The convert_constant_item() function converts constant items to ints on
prepare phase to optimize execution speed. In this case it tries to evaluate
subselect which contains a derived table and is contained in a derived table.
All derived tables are filled only after all derived tables are prepared.
So evaluation of subselect with derived table at the prepare phase will
return a wrong result.
A new flag with_subselect is added to the Item class. It indicates that
expression which this item represents is a subselect or contains a subselect.
It is set to 0 by default. It is set to 1 in the Item_subselect constructor
for subselects.
For Item_func and Item_cond derived classes it is set after fixing any argument
in Item_func::fix_fields() and Item_cond::fix_fields accordingly.
The convert_constant_item() function now doesn't convert a constant item
if the with_subselect flag set in it.
from within triggers
Add support for passing NEW.x as INOUT and OUT parameters to stored
procedures. Passing NEW.x as INOUT parameter requires SELECT and
UPDATE privileges on that column, and passing it as OUT parameter
requires only UPDATE privilege.
In the code that converts IN predicates to EXISTS predicates it is changing
the select list elements to constant 1. Example :
SELECT ... FROM ... WHERE a IN (SELECT c FROM ...)
is transformed to :
SELECT ... FROM ... WHERE EXISTS (SELECT 1 FROM ... HAVING a = c)
However there can be no FROM clause in the IN subquery and it may not be
a simple select : SELECT ... FROM ... WHERE a IN (SELECT f(..) AS
c UNION SELECT ...) This query is transformed to : SELECT ... FROM ...
WHERE EXISTS (SELECT 1 FROM (SELECT f(..) AS c UNION SELECT ...)
x HAVING a = c) In the above query c in the HAVING clause is made to be
an Item_null_helper (a subclass of Item_ref) pointing to the real
Item_field (which is not referenced anywhere else in the query anymore).
This is done because Item_ref_null_helper collects information whether
there are NULL values in the result. This is OK for directly executed
statements, because the Item_field pointed by the Item_null_helper is
already fixed when the transformation is done. But when executed as
a prepared statement all the Item instances are "un-fixed" before the
recompilation of the prepared statement. So when the Item_null_helper
gets fixed it discovers that the Item_field it points to is not fixed
and issues an error. The remedy is to keep the original select list
references when there are no tables in the FROM clause. So the above
becomes : SELECT ... FROM ... WHERE EXISTS (SELECT c FROM (SELECT f(..)
AS c UNION SELECT ...) x HAVING a = c) In this way c is referenced
directly in the select list as well as by reference in the HAVING
clause. So it gets correctly fixed even with prepared statements. And
since the Item_null_helper subclass of Item_ref_null_helper is not used
anywhere else it's taken out.
too much memory. Instead, either create the equvalent SEL_TREE manually, or create only two ranges that
strictly include the area to scan
(Note: just to re-iterate: increasing NOT_IN_IGNORE_THRESHOLD will make optimization run slower for big
IN-lists, but the server will not run out of memory. O(N^2) memory use has been eliminated)
CONNECTION_ID() was implemented as a constant Item, i.e. an instance of
Item_static_int_func class holding value computed at creation time.
Since Items are created on parsing, and trigger statements are parsed
on table open, the first connection to open a particular table would
effectively set its own CONNECTION_ID() inside trigger statements for
that table.
Re-implement CONNECTION_ID() as a class derived from Item_int_func, and
compute connection_id on every call to fix_fields().
The bug was due to a missed case in the detection of whether an index
can be used for loose scan. More precisely, the range optimizer chose
to use loose index scan for queries for which the condition(s) over
an index key part could not be pushed to the index together with the
loose scan.
As a result, loose index scan was selecting the first row in the
index with a given GROUP BY prefix, and was applying the WHERE
clause after that, while it should have inspected all rows with
the given prefix, and apply the WHERE clause to all of them.
The fix detects and skips such cases.
time per connection
Removed const_string() method from Item_string (it was only used in one
place, in a bad way). Defer possible SP variable, and access data directly
instead, in date_format item.
- Added empty constructors and virtual destructors to many classes and structs
- Removed some usage of the offsetof() macro to instead use C++ class pointers
If item->cached_table is set, find_field_in_tables() returns found field
even if it doesn't belong to current select. Because Item_field::fix_fields
doesn't expect such behaviour, reported bug occurs.
Item_field::fix_fields() was modifed to detect when find_field_in_tables()
can return field from outer select and process such fields accordingly.
In order to ease this code which was searching and processing outed fields was
moved into separate function called Item_field::fix_outer_field().
- BUG#15166: Wrong update permissions required to execute triggers
- BUG#15196: Wrong select permission required to execute triggers
The idea of the fix is to check necessary privileges
in Item_trigger_field::fix_fields(), instead of having "special variables"
technique. To achieve this, we should pass to an Item_trigger_field instance
a flag, which will indicate the usage/access type of this trigger variable.
- Fixed tests
- Optimized new code
- Fixed some unlikely core dumps
- Better bug fixes for:
- #14397 - OPTIMIZE TABLE with an open HANDLER causes a crash
- #14850 (ERROR 1062 when a quering a view using a Group By on a column that can be null
libmysqld/Makefile.am:
sql/Makefile.am:
Adding new source files.
Adding new file into build process.
include/my_xml.h:
strings/xml.c:
Adding new XML parse flags to skip text normalization and
to use relative tag names. Adding enum for XML token types.
sql/lex.h:
Making parser aware of new SQL functions.
sqll/item_create.h, sql/item_create.cc:
Adding creators for ExtractValue and UpdateXML.
sql/item.h:
Adding new Item types: nodeset and nodeset comparator.
sql/item_xmlfunc.h
sql/item_xmlfunc.cc
Adding new classes implementing XPath functions.
mysql-test/t/xml.test, mysql-test/r/xml.result:
New files: adding test case
to Crash": the bug was that due to non-standard name
resolution precedence in stored procedures (See Bug#5967)
a stored procedure variable took precedence over a table column
when the arguments for VALUES() function were resolved.
The implementation of VALUES() function was not designed to work
with Item_splocal and crashed.
VALUES() function is non-standard. It can refer to, and
is meaningful for, table columns only. The patch disables SP
variables as possible arguments of VALUES() function.
according to the standard.
The idea is to use Field-classes to implement stored routines
variables. Also, we should provide facade to Item-hierarchy
by Item_field class (it is necessary, since SRVs take part
in expressions).
The patch fixes the following bugs:
- BUG#8702: Stored Procedures: No Error/Warning shown for inappropriate data
type matching;
- BUG#8768: Functions: For any unsigned data type, -ve values can be passed
and returned;
- BUG#8769: Functions: For Int datatypes, out of range values can be passed
and returned;
- BUG#9078: STORED PROCDURE: Decimal digits are not displayed when we use
DECIMAL datatype;
- BUG#9572: Stored procedures: variable type declarations ignored;
- BUG#12903: upper function does not work inside a function;
- BUG#13705: parameters to stored procedures are not verified;
- BUG#13808: ENUM type stored procedure parameter accepts non-enumerated
data;
- BUG#13909: Varchar Stored Procedure Parameter always BINARY string (ignores
CHARACTER SET);
- BUG#14161: Stored procedure cannot retrieve bigint unsigned;
- BUG#14188: BINARY variables have no 0x00 padding;
- BUG#15148: Stored procedure variables accept non-scalar values;
The table opening process now works the following way:
- Create common TABLE_SHARE object
- Read the .frm file and unpack it into the TABLE_SHARE object
- Create a TABLE object based on the information in the TABLE_SHARE
object and open a handler to the table object
Other noteworthy changes:
- In TABLE_SHARE the most common strings are now LEX_STRING's
- Better error message when table is not found
- Variable table_cache is now renamed 'table_open_cache'
- New variable 'table_definition_cache' that is the number of table defintions that will be cached
- strxnmov() calls are now fixed to avoid overflows
- strxnmov() will now always add one end \0 to result
- engine objects are now created with a TABLE_SHARE object instead of a TABLE object.
- After creating a field object one must call field->init(table) before using it
- For a busy system this change will give you:
- Less memory usage for table object
- Faster opening of tables (if it's has been in use or is in table definition cache)
- Allow you to cache many table definitions objects
- Faster drop of table
if inner routine has more local variables than outer one, and
one of its last variables was used as argument to NOT operator".
THD::spcont was non-0 when we were parsing stored routine/trigger
definition during execution of another stored routine. This confused
methods of Item_splocal and forced them use wrong runtime context.
Fix ensures that we always have THD::spcont equal to zero during
routine/trigger body parsing. This also allows to avoid problems
with errors which occur during parsing and SQL exception handlers.
Bad examples of usage of a string with its length fixed.
The incorrect length in the trigger file configuration descriptor
fixed (BUG#14090).
A hook for unknown keys added to the parser to support old .TRG files.
Invalid date like 2000-02-32 wasn't converted to int, which lead to not
using index and comparison with field as astring, which results in slow
query execution.
convert_constatn_item() and get_mm_leaf() now forces MODE_INVALID_DATES to
allow such conversion.
- CHAR() now returns binary string as default
- CHAR(X*65536+Y*256+Z) is now equal to CHAR(X,Y,Z) independent of the character set for CHAR()
- Test for both ETIMEDOUT and ETIME from pthread_cond_timedwait()
(Some old systems returns ETIME and it's safer to test for both values
than to try to write a wrapper for each old system)
- Fixed new introduced bug in NOT BETWEEN X and X
- Ensure we call commit_by_xid or rollback_by_xid for all engines, even if one engine has failed
- Use octet2hex() for all conversion of string to hex
- Simplify and optimize code
Ensure that ccache is also used for C programs
mysql: Ensure that 'delimiter' works the same way in batch mode as in normal mode
mysqldump: Change to use ;; (instead of //) as a stored procedure/trigger delimiter
Fixed test cases by adding missing DROP's and rename views to be of type 'v#'
Removed MY_UNIX_PATH from fn_format()
Removed current_db_used from TABLE_LIST
Removed usage of 'current_thd' in Item_splocal
Removed some compiler warnings
A bit faster longlong2str code
The problem was then when a column reference was resolved to a view column, the new Item
created for this column contained the name of the view, and not the view alias.
Fields of view represented by Item_direct_view_ref. When complex expression
such as if(sum()>...,...) is splited in simpler parts by refs was ignored.
Beside this direct ref doesn't use it's result_field and thus can't store
it's result in tmp table which is needed for sum() ... group.
All this results in reported bug.
Item::split_sum_func2() now converts Item_direct_view_ref to Item_ref to
make fields from view being storable in tmp table.
"Interleaved SPs execution is now binlogged properly, "SELECT spfunc()" is binlogged too.
The known remaining issue is binlogging/replication of "a routine is deleted while it is executed" scenario.
- Corrected problem with N-way nested natural joins in PS mode.
- Code cleanup
- More asserts to check consistency of name resolution contexts
- Fixed potential memory leak of name resolution contexts
In cp932, '\' character can be the second byte in a
multi-byte character stream. This makes it difficult to use
mysql_escape_string. Added flag to indicate which languages allow
'\' as second byte of multibyte sequence so that when putting a prepared
statement into the binlog we can decide at runtime whether hex encoding
is really needed.
"Process NATURAL and USING joins according to SQL:2003".
* Some of the main problems fixed by the patch:
- in "select *" queries the * expanded correctly according to
ANSI for arbitrary natural/using joins
- natural/using joins are correctly transformed into JOIN ... ON
for any number/nesting of the joins.
- column references are correctly resolved against natural joins
of any nesting and combined with arbitrary other joins.
* This patch also contains a fix for name resolution of items
inside the ON condition of JOIN ... ON - in this case items must
be resolved only against the JOIN operands. To support such
'local' name resolution, the patch introduces a stack of
name resolution contexts used at parse time.
NOTICE:
- This patch is not complete in the sense that
- there are 2 test cases that still do not pass -
one in join.test, one in select.test. Both are marked
with a comment "TODO: WL#2486".
- it does not include a new test specific for the task
item.cc:
item.h:
Adding Item_param::safe_charset_converter,
not to return collation mix error if
parameter can be converted into operation
character set.
ctype_utf8.result:
adding test case
ctype_utf8.test:
adding test case
Change bool in C code to my_bool
Added to mysqltest --enable_parsning and --disable_parsing to avoid to have to comment parts of tests
Added comparison of LEX_STRING's and use this to compare file types for view and trigger files.
Adding test
item_sum.cc:
Adding a call for collation/charset aggregation,
to collect attributes from the arguments. The actual bug fix.
item_func.h, item_func.cc, item.h, item.cc:
- Removing collation aggrgation functions from Item_func class
in item.cc, and adding it as non-class functions in item.cc
to be able to reuse this code for group_concat.
- Adding replacement for these functions into Item_func class
as wrappers for moved functions, to minizize patch size,
Fixed bug #11412.
Reversed the patch of cs 1.1934 for the function
create_tmp_table. Modified the function to support
tem_ref objects created for view fields.
item_buff.cc:
Fixed bug #11412.
Modified implementation of new_Cached_item to support
cacheing of view fields.
item.h:
Fixed bug #11412.
Changed implementation of Item_ref::get_tmp_table_field and
added Item_ref::get_tmp_table_item to support Item_ref objects
created for view fields.
view.test, view.result:
Added a test case for bug #11412.
"Triggers have the wrong namespace"
"Triggers: duplicate names allowed"
"Triggers: CREATE TRIGGER does not accept fully qualified names"
"SHOW TRIGGERS"
data": remove the fix for another bug (8807) that
added OUTER_REF_TABLE_BIT to all subqueries that used a placeholder
to prevent their evaluation at prepare. As this bit hanged in
Item_subselect::used_tables_cache for ever, a constant subquery with
a placeholder was never evaluated as such, which caused wrong
choice of the execution plan for the statement.
- to fix Bug#8807 backport a better fix from 5.0
- post-review fixes.
adding test case
sql_table.cc:
sql_table.cc:
- do not create a new item when charsets are the same
- return ER_INVALID_DEFAULT if default value cannot
be converted into the column character set.
item.cc:
- Allow conversion not only to Unicode,
but also to and from "binary".
- Adding safe_charset_converter() for Item_num
and Item_varbinary, returning a fixed const Item.
When searching column to sort on, item was compared to field under view
column, but not the column itself. Because names of view column and underlaid
field may differ, it leads to possibly choosing wrong column for sorting on.
This patch makes Item_direct_view_ref::eq(Item *item,...) compare
item's name with it's own name proir to comparing to *ref item.
Added a test case for bug #11771.
item.h:
Fixed bug #11771.
Added method reset_query_id_processor to be able to adjust
query_id for fields generated from * in queries like this:
SELECT * FROM <view> ...
sql_base.cc:
Fixed bug #11771.
Adjusted query_id for fields generated from * in queries
like this: SELECT * FROM <view> ...
Fixed the results of a test for group_concat.
After the fix foor bug #11639 the results became
correct.
olap.result, olap.test:
Added a test case for bug #11639.
sql_select.cc:
Fixed bug #11639: a wrong result set when using a view
instead of the underlying table in a rollup query
executed through filesort.
The old code did not take into account that we always
use an Item_ref object when referring to a view column.
item.h:
Fixed bug #11639.
Now if two Item_ref items ref1 and ref2 refer to the same field
then ref1->eq(ref2) returns 1.
#9728 'Decreased functionality in "on duplicate key update
#8147 'a column proclaimed ambigous in INSERT ... SELECT .. ON DUPLICATE'
This ensures fields are uniquely qualified and also that one can't update other tables in the ON DUPLICATE KEY UPDATE part
Remove changes made by bug fix#8147. They strips list of insert_table_list to
only insert table, which results in error reported in bug #9728.
Added flag to Item to resolve ambigous fields reported in bug #8147.
when using ORDER BY
Insert were inserting data from last record fetched instead of inserting from
temporary table.
Make Item_ref to save data from result_field if it's available rather then
from *ref on save_in_field() call.
Wrong method for creating temporary field was choosen, which results in
sending int field with int header but lonlong data.
Test case is added to mysql_client_test.c because client library is required
to test the bug.
error for LIMIT placeholder".
The patch adds grammar support for LIMIT ?, ? and changes the
type of ST_SELECT_LEX::select_limit,offset_limit from ha_rows to Item*,
so that it can point to Item_param.
Fixed bug #11088: a crash for queries with GROUP BY a BLOB column
+ COUNT(DISTINCT...) due to an attempt to allocate a too large
buffer for the BLOB field.
Now the size of the buffer is limited by max_sort_length.
group_by.test, group_by.result:
Added a test case for bug #11088.
