Commit 6d6b38c2 authored by unknown's avatar unknown
Browse files

Bug #6737: REGEXP gives wrong result with case sensitive collation: 

- A new flag MY_CS_CSSORT was introduced for case sensitivity.
- Item_func_regexp doesn't substiture ICASE not only
  for binary collations but for case sensitive collations as well.
parent 5a7198f5

include/m_ctype.h

+1 −1
Original line number Diff line number Diff line
@@ -63,7 +63,7 @@ typedef struct unicase_info_st 
#define MY_CS_UNICODE	128    /* is a charset is full unicode   */

#define MY_CS_READY	256    /* if a charset is initialized    */

#define MY_CS_AVAILABLE	512    /* If either compiled-in or loaded*/



#define MY_CS_CSSORT	1024   /* if case sensitive sort order   */	

#define MY_CHARSET_UNDEFINED 0

mysql-test/r/ctype_latin1.result

+9 −0
Original line number Diff line number Diff line
@@ -296,3 +296,12 @@ FD C3BD FD 1 
FE	C3BE	FE	1

FF	C3BF	FF	1

DROP TABLE t1;

select 'a' regexp 'A' collate latin1_general_ci;

'a' regexp 'A' collate latin1_general_ci

1

select 'a' regexp 'A' collate latin1_general_cs;

'a' regexp 'A' collate latin1_general_cs

0

select 'a' regexp 'A' collate latin1_bin;

'a' regexp 'A' collate latin1_bin

0

mysql-test/t/ctype_latin1.test

+7 −0
Original line number Diff line number Diff line
@@ -53,3 +53,10 @@ SELECT 
  hex(@l:=convert(@u using latin1)),

  a=@l FROM t1;

DROP TABLE t1;



#

# Bug #6737: REGEXP gives wrong result with case sensitive collation

#

select 'a' regexp 'A' collate latin1_general_ci;

select 'a' regexp 'A' collate latin1_general_cs;

select 'a' regexp 'A' collate latin1_bin;

mysys/charset.c

+11 −0
Original line number Diff line number Diff line
@@ -228,6 +228,7 @@ static int add_collation(CHARSET_INFO *cs) 
      }

      else

      {

        uchar *sort_order= all_charsets[cs->number]->sort_order;

        simple_cs_init_functions(all_charsets[cs->number]);

        new->mbminlen= 1;

        new->mbmaxlen= 1;
@@ -236,6 +237,16 @@ static int add_collation(CHARSET_INFO *cs) 
          all_charsets[cs->number]->state |= MY_CS_LOADED;

        }

        all_charsets[cs->number]->state|= MY_CS_AVAILABLE;

        

        /*

          Check if case sensitive sort order: A < a < B.

          We need MY_CS_FLAG for regex library, and for

          case sensitivity flag for 5.0 client protocol,

          to support isCaseSensitive() method in JDBC driver 

        */

        if (sort_order && sort_order['A'] < sort_order['a'] &&

                          sort_order['a'] < sort_order['B'])

          all_charsets[cs->number]->state|= MY_CS_CSSORT; 

      }

    }

    else

sql/item_cmpfunc.cc

+11 −9
Original line number Diff line number Diff line
@@ -2365,7 +2365,8 @@ Item_func_regex::fix_fields(THD *thd, TABLE_LIST *tables, Item **ref) 
    }

    int error;

    if ((error= regcomp(&preg,res->c_ptr(),

		       (cmp_collation.collation->state & MY_CS_BINSORT) ?

                        ((cmp_collation.collation->state & MY_CS_BINSORT) ||

                         (cmp_collation.collation->state & MY_CS_CSSORT)) ?

                         REG_EXTENDED | REG_NOSUB :

                         REG_EXTENDED | REG_NOSUB | REG_ICASE,

                        cmp_collation.collation)))
@@ -2416,7 +2417,8 @@ longlong Item_func_regex::val_int() 
	regex_compiled=0;

      }

      if (regcomp(&preg,res2->c_ptr(),

		  (cmp_collation.collation->state & MY_CS_BINSORT) ?

                  ((cmp_collation.collation->state & MY_CS_BINSORT) ||

                   (cmp_collation.collation->state & MY_CS_CSSORT)) ?

                   REG_EXTENDED | REG_NOSUB :

                   REG_EXTENDED | REG_NOSUB | REG_ICASE,

                   cmp_collation.collation))