mariadb/sql/sql_string.h
Alexander Barkov 87cebd8ce2 Cleanup#1 for MDEV-34319: DECLARE TYPE .. TABLE OF .. INDEX BY
- Checking that the key expression is compatible with the INDEX BY data type
  for assignment in expressions:
    assoc_array_variable(key_expr)
    assoc_array_variable(key_expr).field

  in all contexts: SELECT, assignment target, INTO target.
  Raising an error in case it's not compatible.

- Disallowing non-constant expressions as a key,
  as the key is evaluated during the fix_fields() time.

- Disallowing stored functions as a key:
    assoc_array(stored_function())
    assoc_array(stored_function()).field

  The underlying MariaDB code is not ready to call a stored function
  during the fix_fields() time. This will be fixed in a separate MDEV.

- Removing the move Assoc_array_data's constructor.
  Using the usual constructor instead.

- Setting m_key.thread_specific and m_value.thread_specific to true
  in the Assoc_array_data constructor. This is needed to get assoc array
  element data counted by the @@session.memory_used status variable.

  Adding DBUG_ASSERTs to make sure the thread_specific flag never
  disappears in Assoc_array_data members.

- Removing my_free(item) from Field_assoc_array::element_by_key.
  It was a remainder from an earlier patch version.
  In the current patch version all Items behind an assoc array are
  created on a mem_root. It's wrong to use my_free() with them.

- Adding a helper method Field_assoc_array::assoc_tree_search()

- Fixing assoc_array_var.delete() to work as a procedure
  rather than a function. It does not need SELECT/DO any more.

- Fixing the crash in a few ctype_xxx tests, caused by the grammar change.

- Fixing compilation failure on Windows

- Adding a new method LEX::set_field_type_udt_or_typedef()
  and removing duplicate code from sql_yacc.yy

- Renaming the grammar rule field_type_all_with_composites to
  field_type_all_with_typedefs

- Removing the grammar rule assoc_array_index_types.
  Changing the grammar to "INDEX_SYM BY field_type".

  Removing the grammar rule field_type_all_with_record.
  Allow field_type_all_with_typedefs as an assoc array element.

  Catching wrong index and element data types has been moved to
  Type_handler_assoc_array::Column_definition_set_attributes().
  It raises an SQL error on things like:
  * assoc array of assoc arrays in TABLE OF
  * index by a non-supported types in INDEX BY

- Removing four methods:
  * sp_type_def_list::type_defs_add_record()
  * sp_type_def_list::type_defs_add_composite2()
  * sp_pcontext::type_defs_declare_record()
  * sp_type_def_list::type_defs_declare_composite2()
  Adding two methods instead:
  * sp_type_def_list::type_defs_add()
  * sp_pcontext::type_defs_add()
  This allows to get rid of the duplicate code detecting data type
  declarations with the same name in the same sp_pcontext frame.

- Adding new methods:
  * LEX::declare_type_assoc_array()
  * LEX::LEX::declare_type_record()
  They create a type specific sp_type_def_xxx and the call the generic
  sp_pcontext::type_defs_add().

- m_key_def.sp_prepare_create_field() inside
  Field_assoc_array::create_fields() is now called for all key data types
  (not only for integers)

- Removing the assignment of key_def->charset in
  Type_handler_assoc_array::sp_variable_declarations_finalize().
  The charset is now evaluated in m_key_def.sp_prepare_create_field().

- Fixing Item_assoc_array::get_key() to set the character set of the "key"
  to utf8mb3 instead of binary

- Fixing Field_assoc_array::copy_and_convert_key() to set the key length
  limit in terms of the character length as specified in
  INDEX BY VARCHAR(N), instead of octet length. This is needed to make
  keys with multi-byte characters work correctly.
  Also it now raises different errors depending on the reason of the
  key conversion failures:
  * ER_INVALID_CHARACTER_STRING
  * ER_CANNOT_CONVERT_CHARACTER

- Changing the prototype for Type_handler_composite::key_to_lex_cstring() to

   virtual LEX_CSTRING key_to_lex_cstring(THD *thd,
                                          const sp_rcontext_addr &var,
                                          Item **key,
                                          String *buffer) const;
   * Now it returns a LEX_CSTRING, instead of getting it as an out parameter.
   * Gets an sp_rcontext_addr instead of "name" and "def"
   * Gets a String buffer which can be used to be passed to val_str(),
     or for character set conversion purposes.

- Removing Field_assoc_array::m_key_def, as all required information
  is available from Field_assoc_array::m_key_field.
  In Field_assoc_array::create_fields turning m_key_def to a local variable
  key_def.

- Fixing Field_assoc_array::copy_and_convert_key() to follow MariaDB coding
  style: only constants can be passed by-reference, not-constants should
  be passed by-pointer.

- Adding DBUG_ASSERTs into Type_handler_assoc_array::get_item()
  and Type_handler_assoc_array::get_or_create_item() that the passed
  key in "name" is well formed according to the charset of INDEX BY.

- Changing the error ER_TOO_LONG_KEY to ER_WRONG_STRING_LENGTH.
  The former prints length limit in bytes, which is not applicable
  for INDEX BY values, because its limit is in characters.
  Also, the latter is more verbose.

- Fixing the problem that these wrong uses of an assoc array variable:

    BEGIN
      assoc_var;
      assoc_var(1);
    END;

  raised a weird error message:
    ERROR 1054 (42S22): Unknown column 'assoc_var' in '(null)'

  Now a more readable parse error is raised.

- Adding a "Duplicate key" warning for the cases when assigning
  between two assoc arrays rejects some records due to different
  collations in their INDEX BY key definitions.

- Disallow INDEX OF propagation from VARCHAR to TEXT.
  The underlying code cannot handle TEXT.
  Adding tests.

- Adding a helper class StringBufferKey to pass to val_str() when
  a key value is evaluated.
  Fixing all val_str() calls to val_str(&buffer), as the former is
  not desirable.

- Fixing a wrong use of args[0]->null_value in
  Item_func_assoc_array_exists::val_bool()

- Fixing a problem that using TABLE OF TEXT crashed the server.
  Thanks to Iqbal Hassan for the proposed patch.

- Changes in Qualified_ident:
  * Fixing the Qualified_ident constructors to get all parst as
    Lex_ident_cli_st, rather than the first part as Lex_ident_cli_st
    with the following parts Lex_ident_sys.
    This makes the code more symmetric.
  * Fixing the grammar in sql_yacc.yy accordinly.
  * Fixing the data type storing the possition in the client query
    from "const char *" to Lex_ident_cli.
  * Adding a new method Qualified_ident::is_sane().
    It allows to reduce the code side in sql_yacc.yy.
    Thanks to Iqbal Hassan for the idea.

- Replacing qs_append() to append_ulonglong() in:
  * Item_method_func::print()
  * Item_splocal_assoc_array_element::print()
  * Item_splocal_assoc_array_element_field::print()

  These methods do not use reserve()/alloc(), so calling qs_append()
  was wrong and caused a crash.

- Changing the output formats of these methods:
  * Item_splocal_assoc_array_element::print()
  * Item_splocal_assoc_array_element_field::print()
  not to print the key two times.
  Also moving the `@123` part (the variable offset) immediately
  after the variabl name and before the `[key]` part.

- Fixing a memory leak happened when trying to insert a duplicate
  key into an assoc array. Also adding a new "THD *" parameter to
  Field_assoc_array::insert_element(). Thanks to Iqbal Hassan for the fix.
  Adding a test into sp-assoc-array-ctype.test.

- In  Field_assoc_array::create_fields: m_element_field->field_name is now
  set for all element data types (not only for records).
  This fixed a wrong variable name in warnings. Adding tests.

- Adding tests:
  * Adding tests for assoc array elements in UNIONs.

  * Copying from an assoc array with a varchar key
    to an assoc array with a shorter varchar key.

  * A relatively big associative array.

  * Memory usage for x86_64.

  * Package variable as assoc array keys.

  * Character set conversion

  * TABLE OF TEXT

  * TABLE OF VARCHAR(>64k bytes) propagation to TABLE OF TEXT.

  * TEXT element fields in an array of records.

  * VARCHAR->TEXT propagation in elements in an array of records.

  * Some more tests
2025-05-19 19:42:31 +04:00

1302 lines
38 KiB
C++

#ifndef SQL_STRING_INCLUDED
#define SQL_STRING_INCLUDED
/*
Copyright (c) 2000, 2013, Oracle and/or its affiliates.
Copyright (c) 2008, 2020, MariaDB Corporation.
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.
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
Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1335 USA */
/* This file is originally from the mysql distribution. Coded by monty */
#include <my_global.h>
#include <cmath>
#include "m_ctype.h" /* my_charset_bin */
#include <my_sys.h> /* alloc_root, my_free, my_realloc */
#include "m_string.h" /* TRASH */
#include "sql_const.h"
#include "sql_list.h"
class String;
#ifdef MYSQL_SERVER
extern PSI_memory_key key_memory_String_value;
#define STRING_PSI_MEMORY_KEY key_memory_String_value
#else
#define STRING_PSI_MEMORY_KEY PSI_NOT_INSTRUMENTED
#endif
typedef struct st_io_cache IO_CACHE;
typedef struct st_mem_root MEM_ROOT;
#define ASSERT_LENGTH(A) DBUG_ASSERT(str_length + (uint32) (A) <= Alloced_length)
#include "pack.h"
class Binary_string;
int sortcmp(const Binary_string *s, const Binary_string *t, CHARSET_INFO *cs);
int stringcmp(const Binary_string *s, const Binary_string *t);
String *copy_if_not_alloced(String *a,String *b,uint32 arg_length);
inline uint32 copy_and_convert(char *to, size_t to_length, CHARSET_INFO *to_cs,
const char *from, size_t from_length,
CHARSET_INFO *from_cs, uint *errors)
{
return my_convert(to, (uint)to_length, to_cs, from, (uint)from_length,
from_cs, errors);
}
class String_copier: public String_copy_status,
protected MY_STRCONV_STATUS
{
public:
const char *cannot_convert_error_pos() const
{ return m_cannot_convert_error_pos; }
const char *most_important_error_pos() const
{
return well_formed_error_pos() ? well_formed_error_pos() :
cannot_convert_error_pos();
}
/*
Convert a string between character sets.
"dstcs" and "srccs" cannot be &my_charset_bin.
*/
size_t convert_fix(CHARSET_INFO *dstcs, char *dst, size_t dst_length,
CHARSET_INFO *srccs, const char *src, size_t src_length,
size_t nchars)
{
return my_convert_fix(dstcs, dst, dst_length,
srccs, src, src_length, nchars, this, this);
}
/*
Copy a string. Fix bad bytes/characters to '?'.
*/
uint well_formed_copy(CHARSET_INFO *to_cs, char *to, size_t to_length,
CHARSET_INFO *from_cs, const char *from,
size_t from_length, size_t nchars);
// Same as above, but without the "nchars" limit.
uint well_formed_copy(CHARSET_INFO *to_cs, char *to, size_t to_length,
CHARSET_INFO *from_cs, const char *from,
size_t from_length)
{
return well_formed_copy(to_cs, to, to_length,
from_cs, from, from_length,
from_length /* No limit on "nchars"*/);
}
};
size_t my_copy_with_hex_escaping(CHARSET_INFO *cs,
char *dst, size_t dstlen,
const char *src, size_t srclen);
uint convert_to_printable(char *to, size_t to_len,
const char *from, size_t from_len,
CHARSET_INFO *from_cs, size_t nbytes= 0);
size_t convert_to_printable_required_length(uint len);
class Charset
{
CHARSET_INFO *m_charset;
public:
Charset() :m_charset(&my_charset_bin) { }
Charset(CHARSET_INFO *cs) :m_charset(cs) { }
CHARSET_INFO *charset() const { return m_charset; }
bool use_mb() const { return m_charset->use_mb(); }
uint mbminlen() const { return m_charset->mbminlen; }
uint mbmaxlen() const { return m_charset->mbmaxlen; }
bool is_good_for_ft() const
{
// Binary and UCS2/UTF16/UTF32 are not supported
return m_charset != &my_charset_bin && m_charset->mbminlen == 1;
}
size_t numchars(const char *str, const char *end) const
{
return m_charset->numchars(str, end);
}
size_t lengthsp(const char *str, size_t length) const
{
return m_charset->lengthsp(str, length);
}
size_t charpos(const char *str, const char *end, size_t pos) const
{
return m_charset->charpos(str, end, pos);
}
void set_charset(CHARSET_INFO *charset_arg)
{
m_charset= charset_arg;
}
void set_charset(const Charset &other)
{
m_charset= other.m_charset;
}
void swap(Charset &other)
{
swap_variables(CHARSET_INFO*, m_charset, other.m_charset);
}
bool same_encoding(const Charset &other) const
{
return my_charset_same(m_charset, other.m_charset);
}
/*
Collation name without the character set name.
For example, in case of "latin1_swedish_ci",
this method returns "_swedish_ci".
*/
LEX_CSTRING collation_specific_name() const;
bool encoding_allows_reinterpret_as(CHARSET_INFO *cs) const;
bool eq_collation_specific_names(CHARSET_INFO *cs) const;
bool can_have_collate_clause() const
{
return m_charset != &my_charset_bin;
}
/*
The MariaDB version when the last collation change happened,
e.g. due to a bug fix. See functions below.
*/
static ulong latest_mariadb_version_with_collation_change()
{
return 110002;
}
/*
Check if the collation with the given ID changed its order
since the given MariaDB version.
*/
static bool collation_changed_order(ulong mysql_version, uint cs_number)
{
if ((mysql_version < 50048 &&
(cs_number == 11 || /* ascii_general_ci - bug #29499, bug #27562 */
cs_number == 41 || /* latin7_general_ci - bug #29461 */
cs_number == 42 || /* latin7_general_cs - bug #29461 */
cs_number == 20 || /* latin7_estonian_cs - bug #29461 */
cs_number == 21 || /* latin2_hungarian_ci - bug #29461 */
cs_number == 22 || /* koi8u_general_ci - bug #29461 */
cs_number == 23 || /* cp1251_ukrainian_ci - bug #29461 */
cs_number == 26)) || /* cp1250_general_ci - bug #29461 */
(mysql_version < 50124 &&
(cs_number == 33 || /* utf8mb3_general_ci - bug #27877 */
cs_number == 35))) /* ucs2_general_ci - bug #27877 */
return true;
if (cs_number == 159 && /* ucs2_general_mysql500_ci - MDEV-30746 */
((mysql_version >= 100400 && mysql_version < 100429) ||
(mysql_version >= 100500 && mysql_version < 100520) ||
(mysql_version >= 100600 && mysql_version < 100613) ||
(mysql_version >= 100700 && mysql_version < 100708) ||
(mysql_version >= 100800 && mysql_version < 100808) ||
(mysql_version >= 100900 && mysql_version < 100906) ||
(mysql_version >= 101000 && mysql_version < 101004) ||
(mysql_version >= 101100 && mysql_version < 101103) ||
(mysql_version >= 110000 && mysql_version < 110002)))
return true;
return false;
}
/**
Check if a collation has changed ID since the given version.
Return the new ID.
@param mysql_version
@param cs_number - collation ID
@retval the new collation ID (or cs_number, if no change)
*/
static uint upgrade_collation_id(ulong mysql_version, uint cs_number)
{
if (mysql_version >= 50300 && mysql_version <= 50399)
{
switch (cs_number) {
case 149: return MY_PAGE2_COLLATION_ID_UCS2; // ucs2_crotian_ci
case 213: return MY_PAGE2_COLLATION_ID_UTF8; // utf8_crotian_ci
}
}
if ((mysql_version >= 50500 && mysql_version <= 50599) ||
(mysql_version >= 100000 && mysql_version <= 100005))
{
switch (cs_number) {
case 149: return MY_PAGE2_COLLATION_ID_UCS2; // ucs2_crotian_ci
case 213: return MY_PAGE2_COLLATION_ID_UTF8; // utf8_crotian_ci
case 214: return MY_PAGE2_COLLATION_ID_UTF32; // utf32_croatian_ci
case 215: return MY_PAGE2_COLLATION_ID_UTF16; // utf16_croatian_ci
case 245: return MY_PAGE2_COLLATION_ID_UTF8MB4;// utf8mb4_croatian_ci
}
}
return cs_number;
}
};
/**
Storage for strings with both length and allocated length.
Automatically grows on demand.
*/
class Binary_string: public Sql_alloc
{
protected:
char *Ptr;
uint32 str_length, Alloced_length, extra_alloc;
bool alloced, thread_specific;
void init_private_data()
{
Ptr= 0;
Alloced_length= extra_alloc= str_length= 0;
alloced= thread_specific= false;
}
inline void free_buffer()
{
if (alloced)
{
alloced=0;
my_free(Ptr);
}
}
public:
Binary_string()
{
init_private_data();
}
explicit Binary_string(size_t length_arg)
{
init_private_data();
(void) real_alloc(length_arg);
}
/*
NOTE: If one intend to use the c_ptr() method, the following two
constructors need the size of memory for STR to be at least LEN+1 (to make
room for zero termination).
*/
Binary_string(const char *str, size_t len)
{
Ptr= (char*) str;
str_length= (uint32) len;
Alloced_length= 0; /* Memory cannot be written to */
extra_alloc= 0;
alloced= thread_specific= 0;
}
Binary_string(char *str, size_t len)
{
Ptr= str;
str_length= Alloced_length= (uint32) len;
extra_alloc= 0;
alloced= thread_specific= 0;
}
explicit Binary_string(const Binary_string &str)
{
Ptr= str.Ptr;
str_length= str.str_length;
Alloced_length= str.Alloced_length;
extra_alloc= 0;
alloced= thread_specific= 0;
}
~Binary_string()
{
free();
}
inline uint32 length() const { return str_length;}
inline char& operator [] (size_t i) const { return Ptr[i]; }
inline void length(size_t len) { str_length=(uint32)len ; }
inline bool is_empty() const { return (str_length == 0); }
inline const char *ptr() const { return Ptr; }
inline const char *end() const { return Ptr + str_length; }
bool has_8bit_bytes() const
{
for (const char *c= ptr(), *c_end= end(); c < c_end; c++)
{
if (!my_isascii(*c))
return true;
}
return false;
}
bool bin_eq(const Binary_string *other) const
{
return length() == other->length() &&
!memcmp(ptr(), other->ptr(), length());
}
/*
PMG 2004.11.12
This is a method that works the same as perl's "chop". It simply
drops the last character of a string. This is useful in the case
of the federated storage handler where I'm building a unknown
number, list of values and fields to be used in a sql insert
statement to be run on the remote server, and have a comma after each.
When the list is complete, I "chop" off the trailing comma
ex.
String stringobj;
stringobj.append("VALUES ('foo', 'fi', 'fo',");
stringobj.chop();
stringobj.append(")");
In this case, the value of string was:
VALUES ('foo', 'fi', 'fo',
VALUES ('foo', 'fi', 'fo'
VALUES ('foo', 'fi', 'fo')
*/
inline void chop()
{
if (str_length)
{
str_length--;
Ptr[str_length]= '\0';
DBUG_ASSERT(strlen(Ptr) == str_length);
}
}
// Returns offset to substring or -1
int strstr(const Binary_string &search, uint32 offset=0) const;
int strstr(const char *search, uint32 search_length, uint32 offset=0) const;
// Returns offset to substring or -1
int strrstr(const Binary_string &search, uint32 offset=0) const;
/*
The following append operations do not extend the strings and in production
mode do NOT check that alloced memory!
q_*** methods writes values of parameters itself
qs_*** methods writes string representation of value
*/
void q_append(const char c)
{
ASSERT_LENGTH(1);
Ptr[str_length++] = c;
}
void q_append2b(const uint32 n)
{
ASSERT_LENGTH(2);
int2store(Ptr + str_length, n);
str_length += 2;
}
void q_append(const uint32 n)
{
ASSERT_LENGTH(4);
int4store(Ptr + str_length, n);
str_length += 4;
}
void q_append_int64(const longlong n)
{
ASSERT_LENGTH(8);
int8store(Ptr + str_length, n);
str_length += 8;
}
void q_append(double d)
{
ASSERT_LENGTH(8);
float8store(Ptr + str_length, d);
str_length += 8;
}
void q_append(double *d)
{
ASSERT_LENGTH(8);
float8store(Ptr + str_length, *d);
str_length += 8;
}
/*
Append a wide character.
The caller must have allocated at least cs->mbmaxlen bytes.
*/
int q_append_wc(my_wc_t wc, CHARSET_INFO *cs)
{
int mblen;
if ((mblen= cs->cset->wc_mb(cs, wc,
(uchar *) end(),
(uchar *) end() + cs->mbmaxlen)) > 0)
str_length+= (uint32) mblen;
return mblen;
}
void q_append(const char *data, size_t data_len)
{
ASSERT_LENGTH(data_len);
if (data_len)
memcpy(Ptr + str_length, data, data_len);
DBUG_ASSERT(str_length <= UINT_MAX32 - data_len);
str_length += (uint)data_len;
}
void q_append(const LEX_CSTRING *ls)
{
DBUG_ASSERT(ls->length < UINT_MAX32 &&
((ls->length == 0 && !ls->str) ||
ls->length == strlen(ls->str)));
q_append(ls->str, (uint32) ls->length);
}
void write_at_position(uint32 position, uint32 value)
{
DBUG_ASSERT(str_length >= position + 4);
int4store(Ptr + position,value);
}
void qs_append(const LEX_CSTRING *ls)
{
DBUG_ASSERT(ls->length < UINT_MAX32 &&
((ls->length == 0 && !ls->str) ||
ls->length == strlen(ls->str)));
qs_append(ls->str, (uint32)ls->length);
}
void qs_append(const char *str, size_t len);
void qs_append_hex(const char *str, uint32 len);
void qs_append_hex_uint32(uint32 num);
void qs_append(double d);
void qs_append(const double *d);
inline void qs_append(const char c)
{
ASSERT_LENGTH(1);
Ptr[str_length]= c;
str_length++;
}
void qs_append(int i);
void qs_append(uint i)
{
qs_append((ulonglong)i);
}
void qs_append(ulong i)
{
qs_append((ulonglong)i);
}
void qs_append(ulonglong i);
void qs_append(longlong i, int radix)
{
ASSERT_LENGTH(22);
char *buff= Ptr + str_length;
char *end= ll2str(i, buff, radix, 0);
str_length+= (uint32) (end-buff);
}
void qs_append_int64(longlong i);
/* Mark variable thread specific it it's not allocated already */
inline void set_thread_specific()
{
if (!alloced)
thread_specific= 1;
}
bool get_thread_specific() const
{
return thread_specific;
}
bool is_alloced() const { return alloced; }
inline uint32 alloced_length() const { return Alloced_length;}
inline uint32 extra_allocation() const { return extra_alloc;}
inline void extra_allocation(size_t len) { extra_alloc= (uint32)len; }
inline void mark_as_const() { Alloced_length= 0;}
inline bool uses_buffer_owned_by(const Binary_string *s) const
{
return (s->alloced && Ptr >= s->Ptr && Ptr < s->Ptr + s->Alloced_length);
}
/* Swap two string objects. Efficient way to exchange data without memcpy. */
void swap(Binary_string &s)
{
swap_variables(char *, Ptr, s.Ptr);
swap_variables(uint32, str_length, s.str_length);
swap_variables(uint32, Alloced_length, s.Alloced_length);
swap_variables(bool, alloced, s.alloced);
}
/**
Points the internal buffer to the supplied one. The old buffer is freed.
@param str Pointer to the new buffer.
@param arg_length Length of the new buffer in characters, excluding any
null character.
@note The new buffer will not be null terminated.
*/
void set_alloced(char *str, size_t length, size_t alloced_length)
{
free_buffer();
Ptr= str;
str_length= (uint32) length;
DBUG_ASSERT(alloced_length < UINT_MAX32);
Alloced_length= (uint32) alloced_length;
}
inline void set(char *str, size_t arg_length)
{
set_alloced(str, arg_length, arg_length);
}
inline void set(const char *str, size_t length)
{
free_buffer();
Ptr= (char*) str;
str_length= (uint32) length;
Alloced_length= 0;
}
void set(Binary_string &str, size_t offset, size_t length)
{
DBUG_ASSERT(&str != this);
free_buffer();
Ptr= str.Ptr + offset;
str_length= (uint32) length;
Alloced_length= 0;
if (str.Alloced_length)
Alloced_length= (uint32) (str.Alloced_length - offset);
}
LEX_CSTRING to_lex_cstring() const
{
LEX_CSTRING tmp= {Ptr, str_length};
return tmp;
}
inline LEX_CSTRING *get_value(LEX_CSTRING *res) const
{
res->str= Ptr;
res->length= str_length;
return res;
}
/* Take over handling of buffer from some other object */
void reset(char *ptr_arg, size_t length_arg, size_t alloced_length_arg)
{
set_alloced(ptr_arg, length_arg, alloced_length_arg);
alloced= ptr_arg != 0;
}
/* Forget about the buffer, let some other object handle it */
char *release()
{
char *old= Ptr;
init_private_data();
return old;
}
/*
This is used to set a new buffer for String.
However if the String already has an allocated buffer, it will
keep that one.
It's not to be used to set the value or length of the string.
*/
inline void set_buffer_if_not_allocated(char *str, size_t arg_length)
{
if (!alloced)
{
/*
Following should really set str_length= 0, but some code may
depend on that the String length is same as buffer length.
*/
Ptr= str;
str_length= Alloced_length= (uint32) arg_length;
}
/* One should set str_length before using it */
MEM_UNDEFINED(&str_length, sizeof(str_length));
}
inline Binary_string& operator=(const Binary_string &s)
{
if (&s != this)
{
/*
It is forbidden to do assignments like
some_string = substring_of_that_string
*/
DBUG_ASSERT(!s.uses_buffer_owned_by(this));
set_alloced((char *) s.Ptr, s.str_length, s.Alloced_length);
}
return *this;
}
bool set_hex(ulonglong num);
bool set_hex(const char *str, uint32 len);
bool set_fcvt(double num, uint decimals);
bool copy(); // Alloc string if not alloced
bool copy(const Binary_string &s); // Allocate new string
bool copy(const char *s, size_t arg_length); // Allocate new string
bool copy_or_move(const char *s,size_t arg_length);
/**
Convert a string to a printable format.
All non-convertable and control characters are replaced to 5-character
sequences '\hhhh'.
*/
bool copy_printable_hhhh(CHARSET_INFO *to_cs,
CHARSET_INFO *from_cs,
const char *from, size_t from_length);
bool append_ulonglong(ulonglong val);
bool append_longlong(longlong val);
bool append(const char *s, size_t size)
{
if (!size)
return false;
if (realloc_with_extra_if_needed(str_length + size))
return true;
q_append(s, size);
return false;
}
bool append(const LEX_CSTRING &s)
{
return append(s.str, s.length);
}
bool append(const Binary_string &s)
{
return append(s.ptr(), s.length());
}
bool append(IO_CACHE* file, uint32 arg_length);
inline bool append_char(char chr)
{
if (str_length < Alloced_length)
{
Ptr[str_length++]= chr;
}
else
{
if (unlikely(realloc_with_extra(str_length + 1)))
return true;
Ptr[str_length++]= chr;
}
return false;
}
bool append_hex(const char *src, uint32 srclen)
{
for (const char *src_end= src + srclen ; src != src_end ; src++)
{
if (unlikely(append_char(_dig_vec_lower[((uchar) *src) >> 4])) ||
unlikely(append_char(_dig_vec_lower[((uchar) *src) & 0x0F])))
return true;
}
return false;
}
bool append_hex_uint32(uint32 num)
{
if (reserve(8))
return true;
qs_append_hex_uint32(num);
return false;
}
bool append_with_step(const char *s, uint32 arg_length, uint32 step_alloc)
{
uint32 new_length= arg_length + str_length;
if (new_length > Alloced_length &&
unlikely(realloc(new_length + step_alloc)))
return true;
q_append(s, arg_length);
return false;
}
inline char *c_ptr()
{
if (unlikely(!Ptr))
return (char*) "";
/*
Here we assume that any buffer used to initialize String has
an end \0 or have at least an accessable character at end.
This is to handle the case of String("Hello",5) and
String("hello",5) efficiently.
We have two options here. To test for !Alloced_length or !alloced.
Using "Alloced_length" is slightly safer so that we do not read
from potentially uninitialized memory (normally not dangerous but
may give warnings in valgrind), but "alloced" is safer as there
are less change to get memory loss from code that is using
String((char*), length) or String.set((char*), length) and does
not free things properly (and there is several places in the code
where this happens and it is hard to find out if any of these will call
c_ptr().
*/
if (unlikely(!alloced && !Ptr[str_length]))
return Ptr;
if (str_length < Alloced_length)
{
Ptr[str_length]=0;
return Ptr;
}
(void) realloc(str_length); /* This will add end \0 */
return Ptr;
}
/*
One should use c_ptr() instead for most cases. This will be deleted soon,
kept for compatibility.
*/
inline char *c_ptr_quick()
{
return c_ptr_safe();
}
/*
This is to be used only in the case when one cannot use c_ptr().
The cases are:
- When one initializes String with an external buffer and length and
buffer[length] could be uninitialized when c_ptr() is called.
- When valgrind gives warnings about uninitialized memory with c_ptr().
*/
inline char *c_ptr_safe()
{
if (Ptr && str_length < Alloced_length)
Ptr[str_length]=0;
else
(void) realloc(str_length);
return Ptr;
}
inline void free()
{
free_buffer();
/*
We have to clear the values as some Strings, like in Field, are
reused after free(). Because of this we cannot use MEM_UNDEFINED() here.
*/
Ptr= 0;
str_length= 0;
Alloced_length= extra_alloc= 0;
}
inline bool alloc(size_t arg_length)
{
/*
Allocate if we need more space or if we don't have done any
allocation yet (we don't want to have Ptr to be NULL for empty strings).
Note that if arg_length == Alloced_length then we don't allocate.
This ensures we don't do any extra allocations in protocol and String:int,
but the string will not be automatically null terminated if c_ptr() is not
called.
*/
if (arg_length <= Alloced_length && Alloced_length)
return 0;
return real_alloc(arg_length);
}
bool real_alloc(size_t arg_length); // Empties old string
bool realloc_raw(size_t arg_length);
bool realloc(size_t arg_length)
{
if (realloc_raw(arg_length+1))
return TRUE;
Ptr[arg_length]= 0; // This make other funcs shorter
return FALSE;
}
bool realloc_with_extra(size_t arg_length)
{
if (extra_alloc < 4096)
extra_alloc= extra_alloc*2+128;
if (realloc_raw(arg_length + extra_alloc))
return TRUE;
Ptr[arg_length]=0; // This make other funcs shorter
return FALSE;
}
bool realloc_with_extra_if_needed(size_t arg_length)
{
if (arg_length < Alloced_length)
{
Ptr[arg_length]=0; // behave as if realloc was called.
return 0;
}
return realloc_with_extra(arg_length);
}
// Shrink the buffer, but only if it is allocated on the heap.
void shrink(size_t arg_length);
void move(Binary_string &s)
{
set_alloced(s.Ptr, s.str_length, s.Alloced_length);
extra_alloc= s.extra_alloc;
alloced= s.alloced;
thread_specific= s.thread_specific;
s.alloced= 0;
}
bool fill(size_t max_length,char fill);
/*
Replace substring with string
If wrong parameter or not enough memory, do nothing
*/
bool replace(uint32 offset,uint32 arg_length, const char *to, uint32 length);
bool replace(uint32 offset,uint32 arg_length, const Binary_string &to)
{
return replace(offset,arg_length,to.ptr(),to.length());
}
int reserve(size_t space_needed)
{
DBUG_ASSERT((ulonglong) str_length + space_needed < UINT_MAX32);
return realloc(str_length + space_needed);
}
int reserve(size_t space_needed, size_t grow_by);
inline char *prep_append(uint32 arg_length, uint32 step_alloc)
{
uint32 new_length= arg_length + str_length;
if (new_length > Alloced_length)
{
if (unlikely(realloc(new_length + step_alloc)))
return 0;
}
uint32 old_length= str_length;
str_length+= arg_length;
return Ptr + old_length; // Area to use
}
void q_net_store_length(ulonglong length)
{
DBUG_ASSERT(Alloced_length >= (str_length + net_length_size(length)));
char *pos= (char *) net_store_length((uchar *)(Ptr + str_length), length);
str_length= uint32(pos - Ptr);
}
void q_net_store_data(const uchar *from, size_t length)
{
DBUG_ASSERT(length < UINT_MAX32);
DBUG_ASSERT(Alloced_length >= (str_length + length +
net_length_size(length)));
q_net_store_length(length);
q_append((const char *)from, (uint32) length);
}
};
class String: public Charset, public Binary_string
{
public:
String() = default;
String(size_t length_arg) :Binary_string(length_arg)
{ }
/*
NOTE: If one intend to use the c_ptr() method, the following two
constructors need the size of memory for STR to be at least LEN+1 (to make
room for zero termination).
*/
String(const char *str, size_t len, CHARSET_INFO *cs)
:Charset(cs), Binary_string(str, len)
{ }
String(char *str, size_t len, CHARSET_INFO *cs)
:Charset(cs), Binary_string(str, len)
{ }
String(const String &str) = default;
void set(String &str,size_t offset,size_t arg_length)
{
Binary_string::set(str, offset, arg_length);
set_charset(str);
}
inline void set(char *str,size_t arg_length, CHARSET_INFO *cs)
{
Binary_string::set(str, arg_length);
set_charset(cs);
}
inline void set(const char *str,size_t arg_length, CHARSET_INFO *cs)
{
Binary_string::set(str, arg_length);
set_charset(cs);
}
bool set_ascii(const char *str, size_t arg_length);
inline void set_buffer_if_not_allocated(char *str,size_t arg_length,
CHARSET_INFO *cs)
{
Binary_string::set_buffer_if_not_allocated(str, arg_length);
set_charset(cs);
}
bool set_int(longlong num, bool unsigned_flag, CHARSET_INFO *cs);
bool set(int num, CHARSET_INFO *cs) { return set_int(num, false, cs); }
bool set(uint num, CHARSET_INFO *cs) { return set_int(num, true, cs); }
bool set(long num, CHARSET_INFO *cs) { return set_int(num, false, cs); }
bool set(ulong num, CHARSET_INFO *cs) { return set_int(num, true, cs); }
bool set(longlong num, CHARSET_INFO *cs) { return set_int(num, false, cs); }
bool set(ulonglong num, CHARSET_INFO *cs) { return set_int((longlong)num, true, cs); }
bool set_real_with_type(double num, uint decimals, CHARSET_INFO *cs, my_gcvt_arg_type);
bool set_real(double num,uint decimals, CHARSET_INFO *cs)
{ return set_real_with_type(num,decimals,cs,MY_GCVT_ARG_DOUBLE); }
bool set_real(float num,uint decimals, CHARSET_INFO *cs)
{ return set_real_with_type(num,decimals,cs,MY_GCVT_ARG_FLOAT); }
bool set_fcvt(double num, uint decimals)
{
set_charset(&my_charset_latin1);
return Binary_string::set_fcvt(num, decimals);
}
bool set_hex(ulonglong num)
{
set_charset(&my_charset_latin1);
return Binary_string::set_hex(num);
}
bool set_hex(const char *str, uint32 len)
{
set_charset(&my_charset_latin1);
return Binary_string::set_hex(str, len);
}
/* Take over handling of buffer from some other object */
void reset(char *ptr_arg, size_t length_arg, size_t alloced_length_arg,
CHARSET_INFO *cs)
{
Binary_string::reset(ptr_arg, length_arg, alloced_length_arg);
set_charset(cs);
}
inline String& operator = (const String &s)
{
if (&s != this)
{
set_charset(s);
Binary_string::operator=(s);
}
return *this;
}
bool copy()
{
return Binary_string::copy();
}
bool copy(const String &s)
{
set_charset(s);
return Binary_string::copy(s);
}
bool copy(const char *s, size_t arg_length, CHARSET_INFO *cs)
{
set_charset(cs);
return Binary_string::copy(s, arg_length);
}
bool copy_or_move(const char *s, size_t arg_length, CHARSET_INFO *cs)
{
set_charset(cs);
return Binary_string::copy_or_move(s, arg_length);
}
static bool needs_conversion(size_t arg_length,
CHARSET_INFO *cs_from, CHARSET_INFO *cs_to,
uint32 *offset);
static bool needs_conversion_on_storage(size_t arg_length,
CHARSET_INFO *cs_from,
CHARSET_INFO *cs_to);
bool copy_aligned(const char *s, size_t arg_length, size_t offset,
CHARSET_INFO *cs);
bool set_or_copy_aligned(const char *s, size_t arg_length, CHARSET_INFO *cs);
bool can_be_safely_converted_to(CHARSET_INFO *tocs) const
{
if (charset() == &my_charset_bin)
return Well_formed_prefix(tocs, ptr(), length()).length() == length();
String try_val;
uint try_conv_error= 0;
try_val.copy(ptr(), length(), charset(), tocs, &try_conv_error);
return try_conv_error == 0;
}
bool copy(const char*s, size_t arg_length, CHARSET_INFO *csfrom,
CHARSET_INFO *csto, uint *errors);
bool copy(const String *str, CHARSET_INFO *tocs, uint *errors)
{
return copy(str->ptr(), str->length(), str->charset(), tocs, errors);
}
bool copy(CHARSET_INFO *tocs,
CHARSET_INFO *fromcs, const char *src, size_t src_length,
size_t nchars, String_copier *copier)
{
if (unlikely(alloc(tocs->mbmaxlen * src_length)))
return true;
str_length= copier->well_formed_copy(tocs, Ptr, alloced_length(),
fromcs, src, (uint) src_length,
(uint) nchars);
set_charset(tocs);
return false;
}
bool copy_casedn(CHARSET_INFO *cs, const LEX_CSTRING &str)
{
size_t nbytes= str.length * cs->casedn_multiply();
DBUG_ASSERT(nbytes + 1 <= UINT_MAX32);
if (alloc(nbytes))
return true;
str_length= (uint32) cs->casedn_z(str.str, str.length, Ptr, nbytes + 1);
return false;
}
bool copy_caseup(CHARSET_INFO *cs, const LEX_CSTRING &str)
{
size_t nbytes= str.length * cs->caseup_multiply();
DBUG_ASSERT(nbytes + 1 <= UINT_MAX32);
if (alloc(nbytes))
return true;
str_length= (uint32) cs->caseup_z(str.str, str.length, Ptr, nbytes + 1);
return false;
}
// Append without character set conversion
bool append(const String &s)
{
return Binary_string::append(s);
}
inline bool append(char chr)
{
return Binary_string::append_char(chr);
}
bool append_hex(const char *src, uint32 srclen)
{
return Binary_string::append_hex(src, srclen);
}
bool append_hex(const uchar *src, uint32 srclen)
{
return Binary_string::append_hex((const char*)src, srclen);
}
bool append_introducer_and_hex(const String *str)
{
return
append('_') ||
append(str->charset()->cs_name) ||
append(STRING_WITH_LEN(" 0x")) ||
append_hex(str->ptr(), (uint32) str->length());
}
bool append(IO_CACHE* file, uint32 arg_length)
{
return Binary_string::append(file, arg_length);
}
inline bool append(const char *s, uint32 arg_length, uint32 step_alloc)
{
return append_with_step(s, arg_length, step_alloc);
}
// Append with optional character set conversion from ASCII (e.g. to UCS2)
bool append(const LEX_CSTRING *ls)
{
DBUG_ASSERT(ls->length < UINT_MAX32 &&
((ls->length == 0 && !ls->str) ||
ls->length == strlen(ls->str)));
return append(ls->str, (uint32) ls->length);
}
bool append(const LEX_CSTRING &ls)
{
return append(&ls);
}
bool append_name_value(const LEX_CSTRING &name,
const LEX_CSTRING &value,
uchar quot= '\0')
{
return
append(name) ||
append('=') ||
(quot && append(quot)) ||
append(value) ||
(quot && append(quot));
}
bool append(const char *s, size_t size);
bool append_parenthesized(long nr, int radix= 10);
// Append with optional character set conversion from cs to charset()
bool append(const char *s, size_t arg_length, CHARSET_INFO *cs);
bool append(const LEX_CSTRING &s, CHARSET_INFO *cs)
{
return append(s.str, s.length, cs);
}
// Append a wide character
bool append_wc(my_wc_t wc)
{
if (reserve(mbmaxlen()))
return true;
int mblen= q_append_wc(wc, charset());
if (mblen > 0)
return false;
else if (mblen == MY_CS_ILUNI && wc != '?')
return q_append_wc('?', charset()) <= 0;
return true;
}
// Append a number with zero prefilling
bool append_zerofill(uint num, uint width)
{
static const char zeros[15]= "00000000000000";
char intbuff[15];
uint length= (uint) (int10_to_str(num, intbuff, 10) - intbuff);
if (length < width &&
append(zeros, width - length, &my_charset_latin1))
return true;
return append(intbuff, length, &my_charset_latin1);
}
/*
Append a bitmask in an uint32 with a translation into a
C-style human readable representation, e.g.:
0x05 -> "(flag04|flag01)"
@param flags - the flags to translate
@param names - an array of flag names
@param count - the number of available elements in "names"
*/
bool append_flag32_names(uint32 flags, LEX_CSTRING names[], size_t count)
{
bool added= false;
if (flags && append('('))
return true;
for (ulong i= 0; i <= 31; i++)
{
ulong bit= 31 - i;
if (flags & (1 << bit))
{
if (added && append('|'))
return true;
if (bit < count ? append(names[bit]) : append('?'))
return true;
added= true;
}
}
if (flags && append(')'))
return true;
return false;
}
void strip_sp();
friend String *copy_if_not_alloced(String *a,String *b,uint32 arg_length);
friend class Field;
uint32 numchars() const
{
return (uint32) Charset::numchars(ptr(), end());
}
int charpos(longlong i, uint32 offset=0)
{
if (i <= 0)
return (int) i;
return (int) Charset::charpos(ptr() + offset, end(), (size_t) i);
}
size_t lengthsp() const
{
return Charset::lengthsp(Ptr, str_length);
}
void print(String *to) const;
void print_with_conversion(String *to, CHARSET_INFO *cs) const;
void print(String *to, CHARSET_INFO *cs) const
{
if (my_charset_same(charset(), cs))
print(to);
else
print_with_conversion(to, cs);
}
static my_wc_t escaped_wc_for_single_quote(my_wc_t ch)
{
switch (ch) {
case '\\': return '\\';
case '\0': return '0';
case '\'': return '\'';
case '\b': return 'b';
case '\t': return 't';
case '\n': return 'n';
case '\r': return 'r';
case '\032': return 'Z';
}
return 0;
}
// Append for single quote using mb_wc/wc_mb Unicode conversion
bool append_for_single_quote_using_mb_wc(const char *str, size_t length,
CHARSET_INFO *cs);
// Append for single quote with optional mb_wc/wc_mb conversion
bool append_for_single_quote_opt_convert(const char *str,
size_t length,
CHARSET_INFO *cs)
{
return charset() == &my_charset_bin || cs == &my_charset_bin ||
my_charset_same(charset(), cs) ?
append_for_single_quote(str, length) :
append_for_single_quote_using_mb_wc(str, length, cs);
}
bool append_for_single_quote_opt_convert(const String &str)
{
return append_for_single_quote_opt_convert(str.ptr(),
str.length(),
str.charset());
}
bool append_for_single_quote(const char *st, size_t len);
bool append_for_single_quote(const String *s)
{
return append_for_single_quote(s->ptr(), s->length());
}
void swap(String &s)
{
Charset::swap(s);
Binary_string::swap(s);
}
uint well_formed_length() const
{
return (uint) Well_formed_prefix(charset(), ptr(), length()).length();
}
bool is_ascii() const
{
if (length() == 0)
return TRUE;
if (charset()->mbminlen > 1)
return FALSE;
return !has_8bit_bytes();
}
bool eq(const String *other, CHARSET_INFO *cs) const
{
return !sortcmp(this, other, cs);
}
private:
bool append_semi_hex(const char *s, uint len, CHARSET_INFO *cs);
};
// The following class is a backport from MySQL 5.6:
/**
String class wrapper with a preallocated buffer of size buff_sz
This class allows to replace sequences of:
char buff[12345];
String str(buff, sizeof(buff));
str.length(0);
with a simple equivalent declaration:
StringBuffer<12345> str;
*/
template<size_t buff_sz>
class StringBuffer : public String
{
char buff[buff_sz];
public:
StringBuffer() : String(buff, buff_sz, &my_charset_bin) { length(0); }
explicit StringBuffer(CHARSET_INFO *cs) : String(buff, buff_sz, cs)
{
length(0);
}
void set_buffer_if_not_allocated(CHARSET_INFO *cs)
{
if (!is_alloced())
{
Ptr= buff;
Alloced_length= (uint32) buff_sz;
}
str_length= 0; /* Safety, not required */
/* One should set str_length before using it */
MEM_UNDEFINED(&str_length, sizeof(str_length));
set_charset(cs);
}
};
template<size_t buff_sz>
class BinaryStringBuffer : public Binary_string
{
char buff[buff_sz];
public:
BinaryStringBuffer() : Binary_string(buff, buff_sz) { length(0); }
};
static inline bool check_if_only_end_space(CHARSET_INFO *cs,
const char *str,
const char *end)
{
return str + cs->scan(str, end, MY_SEQ_SPACES) == end;
}
int append_query_string(CHARSET_INFO *csinfo, String *to,
const char *str, size_t len, bool no_backslash);
#endif /* SQL_STRING_INCLUDED */