Commit 0bfea087 authored by unknown's avatar unknown
Browse files

backport from 4.1: 

"phrase search" should not match partial words (it should not match 'paraphrase searches')
parent 2d776e36

myisam/ft_boolean_search.c

+19 −10
Original line number Diff line number Diff line
@@ -360,25 +360,34 @@ FT_INFO * ft_init_boolean_search(MI_INFO *info, uint keynr, byte *query, 
}





/* returns 1 if str0 contain str1 */

/* returns 1 if str0 ~= /\<str1\>/ */

static int _ftb_strstr(const byte *s0, const byte *e0,

                const byte *s1, const byte *e1,

                CHARSET_INFO *cs)

{

  const byte *p;

  const byte *p0, *p1;

  my_bool s_after, e_before;



  while (s0 < e0)

  s_after=true_word_char(s1[0]);

  e_before=true_word_char(e1[-1]);

  p0=s0;



  while (p0 < e0)

  {

    while (s0 < e0 && cs->to_upper[(uint) (uchar) *s0++] !=

    while (p0 < e0 && cs->to_upper[(uint) (uchar) *p0++] !=

	   cs->to_upper[(uint) (uchar) *s1])

      /* no-op */;

    if (s0 >= e0)

    if (p0 >= e0)

      return 0;

    p=s1+1;

    while (s0 < e0 && p < e1 && cs->to_upper[(uint) (uchar) *s0] ==

	   cs->to_upper[(uint) (uchar) *p])

      s0++, p++;

    if (p >= e1)



    if (s_after && p0-1 > s0 && true_word_char(p0[-2]))

      continue;



    p1=s1+1;

    while (p0 < e0 && p1 < e1 && cs->to_upper[(uint) (uchar) *p0] ==

	   cs->to_upper[(uint) (uchar) *p1])

      p0++, p1++;

    if (p1 == e1 && (!e_before || p0 == e0 || !true_word_char(p0[0])))

      return 1;

  }

  return 0;

myisam/ft_parser.c

+0 −9
Original line number Diff line number Diff line
@@ -105,15 +105,6 @@ FT_WORD * ft_linearize(TREE *wtree) 
  DBUG_RETURN(wlist);

}



#define true_word_char(X)	(isalnum(X) || (X)=='_')

#ifdef HYPHEN_IS_DELIM

#define misc_word_char(X)	((X)=='\'')

#else

#define misc_word_char(X)	((X)=='\'' || (X)=='-')

#endif

#define word_char(X)		(true_word_char(X) || misc_word_char(X))





/* returns:

 * 0 - eof

 * 1 - word found

myisam/ftdefs.h

+3 −2
Original line number Diff line number Diff line
@@ -22,8 +22,9 @@ 
#include <m_ctype.h>

#include <my_tree.h>



#define HYPHEN_IS_DELIM

#define HYPHEN_IS_CONCAT     /* not used for now */

#define true_word_char(X)       (isalnum(X) || (X)=='_')

#define misc_word_char(X)       ((X)=='\'')

#define word_char(X)            (true_word_char(X) || misc_word_char(X))



#define COMPILE_STOPWORDS_IN

mysql-test/r/fulltext.result

+2 −1
Original line number Diff line number Diff line
@@ -116,7 +116,8 @@ a b 
MySQL has now support	for full-text search

select * from t1 where MATCH a,b AGAINST ('"text i"' IN BOOLEAN MODE);

a	b

Full-text indexes	are called collections

select * from t1 where MATCH a,b AGAINST ('"xt indexes"' IN BOOLEAN MODE);

a	b

select * from t1 where MATCH a,b AGAINST ('+(support collections) +foobar*' IN BOOLEAN MODE);

a	b

select * from t1 where MATCH a,b AGAINST ('+(+(support collections)) +foobar*' IN BOOLEAN MODE);

mysql-test/t/fulltext.test

+1 −0
Original line number Diff line number Diff line
@@ -52,6 +52,7 @@ select * from t1 where MATCH a,b AGAINST ('"text search" "now support"' IN BOOL 
select * from t1 where MATCH a,b AGAINST ('"text search" -"now support"' IN BOOLEAN MODE);

select * from t1 where MATCH a,b AGAINST ('"text search" +"now support"' IN BOOLEAN MODE);

select * from t1 where MATCH a,b AGAINST ('"text i"' IN BOOLEAN MODE);

select * from t1 where MATCH a,b AGAINST ('"xt indexes"' IN BOOLEAN MODE);



select * from t1 where MATCH a,b AGAINST ('+(support collections) +foobar*' IN BOOLEAN MODE);

select * from t1 where MATCH a,b AGAINST ('+(+(support collections)) +foobar*' IN BOOLEAN MODE);