLICIT);
collation.set(entry->collation.collation, DERIVATION_IMPLICIT);
cached_result_type= args[0]->result_type();
return FALSE;
2000-07-31 21:29:14 +02:00
}
void
Item_func_set_user_var::fix_length_and_dec()
{
maybe_null=args[0]->maybe_null;
decimals=args[0]->decimals;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
collation.set(DERIVATION_IMPLICIT);
if (args[0]->collation.derivation == DERIVATION_NUMERIC)
fix_length_and_charset(args[0]->max_char_length(), default_charset());
else
{
fix_length_and_charset(args[0]->max_char_length(),
args[0]->collation.collation);
}
unsigned_flag= args[0]->unsigned_flag;
2000-07-31 21:29:14 +02:00
}
/*
Mark field in read_map
NOTES
This is used by filesort to register used fields in a a temporary
column read set or to register used fields in a view
*/
bool Item_func_set_user_var::register_field_in_read_map(uchar *arg)
{
if (result_field)
{
TABLE *table= (TABLE *) arg;
if (result_field->table == table || !table)
bitmap_set_bit(result_field->table->read_set, result_field->field_index);
}
return 0;
}
2007-10-11 19:29:09 +02:00
/**
Set value to user variable.
2007-10-11 19:29:09 +02:00
@param entry pointer to structure representing variable
@param set_null should we set NULL value ?
@param ptr pointer to buffer with new value
@param length length of new value
@param type type of new value
@param cs charset info for new value
@param dv derivation for new value
@param unsigned_arg indiates if a value of type INT_RESULT is unsigned
@note Sets error and fatal error if allocation fails.
2007-10-11 19:29:09 +02:00
@retval
false success
@retval
true failure
*/
static bool
update_hash(user_var_entry *entry, bool set_null, void *ptr, uint length,
Item_result type, CHARSET_INFO *cs, Derivation dv,
bool unsigned_arg)
2000-07-31 21:29:14 +02:00
{
if (set_null)
2000-07-31 21:29:14 +02:00
{
char *pos= (char*) entry+ ALIGN_SIZE(sizeof(user_var_entry));
if (entry->value && entry->value != pos)
Bug#34043: Server loops excessively in _checkchunk() when safemalloc is enabled Essentially, the problem is that safemalloc is excruciatingly slow as it checks all allocated blocks for overrun at each memory management primitive, yielding a almost exponential slowdown for the memory management functions (malloc, realloc, free). The overrun check basically consists of verifying some bytes of a block for certain magic keys, which catches some simple forms of overrun. Another minor problem is violation of aliasing rules and that its own internal list of blocks is prone to corruption. Another issue with safemalloc is rather the maintenance cost as the tool has a significant impact on the server code. Given the magnitude of memory debuggers available nowadays, especially those that are provided with the platform malloc implementation, maintenance of a in-house and largely obsolete memory debugger becomes a burden that is not worth the effort due to its slowness and lack of support for detecting more common forms of heap corruption. Since there are third-party tools that can provide the same functionality at a lower or comparable performance cost, the solution is to simply remove safemalloc. Third-party tools can provide the same functionality at a lower or comparable performance cost. The removal of safemalloc also allows a simplification of the malloc wrappers, removing quite a bit of kludge: redefinition of my_malloc, my_free and the removal of the unused second argument of my_free. Since free() always check whether the supplied pointer is null, redudant checks are also removed. Also, this patch adds unit testing for my_malloc and moves my_realloc implementation into the same file as the other memory allocation primitives.
2010-07-08 23:20:08 +02:00
my_free(entry->value);
entry->value= 0;
entry->length= 0;
2000-07-31 21:29:14 +02:00
}
else
{
if (type == STRING_RESULT)
length++; // Store strings with end \0
2000-07-31 21:29:14 +02:00
if (length <= extra_size)
{
/* Save value in value struct */
char *pos= (char*) entry+ ALIGN_SIZE(sizeof(user_var_entry));
if (entry->value != pos)
{
if (entry->value)
Bug#34043: Server loops excessively in _checkchunk() when safemalloc is enabled Essentially, the problem is that safemalloc is excruciatingly slow as it checks all allocated blocks for overrun at each memory management primitive, yielding a almost exponential slowdown for the memory management functions (malloc, realloc, free). The overrun check basically consists of verifying some bytes of a block for certain magic keys, which catches some simple forms of overrun. Another minor problem is violation of aliasing rules and that its own internal list of blocks is prone to corruption. Another issue with safemalloc is rather the maintenance cost as the tool has a significant impact on the server code. Given the magnitude of memory debuggers available nowadays, especially those that are provided with the platform malloc implementation, maintenance of a in-house and largely obsolete memory debugger becomes a burden that is not worth the effort due to its slowness and lack of support for detecting more common forms of heap corruption. Since there are third-party tools that can provide the same functionality at a lower or comparable performance cost, the solution is to simply remove safemalloc. Third-party tools can provide the same functionality at a lower or comparable performance cost. The removal of safemalloc also allows a simplification of the malloc wrappers, removing quite a bit of kludge: redefinition of my_malloc, my_free and the removal of the unused second argument of my_free. Since free() always check whether the supplied pointer is null, redudant checks are also removed. Also, this patch adds unit testing for my_malloc and moves my_realloc implementation into the same file as the other memory allocation primitives.
2010-07-08 23:20:08 +02:00
my_free(entry->value);
2000-07-31 21:29:14 +02:00
entry->value=pos;
}
}
else
{
/* Allocate variable */
if (entry->length != length)
{
char *pos= (char*) entry+ ALIGN_SIZE(sizeof(user_var_entry));
if (entry->value == pos)
entry->value=0;
entry->value= (char*) my_realloc(entry->value, length,
MYF(MY_ALLOW_ZERO_PTR | MY_WME |
ME_FATALERROR));
if (!entry->value)
return 1;
2000-07-31 21:29:14 +02:00
}
}
if (type == STRING_RESULT)
{
length--; // Fix length change above
entry->value[length]= 0; // Store end \0
}
memmove(entry->value, ptr, length);
2005-02-08 23:50:45 +01:00
if (type == DECIMAL_RESULT)
((my_decimal*)entry->value)->fix_buffer_pointer();
2000-07-31 21:29:14 +02:00
entry->length= length;
entry->collation.set(cs, dv);
2006-09-13 19:25:33 +02:00
entry->unsigned_flag= unsigned_arg;
2000-07-31 21:29:14 +02:00
}
entry->type=type;
return 0;
}
2000-07-31 21:29:14 +02:00
bool
Item_func_set_user_var::update_hash(void *ptr, uint length,
Item_result res_type,
CHARSET_INFO *cs, Derivation dv,
bool unsigned_arg)
{
/*
If we set a variable explicitely to NULL then keep the old
result type of the variable
*/
if ((null_value= args[0]->null_value) && null_item)
res_type= entry->type; // Don't change type of item
if (::update_hash(entry, (null_value= args[0]->null_value),
ptr, length, res_type, cs, dv, unsigned_arg))
{
null_value= 1;
return 1;
}
return 0;
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/** Get the value of a variable as a double. */
2005-02-08 23:50:45 +01:00
double user_var_entry::val_real(my_bool *null_value)
2000-07-31 21:29:14 +02:00
{
if ((*null_value= (value == 0)))
return 0.0;
switch (type) {
case REAL_RESULT:
return *(double*) value;
case INT_RESULT:
return (double) *(longlong*) value;
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
double result;
my_decimal2double(E_DEC_FATAL_ERROR, (my_decimal *)value, &result);
return result;
}
case STRING_RESULT:
return my_atof(value); // This is null terminated
case ROW_RESULT:
DBUG_ASSERT(1); // Impossible
break;
}
return 0.0; // Impossible
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/** Get the value of a variable as an integer. */
longlong user_var_entry::val_int(my_bool *null_value) const
2000-07-31 21:29:14 +02:00
{
if ((*null_value= (value == 0)))
return LL(0);
switch (type) {
case REAL_RESULT:
return (longlong) *(double*) value;
case INT_RESULT:
return *(longlong*) value;
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
longlong result;
my_decimal2int(E_DEC_FATAL_ERROR, (my_decimal *)value, 0, &result);
2005-02-08 23:50:45 +01:00
return result;
}
case STRING_RESULT:
2004-05-12 01:38:57 +02:00
{
int error;
return my_strtoll10(value, (char**) 0, &error);// String is null terminated
}
case ROW_RESULT:
DBUG_ASSERT(1); // Impossible
break;
}
return LL(0); // Impossible
}
2007-10-11 19:29:09 +02:00
/** Get the value of a variable as a string. */
String *user_var_entry::val_str(my_bool *null_value, String *str,
uint decimals)
{
if ((*null_value= (value == 0)))
return (String*) 0;
switch (type) {
case REAL_RESULT:
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set_real(*(double*) value, decimals, collation.collation);
break;
case INT_RESULT:
if (!unsigned_flag)
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set(*(longlong*) value, collation.collation);
else
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str->set(*(ulonglong*) value, collation.collation);
break;
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
str_set_decimal((my_decimal *) value, str, collation.collation);
2005-02-08 23:50:45 +01:00
break;
case STRING_RESULT:
2003-10-07 14:42:26 +02:00
if (str->copy(value, length, collation.collation))
str= 0; // EOM error
case ROW_RESULT:
DBUG_ASSERT(1); // Impossible
break;
}
return(str);
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/** Get the value of a variable as a decimal. */
2005-02-08 23:50:45 +01:00
my_decimal *user_var_entry::val_decimal(my_bool *null_value, my_decimal *val)
{
if ((*null_value= (value == 0)))
return 0;
switch (type) {
case REAL_RESULT:
double2my_decimal(E_DEC_FATAL_ERROR, *(double*) value, val);
break;
case INT_RESULT:
int2my_decimal(E_DEC_FATAL_ERROR, *(longlong*) value, 0, val);
break;
case DECIMAL_RESULT:
my_decimal2decimal((my_decimal *) value, val);
2005-02-08 23:50:45 +01:00
break;
case STRING_RESULT:
str2my_decimal(E_DEC_FATAL_ERROR, value, length, collation.collation, val);
break;
case ROW_RESULT:
DBUG_ASSERT(1); // Impossible
break;
}
return(val);
}
2007-10-11 19:29:09 +02:00
/**
This functions is invoked on SET \@variable or
\@variable:= expression.
2007-10-11 19:29:09 +02:00
Evaluate (and check expression), store results.
2007-10-11 19:29:09 +02:00
@note
2005-02-08 23:50:45 +01:00
For now it always return OK. All problem with value evaluating
will be caught by thd->is_error() check in sql_set_variables().
2007-10-11 19:29:09 +02:00
@retval
FALSE OK.
*/
bool
Item_func_set_user_var::check(bool use_result_field)
{
DBUG_ENTER("Item_func_set_user_var::check");
if (use_result_field && !result_field)
use_result_field= FALSE;
switch (cached_result_type) {
case REAL_RESULT:
{
save_result.vreal= use_result_field ? result_field->val_real() :
args[0]->val_real();
break;
}
case INT_RESULT:
{
save_result.vint= use_result_field ? result_field->val_int() :
args[0]->val_int();
unsigned_flag= use_result_field ? ((Field_num*)result_field)->unsigned_flag:
args[0]->unsigned_flag;
break;
}
case STRING_RESULT:
{
save_result.vstr= use_result_field ? result_field->val_str(&value) :
args[0]->val_str(&value);
break;
}
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
save_result.vdec= use_result_field ?
result_field->val_decimal(&decimal_buff) :
args[0]->val_decimal(&decimal_buff);
2005-02-08 23:50:45 +01:00
break;
}
case ROW_RESULT:
default:
2005-02-08 23:50:45 +01:00
// This case should never be chosen
DBUG_ASSERT(0);
break;
}
DBUG_RETURN(FALSE);
}
/**
@brief Evaluate and store item's result.
This function is invoked on "SELECT ... INTO @var ...".
@param item An item to get value from.
*/
void Item_func_set_user_var::save_item_result(Item *item)
{
DBUG_ENTER("Item_func_set_user_var::save_item_result");
switch (cached_result_type) {
case REAL_RESULT:
save_result.vreal= item->val_result();
break;
case INT_RESULT:
save_result.vint= item->val_int_result();
unsigned_flag= item->unsigned_flag;
break;
case STRING_RESULT:
save_result.vstr= item->str_result(&value);
break;
case DECIMAL_RESULT:
save_result.vdec= item->val_decimal_result(&decimal_buff);
break;
case ROW_RESULT:
default:
// Should never happen
DBUG_ASSERT(0);
break;
}
Bug#14466617 - INVALID WRITES AND/OR CRASH WITH USER VARIABLES Analysis: ------------- After executing the query, new value of the user defined variables are set in the function "select_dumpvar::send_data". "select_dumpvar::send_data" first calls function "Item_func_set_user_var::save_item_result()". This function checks the nullness of the Item_field passed as parameter to it and saves it. The nullness of item is stored with arg[0]'s null_value flag. Then "select_dumpvar::send_data" calls "Item_func_set_user_var::update()" which notices null result that was saved and calls "Item_func_set_user_var:: update_hash". But here null_value is not set and args[0] is different from that given to function "Item_func_set_user_var:: set_item_result()". This causes "Item_func_set_user_var:: update_hash" function to believe that its getting non-null value. "user_var_entry::length" set to 0 and hence "user_var_entry::value" is made to point to extra_area allocated in "user_var_entry". And "Item_func_set_user_var::update_hash" tries to write at memory beyond extra_area for result type DECIMAL. Because of this invalid write issue is reported by Valgrind. Before this bug was introduced, we avoided this problem by creating "Item_func_set_user_var" object with the same Item_field as arg[0] and as parameter to Item_func_set_user_var::save_item_result(). But now they are refering to different args[0]. Because of this null_value flag set in parameter Item_field in function "Item_func_set_user_var::save_item_result()" is not reflected in "Item_func_set_user_var" object. Fix: ------------ This issue is reported on versions 5.5.24. Issue does not exists in 5.5.23, 5.1, 5.6 and trunk. This issue was introduced by revid:georgi.kodinov@oracle.com-20120309130449-82e3bs5v3et1x0ef (fix for bug #12408412), which was pushed into 5.5 and later releases. This patch has later been reversed in 5.6 and trunk by revid:norvald.ryeng@oracle.com-20121010135242-xj34gg73h04hrmyh (fix for bug #14664077). Backported this patch in 5.5 also to fix this issue.
2012-11-07 14:38:33 +01:00
/*
Set the ID of the query that last updated this variable. This is
usually set by Item_func_set_user_var::set_entry(), but if this
item has delayed setting of non-constness, we must do it now.
*/
if (delayed_non_constness)
entry->update_query_id= current_thd->query_id;
DBUG_VOID_RETURN;
}
2007-10-11 19:29:09 +02:00
/**
This functions is invoked on
SET \@variable or \@variable:= expression.
2007-10-11 19:29:09 +02:00
@note
We have to store the expression as such in the variable, independent of
the value method used by the user
2007-10-11 19:29:09 +02:00
@retval
2005-02-08 23:50:45 +01:00
0 OK
2007-10-11 19:29:09 +02:00
@retval
1 EOM Error
*/
2000-07-31 21:29:14 +02:00
bool
Item_func_set_user_var::update()
{
2009-09-24 15:21:46 +02:00
bool res= 0;
DBUG_ENTER("Item_func_set_user_var::update");
2000-07-31 21:29:14 +02:00
switch (cached_result_type) {
case REAL_RESULT:
{
res= update_hash((void*) &save_result.vreal,sizeof(save_result.vreal),
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
REAL_RESULT, default_charset(), DERIVATION_IMPLICIT, 0);
2000-07-31 21:29:14 +02:00
break;
}
2000-07-31 21:29:14 +02:00
case INT_RESULT:
{
res= update_hash((void*) &save_result.vint, sizeof(save_result.vint),
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
INT_RESULT, default_charset(), DERIVATION_IMPLICIT,
unsigned_flag);
2000-07-31 21:29:14 +02:00
break;
}
2000-07-31 21:29:14 +02:00
case STRING_RESULT:
2002-11-15 19:32:09 +01:00
{
if (!save_result.vstr) // Null value
2003-10-07 14:42:26 +02:00
res= update_hash((void*) 0, 0, STRING_RESULT, &my_charset_bin,
2006-09-13 19:25:33 +02:00
DERIVATION_IMPLICIT, 0);
else
res= update_hash((void*) save_result.vstr->ptr(),
save_result.vstr->length(), STRING_RESULT,
save_result.vstr->charset(),
2006-09-13 19:25:33 +02:00
DERIVATION_IMPLICIT, 0);
2000-07-31 21:29:14 +02:00
break;
}
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
{
if (!save_result.vdec) // Null value
res= update_hash((void*) 0, 0, DECIMAL_RESULT, &my_charset_bin,
2006-09-12 16:28:36 +02:00
DERIVATION_IMPLICIT, 0);
2005-02-08 23:50:45 +01:00
else
res= update_hash((void*) save_result.vdec,
sizeof(my_decimal), DECIMAL_RESULT,
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
default_charset(), DERIVATION_IMPLICIT, 0);
2005-02-08 23:50:45 +01:00
break;
}
case ROW_RESULT:
2003-01-31 11:07:07 +01:00
default:
2005-02-08 23:50:45 +01:00
// This case should never be chosen
2002-11-15 19:32:09 +01:00
DBUG_ASSERT(0);
break;
}
DBUG_RETURN(res);
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_set_user_var::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
check(0);
update(); // Store expression
2005-02-08 23:50:45 +01:00
return entry->val_real(&null_value);
2000-07-31 21:29:14 +02:00
}
longlong Item_func_set_user_var::val_int()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
check(0);
update(); // Store expression
return entry->val_int(&null_value);
2000-07-31 21:29:14 +02:00
}
String *Item_func_set_user_var::val_str(String *str)
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
check(0);
update(); // Store expression
return entry->val_str(&null_value, str, decimals);
2000-07-31 21:29:14 +02:00
}
2005-02-08 23:50:45 +01:00
my_decimal *Item_func_set_user_var::val_decimal(my_decimal *val)
{
DBUG_ASSERT(fixed == 1);
check(0);
2005-02-08 23:50:45 +01:00
update(); // Store expression
return entry->val_decimal(&null_value, val);
}
double Item_func_set_user_var::val_result()
{
DBUG_ASSERT(fixed == 1);
check(TRUE);
update(); // Store expression
return entry->val_real(&null_value);
}
longlong Item_func_set_user_var::val_int_result()
{
DBUG_ASSERT(fixed == 1);
check(TRUE);
update(); // Store expression
return entry->val_int(&null_value);
}
2010-08-13 13:20:49 +02:00
bool Item_func_set_user_var::val_bool_result()
{
DBUG_ASSERT(fixed == 1);
check(TRUE);
update(); // Store expression
return entry->val_int(&null_value) != 0;
}
String *Item_func_set_user_var::str_result(String *str)
{
DBUG_ASSERT(fixed == 1);
check(TRUE);
update(); // Store expression
return entry->val_str(&null_value, str, decimals);
}
my_decimal *Item_func_set_user_var::val_decimal_result(my_decimal *val)
{
DBUG_ASSERT(fixed == 1);
check(TRUE);
update(); // Store expression
return entry->val_decimal(&null_value, val);
2005-02-08 23:50:45 +01:00
}
bool Item_func_set_user_var::is_null_result()
{
DBUG_ASSERT(fixed == 1);
check(TRUE);
update(); // Store expression
return is_null();
}
void Item_func_set_user_var::print(String *str, enum_query_type query_type)
2001-12-27 01:04:27 +01:00
{
str->append(STRING_WITH_LEN("(@"));
str->append(name.str, name.length);
str->append(STRING_WITH_LEN(":="));
args[0]->print(str, query_type);
2001-12-27 01:04:27 +01:00
str->append(')');
}
void Item_func_set_user_var::print_as_stmt(String *str,
enum_query_type query_type)
{
str->append(STRING_WITH_LEN("set @"));
str->append(name.str, name.length);
str->append(STRING_WITH_LEN(":="));
args[0]->print(str, query_type);
str->append(')');
}
bool Item_func_set_user_var::send(Protocol *protocol, String *str_arg)
{
if (result_field)
{
check(1);
update();
return protocol->store(result_field);
}
return Item::send(protocol, str_arg);
}
void Item_func_set_user_var::make_field(Send_field *tmp_field)
{
if (result_field)
{
result_field->make_field(tmp_field);
DBUG_ASSERT(tmp_field->table_name != 0);
if (Item::name)
tmp_field->col_name=Item::name; // Use user supplied name
}
else
Item::make_field(tmp_field);
}
/*
Save the value of a user variable into a field
SYNOPSIS
save_in_field()
field target field to save the value to
no_conversion flag indicating whether conversions are allowed
DESCRIPTION
Save the function value into a field and update the user variable
accordingly. If a result field is defined and the target field doesn't
coincide with it then the value from the result field will be used as
the new value of the user variable.
The reason to have this method rather than simply using the result
field in the val_xxx() methods is that the value from the result field
not always can be used when the result field is defined.
Let's consider the following cases:
1) when filling a tmp table the result field is defined but the value of it
is undefined because it has to be produced yet. Thus we can't use it.
2) on execution of an INSERT ... SELECT statement the save_in_field()
function will be called to fill the data in the new record. If the SELECT
part uses a tmp table then the result field is defined and should be
used in order to get the correct result.
The difference between the SET_USER_VAR function and regular functions
like CONCAT is that the Item_func objects for the regular functions are
replaced by Item_field objects after the values of these functions have
been stored in a tmp table. Yet an object of the Item_field class cannot
be used to update a user variable.
Due to this we have to handle the result field in a special way here and
in the Item_func_set_user_var::send() function.
RETURN VALUES
FALSE Ok
TRUE Error
*/
int Item_func_set_user_var::save_in_field(Field *field, bool no_conversions,
bool can_use_result_field)
{
bool use_result_field= (!can_use_result_field ? 0 :
(result_field && result_field != field));
int error;
/* Update the value of the user variable */
check(use_result_field);
update();
if (result_type() == STRING_RESULT ||
(result_type() == REAL_RESULT &&
field->result_type() == STRING_RESULT))
{
String *result;
CHARSET_INFO *cs= collation.collation;
char buff[MAX_FIELD_WIDTH]; // Alloc buffer for small columns
str_value.set_quick(buff, sizeof(buff), cs);
result= entry->val_str(&null_value, &str_value, decimals);
if (null_value)
{
str_value.set_quick(0, 0, cs);
return set_field_to_null_with_conversions(field, no_conversions);
}
/* NOTE: If null_value == FALSE, "result" must be not NULL. */
field->set_notnull();
error=field->store(result->ptr(),result->length(),cs);
str_value.set_quick(0, 0, cs);
}
else if (result_type() == REAL_RESULT)
{
double nr= entry->val_real(&null_value);
if (null_value)
return set_field_to_null(field);
field->set_notnull();
error=field->store(nr);
}
else if (result_type() == DECIMAL_RESULT)
{
my_decimal decimal_value;
my_decimal *val= entry->val_decimal(&null_value, &decimal_value);
if (null_value)
return set_field_to_null(field);
field->set_notnull();
error=field->store_decimal(val);
}
else
{
longlong nr= entry->val_int(&null_value);
if (null_value)
return set_field_to_null_with_conversions(field, no_conversions);
field->set_notnull();
error=field->store(nr, unsigned_flag);
}
return error;
}
2000-07-31 21:29:14 +02:00
String *
Item_func_get_user_var::val_str(String *str)
{
DBUG_ASSERT(fixed == 1);
DBUG_ENTER("Item_func_get_user_var::val_str");
if (!var_entry)
DBUG_RETURN((String*) 0); // No such variable
DBUG_RETURN(var_entry->val_str(&null_value, str, decimals));
2000-07-31 21:29:14 +02:00
}
2004-11-11 19:39:35 +01:00
double Item_func_get_user_var::val_real()
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 1);
if (!var_entry)
return 0.0; // No such variable
2005-02-08 23:50:45 +01:00
return (var_entry->val_real(&null_value));
}
my_decimal *Item_func_get_user_var::val_decimal(my_decimal *dec)
{
DBUG_ASSERT(fixed == 1);
if (!var_entry)
return 0;
return var_entry->val_decimal(&null_value, dec);
2000-07-31 21:29:14 +02:00
}
longlong Item_func_get_user_var::val_int()
{
DBUG_ASSERT(fixed == 1);
if (!var_entry)
return LL(0); // No such variable
return (var_entry->val_int(&null_value));
2000-07-31 21:29:14 +02:00
}
2007-10-11 19:29:09 +02:00
/**
Get variable by name and, if necessary, put the record of variable
use into the binary log.
When a user variable is invoked from an update query (INSERT, UPDATE etc),
stores this variable and its value in thd->user_var_events, so that it can be
written to the binlog (will be written just before the query is written, see
log.cc).
2007-10-11 19:29:09 +02:00
@param thd Current thread
@param name Variable name
@param[out] out_entry variable structure or NULL. The pointer is set
regardless of whether function succeeded or not.
@retval
0 OK
2007-10-11 19:29:09 +02:00
@retval
2005-02-08 23:50:45 +01:00
1 Failed to put appropriate record into binary log
*/
static int
get_var_with_binlog(THD *thd, enum_sql_command sql_command,
LEX_STRING &name, user_var_entry **out_entry)
2000-07-31 21:29:14 +02:00
{
BINLOG_USER_VAR_EVENT *user_var_event;
user_var_entry *var_entry;
var_entry= get_variable(&thd->user_vars, name, 0);
2007-02-26 20:06:10 +01:00
/*
Any reference to user-defined variable which is done from stored
2007-02-23 18:58:56 +01:00
function or trigger affects their execution and the execution of the
calling statement. We must log all such variables even if they are
not involved in table-updating statements.
2007-02-26 20:06:10 +01:00
*/
if (!(opt_bin_log &&
(is_update_query(sql_command) || thd->in_sub_stmt)))
{
*out_entry= var_entry;
return 0;
}
if (!var_entry)
2003-01-30 18:39:54 +01:00
{
/*
If the variable does not exist, it's NULL, but we want to create it so
that it gets into the binlog (if it didn't, the slave could be
influenced by a variable of the same name previously set by another
thread).
2005-02-08 23:50:45 +01:00
We create it like if it had been explicitly set with SET before.
The 'new' mimics what sql_yacc.yy does when 'SET @a=10;'.
sql_set_variables() is what is called from 'case SQLCOM_SET_OPTION'
in dispatch_command()). Instead of building a one-element list to pass to
sql_set_variables(), we could instead manually call check() and update();
this would save memory and time; but calling sql_set_variables() makes
one unique place to maintain (sql_set_variables()).
Manipulation with lex is necessary since free_underlaid_joins
is going to release memory belonging to the main query.
*/
List<set_var_base> tmp_var_list;
LEX *sav_lex= thd->lex, lex_tmp;
thd->lex= &lex_tmp;
Bug#25411 (trigger code truncated), PART I The issue found with bug 25411 is due to the function skip_rear_comments() which damages the source code while implementing a work around. The root cause of the problem is in the lexical analyser, which does not process special comments properly. For special comments like : [1] aaa /*!50000 bbb */ ccc since 5.0 is a version older that the current code, the parser is in lining the content of the special comment, so that the query to process is [2] aaa bbb ccc However, the text of the query captured when processing a stored procedure, stored function or trigger (or event in 5.1), can be after rebuilding it: [3] aaa bbb */ ccc which is wrong. To fix bug 25411 properly, the lexical analyser needs to return [2] when in lining special comments. In order to implement this, some preliminary cleanup is required in the code, which is implemented by this patch. Before this change, the structure named LEX (or st_lex) contains attributes that belong to lexical analysis, as well as attributes that represents the abstract syntax tree (AST) of a statement. Creating a new LEX structure for each statements (which makes sense for the AST part) also re-initialized the lexical analysis phase each time, which is conceptually wrong. With this patch, the previous st_lex structure has been split in two: - st_lex represents the Abstract Syntax Tree for a statement. The name "lex" has not been changed to avoid a bigger impact in the code base. - class lex_input_stream represents the internal state of the lexical analyser, which by definition should *not* be reinitialized when parsing multiple statements from the same input stream. This change is a pre-requisite for bug 25411, since the implementation of lex_input_stream will later improve to deal properly with special comments, and this processing can not be done with the current implementation of sp_head::reset_lex and sp_head::restore_lex, which interfere with the lexer. This change set alone does not fix bug 25411.
2007-04-24 17:24:21 +02:00
lex_start(thd);
tmp_var_list.push_back(new set_var_user(new Item_func_set_user_var(name,
Bug#14466617 - INVALID WRITES AND/OR CRASH WITH USER VARIABLES Analysis: ------------- After executing the query, new value of the user defined variables are set in the function "select_dumpvar::send_data". "select_dumpvar::send_data" first calls function "Item_func_set_user_var::save_item_result()". This function checks the nullness of the Item_field passed as parameter to it and saves it. The nullness of item is stored with arg[0]'s null_value flag. Then "select_dumpvar::send_data" calls "Item_func_set_user_var::update()" which notices null result that was saved and calls "Item_func_set_user_var:: update_hash". But here null_value is not set and args[0] is different from that given to function "Item_func_set_user_var:: set_item_result()". This causes "Item_func_set_user_var:: update_hash" function to believe that its getting non-null value. "user_var_entry::length" set to 0 and hence "user_var_entry::value" is made to point to extra_area allocated in "user_var_entry". And "Item_func_set_user_var::update_hash" tries to write at memory beyond extra_area for result type DECIMAL. Because of this invalid write issue is reported by Valgrind. Before this bug was introduced, we avoided this problem by creating "Item_func_set_user_var" object with the same Item_field as arg[0] and as parameter to Item_func_set_user_var::save_item_result(). But now they are refering to different args[0]. Because of this null_value flag set in parameter Item_field in function "Item_func_set_user_var::save_item_result()" is not reflected in "Item_func_set_user_var" object. Fix: ------------ This issue is reported on versions 5.5.24. Issue does not exists in 5.5.23, 5.1, 5.6 and trunk. This issue was introduced by revid:georgi.kodinov@oracle.com-20120309130449-82e3bs5v3et1x0ef (fix for bug #12408412), which was pushed into 5.5 and later releases. This patch has later been reversed in 5.6 and trunk by revid:norvald.ryeng@oracle.com-20121010135242-xj34gg73h04hrmyh (fix for bug #14664077). Backported this patch in 5.5 also to fix this issue.
2012-11-07 14:38:33 +01:00
new Item_null(),
false)));
/* Create the variable */
if (sql_set_variables(thd, &tmp_var_list))
{
thd->lex= sav_lex;
goto err;
}
thd->lex= sav_lex;
if (!(var_entry= get_variable(&thd->user_vars, name, 0)))
goto err;
}
else if (var_entry->used_query_id == thd->query_id ||
mysql_bin_log.is_query_in_union(thd, var_entry->used_query_id))
{
/*
If this variable was already stored in user_var_events by this query
(because it's used in more than one place in the query), don't store
it.
*/
*out_entry= var_entry;
return 0;
}
uint size;
/*
First we need to store value of var_entry, when the next situation
2005-02-08 23:50:45 +01:00
appears:
> set @a:=1;
> insert into t1 values (@a), (@a:=@a+1), (@a:=@a+1);
We have to write to binlog value @a= 1.
We allocate the user_var_event on user_var_events_alloc pool, not on
the this-statement-execution pool because in SPs user_var_event objects
may need to be valid after current [SP] statement execution pool is
destroyed.
*/
size= ALIGN_SIZE(sizeof(BINLOG_USER_VAR_EVENT)) + var_entry->length;
if (!(user_var_event= (BINLOG_USER_VAR_EVENT *)
alloc_root(thd->user_var_events_alloc, size)))
goto err;
user_var_event->value= (char*) user_var_event +
ALIGN_SIZE(sizeof(BINLOG_USER_VAR_EVENT));
user_var_event->user_var_event= var_entry;
user_var_event->type= var_entry->type;
user_var_event->charset_number= var_entry->collation.collation->number;
user_var_event->unsigned_flag= var_entry->unsigned_flag;
if (!var_entry->value)
{
/* NULL value*/
user_var_event->length= 0;
user_var_event->value= 0;
2003-01-30 18:39:54 +01:00
}
else
{
user_var_event->length= var_entry->length;
memcpy(user_var_event->value, var_entry->value,
var_entry->length);
}
/* Mark that this variable has been used by this query */
var_entry->used_query_id= thd->query_id;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (insert_dynamic(&thd->user_var_events, (uchar*) &user_var_event))
goto err;
*out_entry= var_entry;
return 0;
2003-01-30 18:39:54 +01:00
err:
*out_entry= var_entry;
return 1;
}
void Item_func_get_user_var::fix_length_and_dec()
{
THD *thd=current_thd;
int error;
maybe_null=1;
decimals=NOT_FIXED_DEC;
max_length=MAX_BLOB_WIDTH;
error= get_var_with_binlog(thd, thd->lex->sql_command, name, &var_entry);
/*
If the variable didn't exist it has been created as a STRING-type.
'var_entry' is NULL only if there occured an error during the call to
get_var_with_binlog.
*/
if (!error && var_entry)
{
m_cached_result_type= var_entry->type;
unsigned_flag= var_entry->unsigned_flag;
max_length= var_entry->length;
collation.set(var_entry->collation);
switch(m_cached_result_type) {
case REAL_RESULT:
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
fix_char_length(DBL_DIG + 8);
break;
case INT_RESULT:
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
fix_char_length(MAX_BIGINT_WIDTH);
decimals=0;
break;
case STRING_RESULT:
max_length= MAX_BLOB_WIDTH - 1;
break;
2005-02-08 23:50:45 +01:00
case DECIMAL_RESULT:
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
fix_char_length(DECIMAL_MAX_STR_LENGTH);
decimals= DECIMAL_MAX_SCALE;
2005-02-08 23:50:45 +01:00
break;
case ROW_RESULT: // Keep compiler happy
2005-02-08 23:50:45 +01:00
default:
DBUG_ASSERT(0);
break;
}
}
else
{
collation.set(&my_charset_bin, DERIVATION_IMPLICIT);
null_value= 1;
m_cached_result_type= STRING_RESULT;
max_length= MAX_BLOB_WIDTH;
}
2000-07-31 21:29:14 +02:00
}
bool Item_func_get_user_var::const_item() const
{
return (!var_entry || current_thd->query_id != var_entry->update_query_id);
}
2000-07-31 21:29:14 +02:00
enum Item_result Item_func_get_user_var::result_type() const
{
return m_cached_result_type;
2000-07-31 21:29:14 +02:00
}
2001-12-27 01:04:27 +01:00
void Item_func_get_user_var::print(String *str, enum_query_type query_type)
2001-12-27 01:04:27 +01:00
{
str->append(STRING_WITH_LEN("(@"));
2001-12-27 01:04:27 +01:00
str->append(name.str,name.length);
str->append(')');
}
bool Item_func_get_user_var::eq(const Item *item, bool binary_cmp) const
2001-12-27 01:04:27 +01:00
{
/* Assume we don't have rtti */
if (this == item)
return 1; // Same item is same.
/* Check if other type is also a get_user_var() object */
if (item->type() != FUNC_ITEM ||
((Item_func*) item)->functype() != functype())
2001-12-27 01:04:27 +01:00
return 0;
Item_func_get_user_var *other=(Item_func_get_user_var*) item;
return (name.length == other->name.length &&
!memcmp(name.str, other->name.str, name.length));
}
bool Item_func_get_user_var::set_value(THD *thd,
sp_rcontext * /*ctx*/, Item **it)
{
Bug#14466617 - INVALID WRITES AND/OR CRASH WITH USER VARIABLES Analysis: ------------- After executing the query, new value of the user defined variables are set in the function "select_dumpvar::send_data". "select_dumpvar::send_data" first calls function "Item_func_set_user_var::save_item_result()". This function checks the nullness of the Item_field passed as parameter to it and saves it. The nullness of item is stored with arg[0]'s null_value flag. Then "select_dumpvar::send_data" calls "Item_func_set_user_var::update()" which notices null result that was saved and calls "Item_func_set_user_var:: update_hash". But here null_value is not set and args[0] is different from that given to function "Item_func_set_user_var:: set_item_result()". This causes "Item_func_set_user_var:: update_hash" function to believe that its getting non-null value. "user_var_entry::length" set to 0 and hence "user_var_entry::value" is made to point to extra_area allocated in "user_var_entry". And "Item_func_set_user_var::update_hash" tries to write at memory beyond extra_area for result type DECIMAL. Because of this invalid write issue is reported by Valgrind. Before this bug was introduced, we avoided this problem by creating "Item_func_set_user_var" object with the same Item_field as arg[0] and as parameter to Item_func_set_user_var::save_item_result(). But now they are refering to different args[0]. Because of this null_value flag set in parameter Item_field in function "Item_func_set_user_var::save_item_result()" is not reflected in "Item_func_set_user_var" object. Fix: ------------ This issue is reported on versions 5.5.24. Issue does not exists in 5.5.23, 5.1, 5.6 and trunk. This issue was introduced by revid:georgi.kodinov@oracle.com-20120309130449-82e3bs5v3et1x0ef (fix for bug #12408412), which was pushed into 5.5 and later releases. This patch has later been reversed in 5.6 and trunk by revid:norvald.ryeng@oracle.com-20121010135242-xj34gg73h04hrmyh (fix for bug #14664077). Backported this patch in 5.5 also to fix this issue.
2012-11-07 14:38:33 +01:00
Item_func_set_user_var *suv= new Item_func_set_user_var(get_name(), *it, false);
/*
Item_func_set_user_var is not fixed after construction, call
fix_fields().
*/
return (!suv || suv->fix_fields(thd, it) || suv->check(0) || suv->update());
}
bool Item_user_var_as_out_param::fix_fields(THD *thd, Item **ref)
{
DBUG_ASSERT(fixed == 0);
DBUG_ASSERT(thd->lex->exchange);
if (Item::fix_fields(thd, ref) ||
!(entry= get_variable(&thd->user_vars, name, 1)))
return TRUE;
entry->type= STRING_RESULT;
/*
Let us set the same collation which is used for loading
of fields in LOAD DATA INFILE.
(Since Item_user_var_as_out_param is used only there).
*/
entry->collation.set(thd->lex->exchange->cs ?
thd->lex->exchange->cs :
thd->variables.collation_database);
entry->update_query_id= thd->query_id;
return FALSE;
}
void Item_user_var_as_out_param::set_null_value(CHARSET_INFO* cs)
{
::update_hash(entry, TRUE, 0, 0, STRING_RESULT, cs,
DERIVATION_IMPLICIT, 0 /* unsigned_arg */);
}
void Item_user_var_as_out_param::set_value(const char *str, uint length,
CHARSET_INFO* cs)
{
::update_hash(entry, FALSE, (void*)str, length, STRING_RESULT, cs,
DERIVATION_IMPLICIT, 0 /* unsigned_arg */);
}
double Item_user_var_as_out_param::val_real()
{
DBUG_ASSERT(0);
return 0.0;
}
longlong Item_user_var_as_out_param::val_int()
{
DBUG_ASSERT(0);
return 0;
}
String* Item_user_var_as_out_param::val_str(String *str)
{
DBUG_ASSERT(0);
return 0;
}
my_decimal* Item_user_var_as_out_param::val_decimal(my_decimal *decimal_buffer)
{
DBUG_ASSERT(0);
return 0;
}
void Item_user_var_as_out_param::print(String *str, enum_query_type query_type)
{
str->append('@');
str->append(name.str,name.length);
}
Item_func_get_system_var::
Item_func_get_system_var(sys_var *var_arg, enum_var_type var_type_arg,
LEX_STRING *component_arg, const char *name_arg,
size_t name_len_arg)
:var(var_arg), var_type(var_type_arg), orig_var_type(var_type_arg),
component(*component_arg), cache_present(0)
{
/* set_name() will allocate the name */
set_name(name_arg, (uint) name_len_arg, system_charset_info);
}
bool Item_func_get_system_var::is_written_to_binlog()
{
return var->is_written_to_binlog(var_type);
}
void Item_func_get_system_var::update_null_value()
{
THD *thd= current_thd;
int save_no_errors= thd->no_errors;
thd->no_errors= TRUE;
Item::update_null_value();
thd->no_errors= save_no_errors;
}
void Item_func_get_system_var::fix_length_and_dec()
{
char *cptr;
maybe_null= TRUE;
max_length= 0;
if (var->check_type(var_type))
{
if (var_type != OPT_DEFAULT)
{
my_error(ER_INCORRECT_GLOBAL_LOCAL_VAR, MYF(0),
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
var->name.str, var_type == OPT_GLOBAL ? "SESSION" : "GLOBAL");
return;
}
/* As there was no local variable, return the global value */
var_type= OPT_GLOBAL;
}
switch (var->show_type())
{
case SHOW_LONG:
case SHOW_INT:
case SHOW_HA_ROWS:
unsigned_flag= TRUE;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
collation.set_numeric();
fix_char_length(MY_INT64_NUM_DECIMAL_DIGITS);
decimals=0;
break;
case SHOW_LONGLONG:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
unsigned_flag= TRUE;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
collation.set_numeric();
fix_char_length(MY_INT64_NUM_DECIMAL_DIGITS);
decimals=0;
break;
case SHOW_CHAR:
case SHOW_CHAR_PTR:
mysql_mutex_lock(&LOCK_global_system_variables);
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
cptr= var->show_type() == SHOW_CHAR ?
(char*) var->value_ptr(current_thd, var_type, &component) :
*(char**) var->value_ptr(current_thd, var_type, &component);
if (cptr)
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
max_length= system_charset_info->cset->numchars(system_charset_info,
cptr,
cptr + strlen(cptr));
mysql_mutex_unlock(&LOCK_global_system_variables);
collation.set(system_charset_info, DERIVATION_SYSCONST);
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
max_length*= system_charset_info->mbmaxlen;
decimals=NOT_FIXED_DEC;
break;
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
case SHOW_LEX_STRING:
{
mysql_mutex_lock(&LOCK_global_system_variables);
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
LEX_STRING *ls= ((LEX_STRING*)var->value_ptr(current_thd, var_type, &component));
max_length= system_charset_info->cset->numchars(system_charset_info,
ls->str,
ls->str + ls->length);
mysql_mutex_unlock(&LOCK_global_system_variables);
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
collation.set(system_charset_info, DERIVATION_SYSCONST);
max_length*= system_charset_info->mbmaxlen;
decimals=NOT_FIXED_DEC;
}
break;
case SHOW_BOOL:
case SHOW_MY_BOOL:
unsigned_flag= FALSE;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
collation.set_numeric();
fix_char_length(1);
decimals=0;
break;
case SHOW_DOUBLE:
unsigned_flag= FALSE;
decimals= 6;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
collation.set_numeric();
fix_char_length(DBL_DIG + 6);
break;
default:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
my_error(ER_VAR_CANT_BE_READ, MYF(0), var->name.str);
break;
}
}
2005-07-19 18:25:05 +02:00
void Item_func_get_system_var::print(String *str, enum_query_type query_type)
{
str->append(name, name_length);
}
enum Item_result Item_func_get_system_var::result_type() const
{
switch (var->show_type())
{
case SHOW_BOOL:
case SHOW_MY_BOOL:
case SHOW_INT:
case SHOW_LONG:
case SHOW_LONGLONG:
case SHOW_HA_ROWS:
return INT_RESULT;
case SHOW_CHAR:
case SHOW_CHAR_PTR:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
case SHOW_LEX_STRING:
return STRING_RESULT;
case SHOW_DOUBLE:
return REAL_RESULT;
default:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
my_error(ER_VAR_CANT_BE_READ, MYF(0), var->name.str);
return STRING_RESULT; // keep the compiler happy
}
}
enum_field_types Item_func_get_system_var::field_type() const
{
switch (var->show_type())
{
case SHOW_BOOL:
case SHOW_MY_BOOL:
case SHOW_INT:
case SHOW_LONG:
case SHOW_LONGLONG:
case SHOW_HA_ROWS:
return MYSQL_TYPE_LONGLONG;
case SHOW_CHAR:
case SHOW_CHAR_PTR:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
case SHOW_LEX_STRING:
return MYSQL_TYPE_VARCHAR;
case SHOW_DOUBLE:
return MYSQL_TYPE_DOUBLE;
default:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
my_error(ER_VAR_CANT_BE_READ, MYF(0), var->name.str);
return MYSQL_TYPE_VARCHAR; // keep the compiler happy
}
}
/*
Uses var, var_type, component, cache_present, used_query_id, thd,
cached_llval, null_value, cached_null_value
*/
#define get_sys_var_safe(type) \
do { \
type value; \
mysql_mutex_lock(&LOCK_global_system_variables); \
value= *(type*) var->value_ptr(thd, var_type, &component); \
mysql_mutex_unlock(&LOCK_global_system_variables); \
cache_present |= GET_SYS_VAR_CACHE_LONG; \
used_query_id= thd->query_id; \
cached_llval= null_value ? 0 : (longlong) value; \
cached_null_value= null_value; \
return cached_llval; \
} while (0)
longlong Item_func_get_system_var::val_int()
{
THD *thd= current_thd;
if (cache_present && thd->query_id == used_query_id)
{
if (cache_present & GET_SYS_VAR_CACHE_LONG)
{
null_value= cached_null_value;
return cached_llval;
}
else if (cache_present & GET_SYS_VAR_CACHE_DOUBLE)
{
null_value= cached_null_value;
cached_llval= (longlong) cached_dval;
cache_present|= GET_SYS_VAR_CACHE_LONG;
return cached_llval;
}
else if (cache_present & GET_SYS_VAR_CACHE_STRING)
{
null_value= cached_null_value;
if (!null_value)
cached_llval= longlong_from_string_with_check (cached_strval.charset(),
cached_strval.c_ptr(),
cached_strval.c_ptr() +
cached_strval.length());
else
cached_llval= 0;
cache_present|= GET_SYS_VAR_CACHE_LONG;
return cached_llval;
}
}
switch (var->show_type())
{
case SHOW_INT: get_sys_var_safe (uint);
case SHOW_LONG: get_sys_var_safe (ulong);
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
case SHOW_LONGLONG: get_sys_var_safe (ulonglong);
case SHOW_HA_ROWS: get_sys_var_safe (ha_rows);
case SHOW_BOOL: get_sys_var_safe (bool);
case SHOW_MY_BOOL: get_sys_var_safe (my_bool);
case SHOW_DOUBLE:
{
double dval= val_real();
used_query_id= thd->query_id;
cached_llval= (longlong) dval;
cache_present|= GET_SYS_VAR_CACHE_LONG;
return cached_llval;
}
case SHOW_CHAR:
case SHOW_CHAR_PTR:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
case SHOW_LEX_STRING:
{
String *str_val= val_str(NULL);
if (str_val && str_val->length())
cached_llval= longlong_from_string_with_check (system_charset_info,
str_val->c_ptr(),
str_val->c_ptr() +
str_val->length());
else
{
null_value= TRUE;
cached_llval= 0;
}
cache_present|= GET_SYS_VAR_CACHE_LONG;
return cached_llval;
}
default:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
my_error(ER_VAR_CANT_BE_READ, MYF(0), var->name.str);
return 0; // keep the compiler happy
}
}
String* Item_func_get_system_var::val_str(String* str)
{
THD *thd= current_thd;
if (cache_present && thd->query_id == used_query_id)
{
if (cache_present & GET_SYS_VAR_CACHE_STRING)
{
null_value= cached_null_value;
return null_value ? NULL : &cached_strval;
}
else if (cache_present & GET_SYS_VAR_CACHE_LONG)
{
null_value= cached_null_value;
if (!null_value)
cached_strval.set (cached_llval, collation.collation);
cache_present|= GET_SYS_VAR_CACHE_STRING;
return null_value ? NULL : &cached_strval;
}
else if (cache_present & GET_SYS_VAR_CACHE_DOUBLE)
{
null_value= cached_null_value;
if (!null_value)
cached_strval.set_real (cached_dval, decimals, collation.collation);
cache_present|= GET_SYS_VAR_CACHE_STRING;
return null_value ? NULL : &cached_strval;
}
}
str= &cached_strval;
switch (var->show_type())
{
case SHOW_CHAR:
case SHOW_CHAR_PTR:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
case SHOW_LEX_STRING:
{
mysql_mutex_lock(&LOCK_global_system_variables);
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
char *cptr= var->show_type() == SHOW_CHAR ?
(char*) var->value_ptr(thd, var_type, &component) :
*(char**) var->value_ptr(thd, var_type, &component);
if (cptr)
{
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
size_t len= var->show_type() == SHOW_LEX_STRING ?
((LEX_STRING*)(var->value_ptr(thd, var_type, &component)))->length :
strlen(cptr);
if (str->copy(cptr, len, collation.collation))
{
null_value= TRUE;
str= NULL;
}
}
else
{
null_value= TRUE;
str= NULL;
}
mysql_mutex_unlock(&LOCK_global_system_variables);
break;
}
case SHOW_INT:
case SHOW_LONG:
case SHOW_LONGLONG:
case SHOW_HA_ROWS:
case SHOW_BOOL:
case SHOW_MY_BOOL:
str->set (val_int(), collation.collation);
break;
case SHOW_DOUBLE:
str->set_real (val_real(), decimals, collation.collation);
break;
default:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
my_error(ER_VAR_CANT_BE_READ, MYF(0), var->name.str);
str= NULL;
break;
}
cache_present|= GET_SYS_VAR_CACHE_STRING;
used_query_id= thd->query_id;
cached_null_value= null_value;
return str;
}
double Item_func_get_system_var::val_real()
{
THD *thd= current_thd;
if (cache_present && thd->query_id == used_query_id)
{
if (cache_present & GET_SYS_VAR_CACHE_DOUBLE)
{
null_value= cached_null_value;
return cached_dval;
}
else if (cache_present & GET_SYS_VAR_CACHE_LONG)
{
null_value= cached_null_value;
cached_dval= (double)cached_llval;
cache_present|= GET_SYS_VAR_CACHE_DOUBLE;
return cached_dval;
}
else if (cache_present & GET_SYS_VAR_CACHE_STRING)
{
null_value= cached_null_value;
if (!null_value)
cached_dval= double_from_string_with_check (cached_strval.charset(),
cached_strval.c_ptr(),
cached_strval.c_ptr() +
cached_strval.length());
else
cached_dval= 0;
cache_present|= GET_SYS_VAR_CACHE_DOUBLE;
return cached_dval;
}
}
switch (var->show_type())
{
case SHOW_DOUBLE:
mysql_mutex_lock(&LOCK_global_system_variables);
cached_dval= *(double*) var->value_ptr(thd, var_type, &component);
mysql_mutex_unlock(&LOCK_global_system_variables);
used_query_id= thd->query_id;
cached_null_value= null_value;
if (null_value)
cached_dval= 0;
cache_present|= GET_SYS_VAR_CACHE_DOUBLE;
return cached_dval;
case SHOW_CHAR:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
case SHOW_LEX_STRING:
case SHOW_CHAR_PTR:
{
mysql_mutex_lock(&LOCK_global_system_variables);
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
char *cptr= var->show_type() == SHOW_CHAR ?
(char*) var->value_ptr(thd, var_type, &component) :
*(char**) var->value_ptr(thd, var_type, &component);
if (cptr)
cached_dval= double_from_string_with_check (system_charset_info,
cptr, cptr + strlen (cptr));
else
{
null_value= TRUE;
cached_dval= 0;
}
mysql_mutex_unlock(&LOCK_global_system_variables);
used_query_id= thd->query_id;
cached_null_value= null_value;
cache_present|= GET_SYS_VAR_CACHE_DOUBLE;
return cached_dval;
}
case SHOW_INT:
case SHOW_LONG:
case SHOW_LONGLONG:
case SHOW_HA_ROWS:
case SHOW_BOOL:
case SHOW_MY_BOOL:
cached_dval= (double) val_int();
cache_present|= GET_SYS_VAR_CACHE_DOUBLE;
used_query_id= thd->query_id;
cached_null_value= null_value;
return cached_dval;
default:
WL#4738 streamline/simplify @@variable creation process Bug#16565 mysqld --help --verbose does not order variablesBug#20413 sql_slave_skip_counter is not shown in show variables Bug#20415 Output of mysqld --help --verbose is incomplete Bug#25430 variable not found in SELECT @@global.ft_max_word_len; Bug#32902 plugin variables don't know their names Bug#34599 MySQLD Option and Variable Reference need to be consistent in formatting! Bug#34829 No default value for variable and setting default does not raise error Bug#34834 ? Is accepted as a valid sql mode Bug#34878 Few variables have default value according to documentation but error occurs Bug#34883 ft_boolean_syntax cant be assigned from user variable to global var. Bug#37187 `INFORMATION_SCHEMA`.`GLOBAL_VARIABLES`: inconsistent status Bug#40988 log_output_basic.test succeeded though syntactically false. Bug#41010 enum-style command-line options are not honoured (maria.maria-recover fails) Bug#42103 Setting key_buffer_size to a negative value may lead to very large allocations Bug#44691 Some plugins configured as MYSQL_PLUGIN_MANDATORY in can be disabled Bug#44797 plugins w/o command-line options have no disabling option in --help Bug#46314 string system variables don't support expressions Bug#46470 sys_vars.max_binlog_cache_size_basic_32 is broken Bug#46586 When using the plugin interface the type "set" for options caused a crash. Bug#47212 Crash in DBUG_PRINT in mysqltest.cc when trying to print octal number Bug#48758 mysqltest crashes on sys_vars.collation_server_basic in gcov builds Bug#49417 some complaints about mysqld --help --verbose output Bug#49540 DEFAULT value of binlog_format isn't the default value Bug#49640 ambiguous option '--skip-skip-myisam' (double skip prefix) Bug#49644 init_connect and \0 Bug#49645 init_slave and multi-byte characters Bug#49646 mysql --show-warnings crashes when server dies
2009-12-22 10:35:56 +01:00
my_error(ER_VAR_CANT_BE_READ, MYF(0), var->name.str);
return 0;
}
}
bool Item_func_get_system_var::eq(const Item *item, bool binary_cmp) const
{
/* Assume we don't have rtti */
if (this == item)
return 1; // Same item is same.
/* Check if other type is also a get_user_var() object */
if (item->type() != FUNC_ITEM ||
((Item_func*) item)->functype() != functype())
return 0;
Item_func_get_system_var *other=(Item_func_get_system_var*) item;
return (var == other->var && var_type == other->var_type);
}
void Item_func_get_system_var::cleanup()
{
Item_func::cleanup();
cache_present= 0;
var_type= orig_var_type;
cached_strval.free();
}
2000-07-31 21:29:14 +02:00
longlong Item_func_inet_aton::val_int()
{
DBUG_ASSERT(fixed == 1);
2000-07-31 21:29:14 +02:00
uint byte_result = 0;
ulonglong result = 0; // We are ready for 64 bit addresses
const char *p,* end;
char c = '.'; // we mark c to indicate invalid IP in case length is 0
char buff[36];
int dot_count= 0;
2000-07-31 21:29:14 +02:00
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
String *s, tmp(buff, sizeof(buff), &my_charset_latin1);
if (!(s = args[0]->val_str_ascii(&tmp))) // If null value
2000-07-31 21:29:14 +02:00
goto err;
null_value=0;
end= (p = s->ptr()) + s->length();
while (p < end)
{
c = *p++;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
int digit = (int) (c - '0');
2000-07-31 21:29:14 +02:00
if (digit >= 0 && digit <= 9)
{
if ((byte_result = byte_result * 10 + digit) > 255)
goto err; // Wrong address
}
else if (c == '.')
{
dot_count++;
2000-07-31 21:29:14 +02:00
result= (result << 8) + (ulonglong) byte_result;
byte_result = 0;
}
else
goto err; // Invalid character
}
if (c != '.') // IP number can't end on '.'
{
/*
Handle short-forms addresses according to standard. Examples:
127 -> 0.0.0.127
127.1 -> 127.0.0.1
127.2.1 -> 127.2.0.1
*/
2004-02-05 15:58:10 +01:00
switch (dot_count) {
case 1: result<<= 8; /* Fall through */
case 2: result<<= 8; /* Fall through */
}
2000-07-31 21:29:14 +02:00
return (result << 8) + (ulonglong) byte_result;
}
2000-07-31 21:29:14 +02:00
err:
null_value=1;
return 0;
}
2000-11-04 15:48:06 +01:00
void Item_func_match::init_search(bool no_order)
{
DBUG_ENTER("Item_func_match::init_search");
/* Check if init_search() has been called before */
if (ft_handler)
{
/*
We should reset ft_handler as it is cleaned up
on destruction of FT_SELECT object
(necessary in case of re-execution of subquery).
TODO: FT_SELECT should not clean up ft_handler.
*/
if (join_key)
table->file->ft_handler= ft_handler;
DBUG_VOID_RETURN;
}
if (key == NO_SUCH_KEY)
{
List<Item> fields;
fields.push_back(new Item_string(" ",1, cmp_collation.collation));
for (uint i=1; i < arg_count; i++)
fields.push_back(args[i]);
concat_ws=new Item_func_concat_ws(fields);
2004-03-20 12:36:26 +01:00
/*
Above function used only to get value and do not need fix_fields for it:
Item_string - basic constant
2004-03-25 21:11:22 +01:00
fields - fix_fields() was already called for this arguments
Item_func_concat_ws - do not need fix_fields() to produce value
2004-03-20 12:36:26 +01:00
*/
concat_ws->quick_fix_field();
}
if (master)
{
2000-11-04 15:48:06 +01:00
join_key=master->join_key=join_key|master->join_key;
master->init_search(no_order);
ft_handler=master->ft_handler;
join_key=master->join_key;
DBUG_VOID_RETURN;
}
String *ft_tmp= 0;
// MATCH ... AGAINST (NULL) is meaningless, but possible
if (!(ft_tmp=key_item()->val_str(&value)))
{
ft_tmp= &value;
value.set("",0,cmp_collation.collation);
}
if (ft_tmp->charset() != cmp_collation.collation)
{
uint dummy_errors;
search_value.copy(ft_tmp->ptr(), ft_tmp->length(), ft_tmp->charset(),
cmp_collation.collation, &dummy_errors);
ft_tmp= &search_value;
}
if (join_key && !no_order)
flags|=FT_SORTED;
ft_handler=table->file->ft_init_ext(flags, key, ft_tmp);
if (join_key)
table->file->ft_handler=ft_handler;
DBUG_VOID_RETURN;
}
bool Item_func_match::fix_fields(THD *thd, Item **ref)
2000-07-31 21:29:14 +02:00
{
DBUG_ASSERT(fixed == 0);
Item *UNINIT_VAR(item); // Safe as arg_count is > 1
2000-07-31 21:29:14 +02:00
maybe_null=1;
join_key=0;
/*
const_item is assumed in quite a bit of places, so it would be difficult
to remove; If it would ever to be removed, this should include
modifications to find_best and auto_close as complement to auto_init code
above.
*/
if (Item_func::fix_fields(thd, ref) ||
!args[0]->const_during_execution())
{
my_error(ER_WRONG_ARGUMENTS,MYF(0),"AGAINST");
return TRUE;
}
2000-07-31 21:29:14 +02:00
const_item_cache=0;
for (uint i=1 ; i < arg_count ; i++)
2000-07-31 21:29:14 +02:00
{
item=args[i];
2001-04-17 14:17:22 +02:00
if (item->type() == Item::REF_ITEM)
args[i]= item= *((Item_ref *)item)->ref;
if (item->type() != Item::FIELD_ITEM)
{
my_error(ER_WRONG_ARGUMENTS, MYF(0), "AGAINST");
return TRUE;
}
2000-07-31 21:29:14 +02:00
}
/*
Check that all columns come from the same table.
We've already checked that columns in MATCH are fields so
PARAM_TABLE_BIT can only appear from AGAINST argument.
*/
if ((used_tables_cache & ~PARAM_TABLE_BIT) != item->used_tables())
key=NO_SUCH_KEY;
if (key == NO_SUCH_KEY && !(flags & FT_BOOL))
{
my_error(ER_WRONG_ARGUMENTS,MYF(0),"MATCH");
return TRUE;
}
table=((Item_field *)item)->field->table;
This changeset is largely a handler cleanup changeset (WL#3281), but includes fixes and cleanups that was found necessary while testing the handler changes Changes that requires code changes in other code of other storage engines. (Note that all changes are very straightforward and one should find all issues by compiling a --debug build and fixing all compiler errors and all asserts in field.cc while running the test suite), - New optional handler function introduced: reset() This is called after every DML statement to make it easy for a handler to statement specific cleanups. (The only case it's not called is if force the file to be closed) - handler::extra(HA_EXTRA_RESET) is removed. Code that was there before should be moved to handler::reset() - table->read_set contains a bitmap over all columns that are needed in the query. read_row() and similar functions only needs to read these columns - table->write_set contains a bitmap over all columns that will be updated in the query. write_row() and update_row() only needs to update these columns. The above bitmaps should now be up to date in all context (including ALTER TABLE, filesort()). The handler is informed of any changes to the bitmap after fix_fields() by calling the virtual function handler::column_bitmaps_signal(). If the handler does caching of these bitmaps (instead of using table->read_set, table->write_set), it should redo the caching in this code. as the signal() may be sent several times, it's probably best to set a variable in the signal and redo the caching on read_row() / write_row() if the variable was set. - Removed the read_set and write_set bitmap objects from the handler class - Removed all column bit handling functions from the handler class. (Now one instead uses the normal bitmap functions in my_bitmap.c instead of handler dedicated bitmap functions) - field->query_id is removed. One should instead instead check table->read_set and table->write_set if a field is used in the query. - handler::extra(HA_EXTRA_RETRIVE_ALL_COLS) and handler::extra(HA_EXTRA_RETRIEVE_PRIMARY_KEY) are removed. One should now instead use table->read_set to check for which columns to retrieve. - If a handler needs to call Field->val() or Field->store() on columns that are not used in the query, one should install a temporary all-columns-used map while doing so. For this, we provide the following functions: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->read_set); field->val(); dbug_tmp_restore_column_map(table->read_set, old_map); and similar for the write map: my_bitmap_map *old_map= dbug_tmp_use_all_columns(table, table->write_set); field->val(); dbug_tmp_restore_column_map(table->write_set, old_map); If this is not done, you will sooner or later hit a DBUG_ASSERT in the field store() / val() functions. (For not DBUG binaries, the dbug_tmp_restore_column_map() and dbug_tmp_restore_column_map() are inline dummy functions and should be optimized away be the compiler). - If one needs to temporary set the column map for all binaries (and not just to avoid the DBUG_ASSERT() in the Field::store() / Field::val() methods) one should use the functions tmp_use_all_columns() and tmp_restore_column_map() instead of the above dbug_ variants. - All 'status' fields in the handler base class (like records, data_file_length etc) are now stored in a 'stats' struct. This makes it easier to know what status variables are provided by the base handler. This requires some trivial variable names in the extra() function. - New virtual function handler::records(). This is called to optimize COUNT(*) if (handler::table_flags() & HA_HAS_RECORDS()) is true. (stats.records is not supposed to be an exact value. It's only has to be 'reasonable enough' for the optimizer to be able to choose a good optimization path). - Non virtual handler::init() function added for caching of virtual constants from engine. - Removed has_transactions() virtual method. Now one should instead return HA_NO_TRANSACTIONS in table_flags() if the table handler DOES NOT support transactions. - The 'xxxx_create_handler()' function now has a MEM_ROOT_root argument that is to be used with 'new handler_name()' to allocate the handler in the right area. The xxxx_create_handler() function is also responsible for any initialization of the object before returning. For example, one should change: static handler *myisam_create_handler(TABLE_SHARE *table) { return new ha_myisam(table); } -> static handler *myisam_create_handler(TABLE_SHARE *table, MEM_ROOT *mem_root) { return new (mem_root) ha_myisam(table); } - New optional virtual function: use_hidden_primary_key(). This is called in case of an update/delete when (table_flags() and HA_PRIMARY_KEY_REQUIRED_FOR_DELETE) is defined but we don't have a primary key. This allows the handler to take precisions in remembering any hidden primary key to able to update/delete any found row. The default handler marks all columns to be read. - handler::table_flags() now returns a ulonglong (to allow for more flags). - New/changed table_flags() - HA_HAS_RECORDS Set if ::records() is supported - HA_NO_TRANSACTIONS Set if engine doesn't support transactions - HA_PRIMARY_KEY_REQUIRED_FOR_DELETE Set if we should mark all primary key columns for read when reading rows as part of a DELETE statement. If there is no primary key, all columns are marked for read. - HA_PARTIAL_COLUMN_READ Set if engine will not read all columns in some cases (based on table->read_set) - HA_PRIMARY_KEY_ALLOW_RANDOM_ACCESS Renamed to HA_PRIMARY_KEY_REQUIRED_FOR_POSITION. - HA_DUPP_POS Renamed to HA_DUPLICATE_POS - HA_REQUIRES_KEY_COLUMNS_FOR_DELETE Set this if we should mark ALL key columns for read when when reading rows as part of a DELETE statement. In case of an update we will mark all keys for read for which key part changed value. - HA_STATS_RECORDS_IS_EXACT Set this if stats.records is exact. (This saves us some extra records() calls when optimizing COUNT(*)) - Removed table_flags() - HA_NOT_EXACT_COUNT Now one should instead use HA_HAS_RECORDS if handler::records() gives an exact count() and HA_STATS_RECORDS_IS_EXACT if stats.records is exact. - HA_READ_RND_SAME Removed (no one supported this one) - Removed not needed functions ha_retrieve_all_cols() and ha_retrieve_all_pk() - Renamed handler::dupp_pos to handler::dup_pos - Removed not used variable handler::sortkey Upper level handler changes: - ha_reset() now does some overall checks and calls ::reset() - ha_table_flags() added. This is a cached version of table_flags(). The cache is updated on engine creation time and updated on open. MySQL level changes (not obvious from the above): - DBUG_ASSERT() added to check that column usage matches what is set in the column usage bit maps. (This found a LOT of bugs in current column marking code). - In 5.1 before, all used columns was marked in read_set and only updated columns was marked in write_set. Now we only mark columns for which we need a value in read_set. - Column bitmaps are created in open_binary_frm() and open_table_from_share(). (Before this was in table.cc) - handler::table_flags() calls are replaced with handler::ha_table_flags() - For calling field->val() you must have the corresponding bit set in table->read_set. For calling field->store() you must have the corresponding bit set in table->write_set. (There are asserts in all store()/val() functions to catch wrong usage) - thd->set_query_id is renamed to thd->mark_used_columns and instead of setting this to an integer value, this has now the values: MARK_COLUMNS_NONE, MARK_COLUMNS_READ, MARK_COLUMNS_WRITE Changed also all variables named 'set_query_id' to mark_used_columns. - In filesort() we now inform the handler of exactly which columns are needed doing the sort and choosing the rows. - The TABLE_SHARE object has a 'all_set' column bitmap one can use when one needs a column bitmap with all columns set. (This is used for table->use_all_columns() and other places) - The TABLE object has 3 column bitmaps: - def_read_set Default bitmap for columns to be read - def_write_set Default bitmap for columns to be written - tmp_set Can be used as a temporary bitmap when needed. The table object has also two pointer to bitmaps read_set and write_set that the handler should use to find out which columns are used in which way. - count() optimization now calls handler::records() instead of using handler->stats.records (if (table_flags() & HA_HAS_RECORDS) is true). - Added extra argument to Item::walk() to indicate if we should also traverse sub queries. - Added TABLE parameter to cp_buffer_from_ref() - Don't close tables created with CREATE ... SELECT but keep them in the table cache. (Faster usage of newly created tables). New interfaces: - table->clear_column_bitmaps() to initialize the bitmaps for tables at start of new statements. - table->column_bitmaps_set() to set up new column bitmaps and signal the handler about this. - table->column_bitmaps_set_no_signal() for some few cases where we need to setup new column bitmaps but don't signal the handler (as the handler has already been signaled about these before). Used for the momement only in opt_range.cc when doing ROR scans. - table->use_all_columns() to install a bitmap where all columns are marked as use in the read and the write set. - table->default_column_bitmaps() to install the normal read and write column bitmaps, but not signaling the handler about this. This is mainly used when creating TABLE instances. - table->mark_columns_needed_for_delete(), table->mark_columns_needed_for_delete() and table->mark_columns_needed_for_insert() to allow us to put additional columns in column usage maps if handler so requires. (The handler indicates what it neads in handler->table_flags()) - table->prepare_for_position() to allow us to tell handler that it needs to read primary key parts to be able to store them in future table->position() calls. (This replaces the table->file->ha_retrieve_all_pk function) - table->mark_auto_increment_column() to tell handler are going to update columns part of any auto_increment key. - table->mark_columns_used_by_index() to mark all columns that is part of an index. It will also send extra(HA_EXTRA_KEYREAD) to handler to allow it to quickly know that it only needs to read colums that are part of the key. (The handler can also use the column map for detecting this, but simpler/faster handler can just monitor the extra() call). - table->mark_columns_used_by_index_no_reset() to in addition to other columns, also mark all columns that is used by the given key. - table->restore_column_maps_after_mark_index() to restore to default column maps after a call to table->mark_columns_used_by_index(). - New item function register_field_in_read_map(), for marking used columns in table->read_map. Used by filesort() to mark all used columns - Maintain in TABLE->merge_keys set of all keys that are used in query. (Simplices some optimization loops) - Maintain Field->part_of_key_not_clustered which is like Field->part_of_key but the field in the clustered key is not assumed to be part of all index. (used in opt_range.cc for faster loops) - dbug_tmp_use_all_columns(), dbug_tmp_restore_column_map() tmp_use_all_columns() and tmp_restore_column_map() functions to temporally mark all columns as usable. The 'dbug_' version is primarily intended inside a handler when it wants to just call Field:store() & Field::val() functions, but don't need the column maps set for any other usage. (ie:: bitmap_is_set() is never called) - We can't use compare_records() to skip updates for handlers that returns a partial column set and the read_set doesn't cover all columns in the write set. The reason for this is that if we have a column marked only for write we can't in the MySQL level know if the value changed or not. The reason this worked before was that MySQL marked all to be written columns as also to be read. The new 'optimal' bitmaps exposed this 'hidden bug'. - open_table_from_share() does not anymore setup temporary MEM_ROOT object as a thread specific variable for the handler. Instead we send the to-be-used MEMROOT to get_new_handler(). (Simpler, faster code) Bugs fixed: - Column marking was not done correctly in a lot of cases. (ALTER TABLE, when using triggers, auto_increment fields etc) (Could potentially result in wrong values inserted in table handlers relying on that the old column maps or field->set_query_id was correct) Especially when it comes to triggers, there may be cases where the old code would cause lost/wrong values for NDB and/or InnoDB tables. - Split thd->options flag OPTION_STATUS_NO_TRANS_UPDATE to two flags: OPTION_STATUS_NO_TRANS_UPDATE and OPTION_KEEP_LOG. This allowed me to remove some wrong warnings about: "Some non-transactional changed tables couldn't be rolled back" - Fixed handling of INSERT .. SELECT and CREATE ... SELECT that wrongly reset (thd->options & OPTION_STATUS_NO_TRANS_UPDATE) which caused us to loose some warnings about "Some non-transactional changed tables couldn't be rolled back") - Fixed use of uninitialized memory in ha_ndbcluster.cc::delete_table() which could cause delete_table to report random failures. - Fixed core dumps for some tests when running with --debug - Added missing FN_LIBCHAR in mysql_rm_tmp_tables() (This has probably caused us to not properly remove temporary files after crash) - slow_logs was not properly initialized, which could maybe cause extra/lost entries in slow log. - If we get an duplicate row on insert, change column map to read and write all columns while retrying the operation. This is required by the definition of REPLACE and also ensures that fields that are only part of UPDATE are properly handled. This fixed a bug in NDB and REPLACE where REPLACE wrongly copied some column values from the replaced row. - For table handler that doesn't support NULL in keys, we would give an error when creating a primary key with NULL fields, even after the fields has been automaticly converted to NOT NULL. - Creating a primary key on a SPATIAL key, would fail if field was not declared as NOT NULL. Cleanups: - Removed not used condition argument to setup_tables - Removed not needed item function reset_query_id_processor(). - Field->add_index is removed. Now this is instead maintained in (field->flags & FIELD_IN_ADD_INDEX) - Field->fieldnr is removed (use field->field_index instead) - New argument to filesort() to indicate that it should return a set of row pointers (not used columns). This allowed me to remove some references to sql_command in filesort and should also enable us to return column results in some cases where we couldn't before. - Changed column bitmap handling in opt_range.cc to be aligned with TABLE bitmap, which allowed me to use bitmap functions instead of looping over all fields to create some needed bitmaps. (Faster and smaller code) - Broke up found too long lines - Moved some variable declaration at start of function for better code readability. - Removed some not used arguments from functions. (setup_fields(), mysql_prepare_insert_check_table()) - setup_fields() now takes an enum instead of an int for marking columns usage. - For internal temporary tables, use handler::write_row(), handler::delete_row() and handler::update_row() instead of handler::ha_xxxx() for faster execution. - Changed some constants to enum's and define's. - Using separate column read and write sets allows for easier checking of timestamp field was set by statement. - Remove calls to free_io_cache() as this is now done automaticly in ha_reset() - Don't build table->normalized_path as this is now identical to table->path (after bar's fixes to convert filenames) - Fixed some missed DBUG_PRINT(.."%lx") to use "0x%lx" to make it easier to do comparision with the 'convert-dbug-for-diff' tool. Things left to do in 5.1: - We wrongly log failed CREATE TABLE ... SELECT in some cases when using row based logging (as shown by testcase binlog_row_mix_innodb_myisam.result) Mats has promised to look into this. - Test that my fix for CREATE TABLE ... SELECT is indeed correct. (I added several test cases for this, but in this case it's better that someone else also tests this throughly). Lars has promosed to do this.
2006-06-04 17:52:22 +02:00
if (!(table->file->ha_table_flags() & HA_CAN_FULLTEXT))
{
2005-05-14 15:24:36 +02:00
my_error(ER_TABLE_CANT_HANDLE_FT, MYF(0));
2000-07-31 21:29:14 +02:00
return 1;
}
table->fulltext_searched=1;
WL#2649 Number-to-string conversions added: include/ctype_numconv.inc mysql-test/include/ctype_numconv.inc mysql-test/r/ctype_binary.result mysql-test/t/ctype_binary.test Adding tests modified: mysql-test/r/bigint.result mysql-test/r/case.result mysql-test/r/create.result mysql-test/r/ctype_cp1251.result mysql-test/r/ctype_latin1.result mysql-test/r/ctype_ucs.result mysql-test/r/func_gconcat.result mysql-test/r/func_str.result mysql-test/r/metadata.result mysql-test/r/ps_1general.result mysql-test/r/ps_2myisam.result mysql-test/r/ps_3innodb.result mysql-test/r/ps_4heap.result mysql-test/r/ps_5merge.result mysql-test/r/show_check.result mysql-test/r/type_datetime.result mysql-test/r/type_ranges.result mysql-test/r/union.result mysql-test/suite/ndb/r/ps_7ndb.result mysql-test/t/ctype_cp1251.test mysql-test/t/ctype_latin1.test mysql-test/t/ctype_ucs.test mysql-test/t/func_str.test Fixing tests @ sql/field.cc - Return str result using my_charset_numeric. - Using real multi-byte aware str_to_XXX functions to handle tricky charset values propely (e.g. UCS2) @ sql/field.h - Changing derivation of non-string field types to DERIVATION_NUMERIC. - Changing binary() for numeric/datetime fields to always return TRUE even if charset is not my_charset_bin. We need this to keep ha_base_keytype() return HA_KEYTYPE_BINARY. - Adding BINARY_FLAG into some fields, because it's not being set automatically anymore with "my_charset_bin to my_charset_numeric" change. - Changing derivation for numeric/datetime datatypes to a weaker value, to make "SELECT concat('string', field)" use character set of the string literal for the result of the function. @ sql/item.cc - Implementing generic val_str_ascii(). - Using max_char_length() instead of direct read of max_length to make "tricky" charsets like UCS2 work. NOTE: in the future we'll possibly remove all direct reads of max_length - Fixing Item_num::safe_charset_converter(). Previously it alligned binary string to character string (for example by adding leading 0x00 when doing binary->UCS2 conversion). Now it just converts from my_charset_numbner to "tocs". - Using val_str_ascii() in Item::get_time() to make UCS2 arguments work. - Other misc changes @ sql/item.h - Changing MY_COLL_CMP_CONV and MY_COLL_ALLOW_CONV to bit operations instead of hard-coded bit masks. - Addding new method DTCollation.set_numeric(). - Adding new methods to Item. - Adding helper functions to make code look nicer: agg_item_charsets_for_string_result() agg_item_charsets_for_comparison() - Changing charset for Item_num-derived items from my_charset_bin to my_charset_numeric (which is an alias for latin1). @ sql/item_cmpfunc.cc - Using new helper functions - Other misc changes @ sql/item_cmpfunc.h - Fixing strcmp() to return max_length=2. Previously it returned 1, which was wrong, because it did not fit '-1'. @ sql/item_func.cc - Using new helper functions - Other minor changes @ sql/item_func.h - Removing unused functions - Adding helper functions agg_arg_charsets_for_string_result() agg_arg_charsets_for_comparison() - Adding set_numeric() into constructors of numeric items. - Using fix_length_and_charset() and fix_char_length() instead of direct write to max_length. @ sql/item_geofunc.cc - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when character_set_connection=ucs2). @ sql/item_geofunc.h - Changing class for Item_func_geometry_type and Item_func_as_wkt from Item_str_func to Item_str_ascii_func, to make them return UCS2 result properly (when @@character_set_connection=ucs2). @ sql/item_strfunc.cc - Implementing Item_str_func::val_str(). - Renaming val_str to val_str_ascii for some items, to make them work with UCS2 properly. - Using new helper functions - All single-argument functions that expect string result now call this method: agg_arg_charsets_for_string_result(collation, args, 1); This enables character set conversion to @@character_set_connection in case of pure numeric input. @ sql/item_strfunc.h - Introducing Item_str_ascii_func - for functions which return pure ASCII data, for performance purposes, as well as for the cases when the old implementation of val_str() was heavily 8-bit oriented and implementing a UCS2-aware version is tricky. @ sql/item_sum.cc - Using new helper functions. @ sql/item_timefunc.cc - Using my_charset_numeric instead of my_charset_bin. - Using fix_char_length(), fix_length_and_charset() and fix_length_and_charset_datetime() instead of direct write to max_length. - Using tricky-charset aware function str_to_time_with_warn() @ sql/item_timefunc.h - Using new helper functions for charset and length initialization. - Changing base class for Item_func_get_format() to make it return UCS2 properly (when character_set_connection=ucs2). @ sql/item_xmlfunc.cc - Using new helper function @ sql/my_decimal.cc - Adding a new DECIMAL to CHAR converter with real multibyte support (e.g. UCS2) @ sql/mysql_priv.h - Introducing a new derivation level for numeric/datetime data types. - Adding macros for my_charset_numeric and MY_REPERTOIRE_NUMERIC. - Adding prototypes for str_set_decimal() - Adding prototypes for character-set aware str_to_xxx() functions. @ sql/protocol.cc - Changing charsetnr to "binary" client-side metadata for numeric/datetime data types. @ sql/time.cc - Adding to_ascii() helper function, to convert a string in any character set to ascii representation. In the future can be extended to understand digits written in various non-Latin word scripts. - Adding real multy-byte character set aware versions for str_to_XXXX, to make these these type of queries work correct: INSERT INTO t1 SET datetime_column=ucs2_expression; @ strings/ctype-ucs2.c - endptr was not calculated correctly. INSERTing of UCS2 values into numeric columns returned warnings about truncated wrong data.
2010-02-11 05:17:25 +01:00
return agg_item_collations_for_comparison(cmp_collation, func_name(),
args+1, arg_count-1, 0);
}
2000-07-31 21:29:14 +02:00
bool Item_func_match::fix_index()
{
Item_field *item;
uint ft_to_key[MAX_KEY], ft_cnt[MAX_KEY], fts=0, keynr;
uint max_cnt=0, mkeys=0, i;
if (key == NO_SUCH_KEY)
return 0;
if (!table)
goto err;
2000-07-31 21:29:14 +02:00
for (keynr=0 ; keynr < table->s->keys ; keynr++)
2000-07-31 21:29:14 +02:00
{
if ((table->key_info[keynr].flags & HA_FULLTEXT) &&
(flags & FT_BOOL ? table->keys_in_use_for_query.is_set(keynr) :
table->s->keys_in_use.is_set(keynr)))
2000-07-31 21:29:14 +02:00
{
ft_to_key[fts]=keynr;
2000-07-31 21:29:14 +02:00
ft_cnt[fts]=0;
fts++;
}
}
if (!fts)
goto err;
2000-07-31 21:29:14 +02:00
for (i=1; i < arg_count; i++)
2000-07-31 21:29:14 +02:00
{
item=(Item_field*)args[i];
for (keynr=0 ; keynr < fts ; keynr++)
2000-07-31 21:29:14 +02:00
{
KEY *ft_key=&table->key_info[ft_to_key[keynr]];
2000-07-31 21:29:14 +02:00
uint key_parts=ft_key->key_parts;
for (uint part=0 ; part < key_parts ; part++)
{
if (item->field->eq(ft_key->key_part[part].field))
ft_cnt[keynr]++;
2000-07-31 21:29:14 +02:00
}
}
}
for (keynr=0 ; keynr < fts ; keynr++)
2000-07-31 21:29:14 +02:00
{
if (ft_cnt[keynr] > max_cnt)
2000-07-31 21:29:14 +02:00
{
mkeys=0;
max_cnt=ft_cnt[mkeys]=ft_cnt[keynr];
ft_to_key[mkeys]=ft_to_key[keynr];
continue;
}
if (max_cnt && ft_cnt[keynr] == max_cnt)
{
mkeys++;
ft_cnt[mkeys]=ft_cnt[keynr];
ft_to_key[mkeys]=ft_to_key[keynr];
continue;
2000-07-31 21:29:14 +02:00
}
}
for (keynr=0 ; keynr <= mkeys ; keynr++)
2000-07-31 21:29:14 +02:00
{
// partial keys doesn't work
if (max_cnt < arg_count-1 ||
max_cnt < table->key_info[ft_to_key[keynr]].key_parts)
continue;
2000-07-31 21:29:14 +02:00
key=ft_to_key[keynr];
return 0;
}
err:
if (flags & FT_BOOL)
{
key=NO_SUCH_KEY;
return 0;
}
2004-11-12 13:34:00 +01:00
my_message(ER_FT_MATCHING_KEY_NOT_FOUND,
ER(ER_FT_MATCHING_KEY_NOT_FOUND), MYF(0));
return 1;
}
bool Item_func_match::eq(const Item *item, bool binary_cmp) const
{
if (item->type() != FUNC_ITEM ||
((Item_func*)item)->functype() != FT_FUNC ||
flags != ((Item_func_match*)item)->flags)
return 0;
Item_func_match *ifm=(Item_func_match*) item;
if (key == ifm->key && table == ifm->table &&
key_item()->eq(ifm->key_item(), binary_cmp))
return 1;
2000-07-31 21:29:14 +02:00
return 0;
}
2004-11-11 19:39:35 +01:00
double Item_func_match::val_real()
2001-10-09 14:53:54 +02:00
{
DBUG_ASSERT(fixed == 1);
DBUG_ENTER("Item_func_match::val");
2001-12-02 13:34:01 +01:00
if (ft_handler == NULL)
DBUG_RETURN(-1.0);
2001-10-09 14:53:54 +02:00
if (key != NO_SUCH_KEY && table->null_row) /* NULL row from an outer join */
DBUG_RETURN(0.0);
2003-06-16 23:05:45 +02:00
2001-10-09 14:53:54 +02:00
if (join_key)
{
if (table->file->ft_handler)
DBUG_RETURN(ft_handler->please->get_relevance(ft_handler));
2001-10-09 14:53:54 +02:00
join_key=0;
}
if (key == NO_SUCH_KEY)
2001-10-09 14:53:54 +02:00
{
String *a= concat_ws->val_str(&value);
if ((null_value= (a == 0)) || !a->length())
DBUG_RETURN(0);
DBUG_RETURN(ft_handler->please->find_relevance(ft_handler,
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
(uchar *)a->ptr(), a->length()));
2001-10-09 14:53:54 +02:00
}
DBUG_RETURN(ft_handler->please->find_relevance(ft_handler,
table->record[0], 0));
2001-10-09 14:53:54 +02:00
}
void Item_func_match::print(String *str, enum_query_type query_type)
{
str->append(STRING_WITH_LEN("(match "));
print_args(str, 1, query_type);
str->append(STRING_WITH_LEN(" against ("));
args[0]->print(str, query_type);
2003-11-03 11:28:36 +01:00
if (flags & FT_BOOL)
str->append(STRING_WITH_LEN(" in boolean mode"));
2003-11-03 11:28:36 +01:00
else if (flags & FT_EXPAND)
str->append(STRING_WITH_LEN(" with query expansion"));
str->append(STRING_WITH_LEN("))"));
}
longlong Item_func_bit_xor::val_int()
{
DBUG_ASSERT(fixed == 1);
ulonglong arg1= (ulonglong) args[0]->val_int();
ulonglong arg2= (ulonglong) args[1]->val_int();
if ((null_value= (args[0]->null_value || args[1]->null_value)))
return 0;
return (longlong) (arg1 ^ arg2);
}
/***************************************************************************
System variables
****************************************************************************/
2007-10-11 19:29:09 +02:00
/**
Return value of an system variable base[.name] as a constant item.
2007-10-11 19:29:09 +02:00
@param thd Thread handler
@param var_type global / session
@param name Name of base or system variable
@param component Component.
2007-10-11 19:29:09 +02:00
@note
If component.str = 0 then the variable name is in 'name'
2007-10-11 19:29:09 +02:00
@return
- 0 : error
- # : constant item
*/
Item *get_system_var(THD *thd, enum_var_type var_type, LEX_STRING name,
LEX_STRING component)
{
sys_var *var;
LEX_STRING *base_name, *component_name;
if (component.str)
{
base_name= &component;
component_name= &name;
}
else
{
base_name= &name;
component_name= &component; // Empty string
}
if (!(var= find_sys_var(thd, base_name->str, base_name->length)))
return 0;
if (component.str)
{
if (!var->is_struct())
{
my_error(ER_VARIABLE_IS_NOT_STRUCT, MYF(0), base_name->str);
return 0;
}
}
2003-11-19 15:19:46 +01:00
thd->lex->uncacheable(UNCACHEABLE_SIDEEFFECT);
set_if_smaller(component_name->length, MAX_SYS_VAR_LENGTH);
return new Item_func_get_system_var(var, var_type, component_name,
NULL, 0);
}
2007-10-11 19:29:09 +02:00
/**
Check a user level lock.
2007-10-11 19:29:09 +02:00
Sets null_value=TRUE on error.
2007-10-11 19:29:09 +02:00
@retval
1 Available
2007-10-11 19:29:09 +02:00
@retval
0 Already taken, or error
*/
longlong Item_func_is_free_lock::val_int()
{
DBUG_ASSERT(fixed == 1);
String *res=args[0]->val_str(&value);
User_level_lock *ull;
null_value=0;
if (!res || !res->length())
{
null_value=1;
return 0;
}
mysql_mutex_lock(&LOCK_user_locks);
ull= (User_level_lock *) my_hash_search(&hash_user_locks, (uchar*) res->ptr(),
(size_t) res->length());
mysql_mutex_unlock(&LOCK_user_locks);
if (!ull || !ull->locked)
return 1;
return 0;
}
2002-02-22 12:24:42 +01:00
longlong Item_func_is_used_lock::val_int()
{
DBUG_ASSERT(fixed == 1);
String *res=args[0]->val_str(&value);
User_level_lock *ull;
null_value=1;
if (!res || !res->length())
return 0;
mysql_mutex_lock(&LOCK_user_locks);
ull= (User_level_lock *) my_hash_search(&hash_user_locks, (uchar*) res->ptr(),
(size_t) res->length());
mysql_mutex_unlock(&LOCK_user_locks);
if (!ull || !ull->locked)
return 0;
null_value=0;
return ull->thread_id;
}
longlong Item_func_row_count::val_int()
{
DBUG_ASSERT(fixed == 1);
THD *thd= current_thd;
return thd->get_row_count_func();
}
Item_func_sp::Item_func_sp(Name_resolution_context *context_arg, sp_name *name)
:Item_func(), context(context_arg), m_name(name), m_sp(NULL), sp_result_field(NULL)
{
maybe_null= 1;
m_name->init_qname(current_thd);
dummy_table= (TABLE*) sql_calloc(sizeof(TABLE)+ sizeof(TABLE_SHARE));
dummy_table->s= (TABLE_SHARE*) (dummy_table+1);
}
Item_func_sp::Item_func_sp(Name_resolution_context *context_arg,
sp_name *name, List<Item> &list)
:Item_func(list), context(context_arg), m_name(name), m_sp(NULL),sp_result_field(NULL)
{
maybe_null= 1;
m_name->init_qname(current_thd);
dummy_table= (TABLE*) sql_calloc(sizeof(TABLE)+ sizeof(TABLE_SHARE));
dummy_table->s= (TABLE_SHARE*) (dummy_table+1);
}
void
Item_func_sp::cleanup()
{
if (sp_result_field)
{
delete sp_result_field;
sp_result_field= NULL;
}
m_sp= NULL;
dummy_table->alias= NULL;
Item_func::cleanup();
}
const char *
Item_func_sp::func_name() const
{
THD *thd= current_thd;
2005-02-08 23:50:45 +01:00
/* Calculate length to avoid reallocation of string for sure */
uint len= (((m_name->m_explicit_name ? m_name->m_db.length : 0) +
m_name->m_name.length)*2 + //characters*quoting
2 + // ` and `
(m_name->m_explicit_name ?
3 : 0) + // '`', '`' and '.' for the db
1 + // end of string
ALIGN_SIZE(1)); // to avoid String reallocation
String qname((char *)alloc_root(thd->mem_root, len), len,
system_charset_info);
qname.length(0);
if (m_name->m_explicit_name)
{
append_identifier(thd, &qname, m_name->m_db.str, m_name->m_db.length);
qname.append('.');
}
append_identifier(thd, &qname, m_name->m_name.str, m_name->m_name.length);
return qname.ptr();
}
void my_missing_function_error(const LEX_STRING &token, const char *func_name)
{
if (token.length && is_lex_native_function (&token))
my_error(ER_FUNC_INEXISTENT_NAME_COLLISION, MYF(0), func_name);
else
my_error(ER_SP_DOES_NOT_EXIST, MYF(0), "FUNCTION", func_name);
}
/**
@brief Initialize the result field by creating a temporary dummy table
and assign it to a newly created field object. Meta data used to
create the field is fetched from the sp_head belonging to the stored
proceedure found in the stored procedure functon cache.
@note This function should be called from fix_fields to init the result
field. It is some what related to Item_field.
@see Item_field
@param thd A pointer to the session and thread context.
@return Function return error status.
@retval TRUE is returned on an error
@retval FALSE is returned on success.
*/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
bool
Item_func_sp::init_result_field(THD *thd)
{
2007-03-27 19:57:33 +02:00
LEX_STRING empty_name= { C_STRING_WITH_LEN("") };
TABLE_SHARE *share;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
DBUG_ENTER("Item_func_sp::init_result_field");
DBUG_ASSERT(m_sp == NULL);
DBUG_ASSERT(sp_result_field == NULL);
if (!(m_sp= sp_find_routine(thd, TYPE_ENUM_FUNCTION, m_name,
&thd->sp_func_cache, TRUE)))
2005-03-09 22:53:04 +01:00
{
my_missing_function_error (m_name->m_name, m_name->m_qname.str);
context->process_error(thd);
DBUG_RETURN(TRUE);
2005-03-09 22:53:04 +01:00
}
/*
A Field need to be attached to a Table.
Below we "create" a dummy table by initializing
the needed pointers.
*/
2007-03-19 16:45:09 +01:00
share= dummy_table->s;
dummy_table->alias = "";
dummy_table->maybe_null = maybe_null;
dummy_table->in_use= thd;
dummy_table->copy_blobs= TRUE;
share->table_cache_key = empty_name;
share->table_name = empty_name;
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
if (!(sp_result_field= m_sp->create_result_field(max_length, name,
dummy_table)))
{
DBUG_RETURN(TRUE);
}
if (sp_result_field->pack_length() > sizeof(result_buf))
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
void *tmp;
if (!(tmp= sql_alloc(sp_result_field->pack_length())))
DBUG_RETURN(TRUE);
sp_result_field->move_field((uchar*) tmp);
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
else
sp_result_field->move_field(result_buf);
sp_result_field->null_ptr= (uchar *) &null_value;
sp_result_field->null_bit= 1;
DBUG_RETURN(FALSE);
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
/**
@brief Initialize local members with values from the Field interface.
@note called from Item::fix_fields.
*/
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
void Item_func_sp::fix_length_and_dec()
{
DBUG_ENTER("Item_func_sp::fix_length_and_dec");
DBUG_ASSERT(sp_result_field);
decimals= sp_result_field->decimals();
max_length= sp_result_field->field_length;
collation.set(sp_result_field->charset());
maybe_null= 1;
unsigned_flag= test(sp_result_field->flags & UNSIGNED_FLAG);
DBUG_VOID_RETURN;
}
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
/**
@brief Execute function & store value in field.
@return Function returns error status.
@retval FALSE on success.
@retval TRUE if an error occurred.
*/
bool
Item_func_sp::execute()
{
THD *thd= current_thd;
/* Execute function and store the return value in the field. */
if (execute_impl(thd))
{
null_value= 1;
context->process_error(thd);
if (thd->killed)
thd->send_kill_message();
return TRUE;
}
/* Check that the field (the value) is not NULL. */
null_value= sp_result_field->is_null();
return null_value;
}
/**
@brief Execute function and store the return value in the field.
@note This function was intended to be the concrete implementation of
the interface function execute. This was never realized.
@return The error state.
@retval FALSE on success
@retval TRUE if an error occurred.
*/
bool
Item_func_sp::execute_impl(THD *thd)
{
bool err_status= TRUE;
Sub_statement_state statement_state;
#ifndef NO_EMBEDDED_ACCESS_CHECKS
Security_context *save_security_ctx= thd->security_ctx;
#endif
enum enum_sp_data_access access=
(m_sp->m_chistics->daccess == SP_DEFAULT_ACCESS) ?
SP_DEFAULT_ACCESS_MAPPING : m_sp->m_chistics->daccess;
DBUG_ENTER("Item_func_sp::execute_impl");
#ifndef NO_EMBEDDED_ACCESS_CHECKS
if (context->security_ctx)
{
/* Set view definer security context */
thd->security_ctx= context->security_ctx;
}
#endif
if (sp_check_access(thd))
goto error;
/*
Throw an error if a non-deterministic function is called while
statement-based replication (SBR) is active.
*/
if (!m_sp->m_chistics->detistic && !trust_function_creators &&
(access == SP_CONTAINS_SQL || access == SP_MODIFIES_SQL_DATA) &&
(mysql_bin_log.is_open() &&
thd->variables.binlog_format == BINLOG_FORMAT_STMT))
{
my_error(ER_BINLOG_UNSAFE_ROUTINE, MYF(0));
goto error;
}
/*
Disable the binlogging if this is not a SELECT statement. If this is a
SELECT, leave binlogging on, so execute_function() code writes the
function call into binlog.
*/
thd->reset_sub_statement_state(&statement_state, SUB_STMT_FUNCTION);
err_status= m_sp->execute_function(thd, args, arg_count, sp_result_field);
thd->restore_sub_statement_state(&statement_state);
2005-09-21 07:29:47 +02:00
error:
#ifndef NO_EMBEDDED_ACCESS_CHECKS
2005-10-31 21:14:27 +01:00
thd->security_ctx= save_security_ctx;
#endif
DBUG_RETURN(err_status);
}
void
Item_func_sp::make_field(Send_field *tmp_field)
{
DBUG_ENTER("Item_func_sp::make_field");
DBUG_ASSERT(sp_result_field);
sp_result_field->make_field(tmp_field);
if (name)
tmp_field->col_name= name;
DBUG_VOID_RETURN;
}
enum enum_field_types
Item_func_sp::field_type() const
{
DBUG_ENTER("Item_func_sp::field_type");
DBUG_ASSERT(sp_result_field);
DBUG_RETURN(sp_result_field->type());
}
Item_result
Item_func_sp::result_type() const
{
DBUG_ENTER("Item_func_sp::result_type");
DBUG_PRINT("info", ("m_sp = %p", (void *) m_sp));
DBUG_ASSERT(sp_result_field);
DBUG_RETURN(sp_result_field->result_type());
}
longlong Item_func_found_rows::val_int()
{
DBUG_ASSERT(fixed == 1);
return current_thd->found_rows();
}
Field *
Item_func_sp::tmp_table_field(TABLE *t_arg)
{
DBUG_ENTER("Item_func_sp::tmp_table_field");
DBUG_ASSERT(sp_result_field);
DBUG_RETURN(sp_result_field);
}
2005-09-21 07:29:47 +02:00
/**
@brief Checks if requested access to function can be granted to user.
If function isn't found yet, it searches function first.
If function can't be found or user don't have requested access
2005-09-21 07:29:47 +02:00
error is raised.
@param thd thread handler
@return Indication if the access was granted or not.
@retval FALSE Access is granted.
@retval TRUE Requested access can't be granted or function doesn't exists.
*/
2005-09-21 07:29:47 +02:00
bool
Item_func_sp::sp_check_access(THD *thd)
{
DBUG_ENTER("Item_func_sp::sp_check_access");
DBUG_ASSERT(m_sp);
2005-09-21 07:29:47 +02:00
#ifndef NO_EMBEDDED_ACCESS_CHECKS
if (check_routine_access(thd, EXECUTE_ACL,
2005-09-21 07:29:47 +02:00
m_sp->m_db.str, m_sp->m_name.str, 0, FALSE))
DBUG_RETURN(TRUE);
#endif
DBUG_RETURN(FALSE);
}
bool
Item_func_sp::fix_fields(THD *thd, Item **ref)
{
bool res;
DBUG_ENTER("Item_func_sp::fix_fields");
DBUG_ASSERT(fixed == 0);
/*
We must call init_result_field before Item_func::fix_fields()
to make m_sp and result_field members available to fix_length_and_dec(),
which is called from Item_func::fix_fields().
*/
res= init_result_field(thd);
if (res)
DBUG_RETURN(res);
res= Item_func::fix_fields(thd, ref);
if (res)
DBUG_RETURN(res);
Fixed following problems: --Bug#52157 various crashes and assertions with multi-table update, stored function --Bug#54475 improper error handling causes cascading crashing failures in innodb/ndb --Bug#57703 create view cause Assertion failed: 0, file .\item_subselect.cc, line 846 --Bug#57352 valgrind warnings when creating view --Recently discovered problem when a nested materialized derived table is used before being populated and it leads to incorrect result We have several modes when we should disable subquery evaluation. The reasons for disabling are different. It could be uselessness of the evaluation as in case of 'CREATE VIEW' or 'PREPARE stmt', or we should disable subquery evaluation if tables are not locked yet as it happens in bug#54475, or too early evaluation of subqueries can lead to wrong result as it happened in Bug#19077. Main problem is that if subquery items are treated as const they are evaluated in ::fix_fields(), ::fix_length_and_dec() of the parental items as a lot of these methods have Item::val_...() calls inside. We have to make subqueries non-const to prevent unnecessary subquery evaluation. At the moment we have different methods for this. Here is a list of these modes: 1. PREPARE stmt; We use UNCACHEABLE_PREPARE flag. It is set during parsing in sql_parse.cc, mysql_new_select() for each SELECT_LEX object and cleared at the end of PREPARE in sql_prepare.cc, init_stmt_after_parse(). If this flag is set subquery becomes non-const and evaluation does not happen. 2. CREATE|ALTER VIEW, SHOW CREATE VIEW, I_S tables which process FRM files We use LEX::view_prepare_mode field. We set it before view preparation and check this flag in ::fix_fields(), ::fix_length_and_dec(). Some bugs are fixed using this approach, some are not(Bug#57352, Bug#57703). The problem here is that we have a lot of ::fix_fields(), ::fix_length_and_dec() where we use Item::val_...() calls for const items. 3. Derived tables with subquery = wrong result(Bug19077) The reason of this bug is too early subquery evaluation. It was fixed by adding Item::with_subselect field The check of this field in appropriate places prevents const item evaluation if the item have subquery. The fix for Bug19077 fixes only the problem with convert_constant_item() function and does not cover other places(::fix_fields(), ::fix_length_and_dec() again) where subqueries could be evaluated. Example: CREATE TABLE t1 (i INT, j BIGINT); INSERT INTO t1 VALUES (1, 2), (2, 2), (3, 2); SELECT * FROM (SELECT MIN(i) FROM t1 WHERE j = SUBSTRING('12', (SELECT * FROM (SELECT MIN(j) FROM t1) t2))) t3; DROP TABLE t1; 4. Derived tables with subquery where subquery is evaluated before table locking(Bug#54475, Bug#52157) Suggested solution is following: -Introduce new field LEX::context_analysis_only with the following possible flags: #define CONTEXT_ANALYSIS_ONLY_PREPARE 1 #define CONTEXT_ANALYSIS_ONLY_VIEW 2 #define CONTEXT_ANALYSIS_ONLY_DERIVED 4 -Set/clean these flags when we perform context analysis operation -Item_subselect::const_item() returns result depending on LEX::context_analysis_only. If context_analysis_only is set then we return FALSE that means that subquery is non-const. As all subquery types are wrapped by Item_subselect it allow as to make subquery non-const when it's necessary.
2010-12-14 10:33:03 +01:00
if (thd->lex->context_analysis_only & CONTEXT_ANALYSIS_ONLY_VIEW)
{
/*
Here we check privileges of the stored routine only during view
creation, in order to validate the view. A runtime check is
perfomed in Item_func_sp::execute(), and this method is not
called during context analysis. Notice, that during view
creation we do not infer into stored routine bodies and do not
check privileges of its statements, which would probably be a
good idea especially if the view has SQL SECURITY DEFINER and
the used stored procedure has SQL SECURITY DEFINER.
*/
res= sp_check_access(thd);
#ifndef NO_EMBEDDED_ACCESS_CHECKS
/*
Try to set and restore the security context to see whether it's valid
*/
Security_context *save_secutiry_ctx;
res= set_routine_security_ctx(thd, m_sp, false, &save_secutiry_ctx);
if (!res)
m_sp->m_security_ctx.restore_security_context(thd, save_secutiry_ctx);
#endif /* ! NO_EMBEDDED_ACCESS_CHECKS */
}
if (!m_sp->m_chistics->detistic)
{
used_tables_cache |= RAND_TABLE_BIT;
const_item_cache= FALSE;
}
DBUG_RETURN(res);
}
void Item_func_sp::update_used_tables()
{
Item_func::update_used_tables();
if (!m_sp->m_chistics->detistic)
{
used_tables_cache |= RAND_TABLE_BIT;
const_item_cache= FALSE;
}
}
/*
uuid_short handling.
The short uuid is defined as a longlong that contains the following bytes:
Bytes Comment
1 Server_id & 255
4 Startup time of server in seconds
3 Incrementor
This means that an uuid is guaranteed to be unique
even in a replication environment if the following holds:
- The last byte of the server id is unique
- If you between two shutdown of the server don't get more than
an average of 2^24 = 16M calls to uuid_short() per second.
*/
ulonglong uuid_value;
void uuid_short_init()
{
WL#3817: Simplify string / memory area types and make things more consistent (first part) The following type conversions was done: - Changed byte to uchar - Changed gptr to uchar* - Change my_string to char * - Change my_size_t to size_t - Change size_s to size_t Removed declaration of byte, gptr, my_string, my_size_t and size_s. Following function parameter changes was done: - All string functions in mysys/strings was changed to use size_t instead of uint for string lengths. - All read()/write() functions changed to use size_t (including vio). - All protocoll functions changed to use size_t instead of uint - Functions that used a pointer to a string length was changed to use size_t* - Changed malloc(), free() and related functions from using gptr to use void * as this requires fewer casts in the code and is more in line with how the standard functions work. - Added extra length argument to dirname_part() to return the length of the created string. - Changed (at least) following functions to take uchar* as argument: - db_dump() - my_net_write() - net_write_command() - net_store_data() - DBUG_DUMP() - decimal2bin() & bin2decimal() - Changed my_compress() and my_uncompress() to use size_t. Changed one argument to my_uncompress() from a pointer to a value as we only return one value (makes function easier to use). - Changed type of 'pack_data' argument to packfrm() to avoid casts. - Changed in readfrm() and writefrom(), ha_discover and handler::discover() the type for argument 'frmdata' to uchar** to avoid casts. - Changed most Field functions to use uchar* instead of char* (reduced a lot of casts). - Changed field->val_xxx(xxx, new_ptr) to take const pointers. Other changes: - Removed a lot of not needed casts - Added a few new cast required by other changes - Added some cast to my_multi_malloc() arguments for safety (as string lengths needs to be uint, not size_t). - Fixed all calls to hash-get-key functions to use size_t*. (Needed to be done explicitely as this conflict was often hided by casting the function to hash_get_key). - Changed some buffers to memory regions to uchar* to avoid casts. - Changed some string lengths from uint to size_t. - Changed field->ptr to be uchar* instead of char*. This allowed us to get rid of a lot of casts. - Some changes from true -> TRUE, false -> FALSE, unsigned char -> uchar - Include zlib.h in some files as we needed declaration of crc32() - Changed MY_FILE_ERROR to be (size_t) -1. - Changed many variables to hold the result of my_read() / my_write() to be size_t. This was needed to properly detect errors (which are returned as (size_t) -1). - Removed some very old VMS code - Changed packfrm()/unpackfrm() to not be depending on uint size (portability fix) - Removed windows specific code to restore cursor position as this causes slowdown on windows and we should not mix read() and pread() calls anyway as this is not thread safe. Updated function comment to reflect this. Changed function that depended on original behavior of my_pwrite() to itself restore the cursor position (one such case). - Added some missing checking of return value of malloc(). - Changed definition of MOD_PAD_CHAR_TO_FULL_LENGTH to avoid 'long' overflow. - Changed type of table_def::m_size from my_size_t to ulong to reflect that m_size is the number of elements in the array, not a string/memory length. - Moved THD::max_row_length() to table.cc (as it's not depending on THD). Inlined max_row_length_blob() into this function. - More function comments - Fixed some compiler warnings when compiled without partitions. - Removed setting of LEX_STRING() arguments in declaration (portability fix). - Some trivial indentation/variable name changes. - Some trivial code simplifications: - Replaced some calls to alloc_root + memcpy to use strmake_root()/strdup_root(). - Changed some calls from memdup() to strmake() (Safety fix) - Simpler loops in client-simple.c
2007-05-10 11:59:39 +02:00
uuid_value= ((((ulonglong) server_id) << 56) +
(((ulonglong) server_start_time) << 24));
}
longlong Item_func_uuid_short::val_int()
{
ulonglong val;
mysql_mutex_lock(&LOCK_uuid_generator);
val= uuid_value++;
mysql_mutex_unlock(&LOCK_uuid_generator);
return (longlong) val;
}