• unknown's avatar
    fix for bug #12595 (ESCAPE must be exactly 1 character long) · ccc051cd
    unknown authored
    ESCAPE has length of 1 if specified and sql_mode is NO_BACKSLASH_ESCAPES
    or has length of 0 or 1 in every other situation.
    (approved patch applied on a up-to-date tree re-commit) 
    
    
    mysql-test/r/select.result:
      results of test for bug 12595
    mysql-test/t/select.test:
      test for bug #12595 (ESCAPE must be exactly one character long)
    sql/item_cmpfunc.cc:
      if ESCAPE was in the statement check whether its length is
      different than 1. In NO_BACKSLASH_ESCAPES mode only length of 1 is
      allowed, otherwise the length could be 0 or 1 character (code point
      in the sense of Unicode).
    sql/item_cmpfunc.h:
      pass variable from the parsing stage - whether
      ESCAPE clause was found in the statement
    sql/sql_help.cc:
      pass FALSE for escape_used_in_parsing because we
      want the default mode of no error checking - our internal code.
    sql/sql_lex.cc:
      initialized variable used to transfer information during
      parsing up in the stack when reducing in the grammar
    sql/sql_lex.h:
      new variable used for transfering information when
      reducing in the grammar.
    sql/sql_yacc.yy:
      initialize Lex->escape_used and then use it when reducing.
      This is needed as fix for bug #12595 to distinguish between
      situation where ESCAPE was found and when not because internally
      we may pass a string an empty string and there is no other way
      to find out whether this is correct or not in case of 
      NO_BACKSLASH_ESCAPES mode, which allows only length of 1 if
      ESCAPE is part of the SQL statement.
    ccc051cd
item_cmpfunc.h 37.9 KB