mariadb/myisam/ft_parser.c
unknown 5ab1dbf567 some DBUG_ENTER/RETURN macro added
safer GWS_PROB, avoid log(0)


myisam/ft_nlq_search.c:
  some DBUG_ENTER/RETURN macro added
myisam/ft_parser.c:
  some DBUG_ENTER/RETURN macro added
myisam/ftdefs.h:
  safer GWS_PROB, avoid log(0).
sql/item_func.cc:
  some DBUG_ENTER/RETURN macro added
2002-10-18 14:53:46 +00:00

242 lines
5.9 KiB
C

/* Copyright (C) 2000 MySQL AB & MySQL Finland AB & TCX DataKonsult AB
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; either version 2 of the License, or
(at your option) any later version.
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., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
/* Written by Sergei A. Golubchik, who has a shared copyright to this code */
#include "ftdefs.h"
#ifdef EVAL_RUN
#ifdef PIVOT_STAT
ulong collstat=0;
#endif
#endif /* EVAL_RUN */
typedef struct st_ft_docstat {
FT_WORD *list;
uint uniq;
double sum;
#ifdef EVAL_RUN
uint words, totlen;
double max, nsum, nsum2;
#endif /* EVAL_RUN */
} FT_DOCSTAT;
static int FT_WORD_cmp(CHARSET_INFO* cs, FT_WORD *w1, FT_WORD *w2)
{
return _mi_compare_text(cs, (uchar*) w1->pos, w1->len,
(uchar*) w2->pos, w2->len, 0);
}
static int walk_and_copy(FT_WORD *word,uint32 count,FT_DOCSTAT *docstat)
{
word->weight=LWS_IN_USE;
#ifdef EVAL_RUN
word->cnt= (uchar) count;
if(docstat->max < word->weight) docstat->max=word->weight;
docstat->words+=count;
docstat->totlen+=word->len;
#endif /* EVAL_RUN */
docstat->sum+=word->weight;
memcpy_fixed((docstat->list)++,word,sizeof(FT_WORD));
return 0;
}
/* transforms tree of words into the array, applying normalization */
FT_WORD * ft_linearize(TREE *wtree)
{
FT_WORD *wlist,*p;
FT_DOCSTAT docstat;
DBUG_ENTER("ft_linearize");
if ((wlist=(FT_WORD *) my_malloc(sizeof(FT_WORD)*
(1+wtree->elements_in_tree),MYF(0))))
{
docstat.list=wlist;
docstat.uniq=wtree->elements_in_tree;
#ifdef EVAL_RUN
docstat.nsum=docstat.nsum2=docstat.max=docstat.words=docstat.totlen=
#endif /* EVAL_RUN */
docstat.sum=0;
tree_walk(wtree,(tree_walk_action)&walk_and_copy,&docstat,left_root_right);
}
delete_tree(wtree);
if (!wlist)
DBUG_RETURN(NULL);
docstat.list->pos=NULL;
for (p=wlist;p->pos;p++)
{
p->weight=PRENORM_IN_USE;
#ifdef EVAL_RUN
docstat.nsum+=p->weight;
docstat.nsum2+=p->weight*p->weight;
#endif /* EVAL_RUN */
}
#ifdef EVAL_RUN
#ifdef PIVOT_STAT
collstat+=PIVOT_STAT;
#endif
#endif /* EVAL_RUN */
for (p=wlist;p->pos;p++)
{
p->weight/=NORM_IN_USE;
}
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
* 2 - left bracket
* 3 - right bracket
*/
byte ft_get_word(byte **start, byte *end, FT_WORD *word, FTB_PARAM *param)
{
byte *doc=*start;
int mwc;
param->yesno=(FTB_YES==' ') ? 1 : (param->quot != 0);
param->plusminus=param->pmsign=0;
while (doc<end)
{
for (;doc<end;doc++)
{
if (true_word_char(*doc)) break;
if (*doc == FTB_RQUOT && param->quot) {
param->quot=doc;
*start=doc+1;
return 3; /* FTB_RBR */
}
if ((*doc == FTB_LBR || *doc == FTB_RBR || *doc == FTB_LQUOT)
&& !param->quot)
{
/* param->prev=' '; */
*start=doc+1;
if (*doc == FTB_LQUOT) param->quot=*start;
return (*doc == FTB_RBR)+2;
}
if (param->prev == ' ' && !param->quot)
{
if (*doc == FTB_YES ) { param->yesno=+1; continue; } else
if (*doc == FTB_EGAL) { param->yesno= 0; continue; } else
if (*doc == FTB_NO ) { param->yesno=-1; continue; } else
if (*doc == FTB_INC ) { param->plusminus++; continue; } else
if (*doc == FTB_DEC ) { param->plusminus--; continue; } else
if (*doc == FTB_NEG ) { param->pmsign=!param->pmsign; continue; }
}
param->prev=*doc;
param->yesno=(FTB_YES==' ') ? 1 : (param->quot != 0);
param->plusminus=param->pmsign=0;
}
mwc=0;
for (word->pos=doc; doc<end; doc++)
if (true_word_char(*doc))
mwc=0;
else if (!misc_word_char(*doc) || mwc++)
break;
param->prev='A'; /* be sure *prev is true_word_char */
word->len= (uint)(doc-word->pos) - mwc;
if ((param->trunc=(doc<end && *doc == FTB_TRUNC)))
doc++;
if (((word->len >= ft_min_word_len && !is_stopword(word->pos, word->len))
|| param->trunc) && word->len < ft_max_word_len)
{
*start=doc;
return 1;
}
}
return 0;
}
byte ft_simple_get_word(byte **start, byte *end, FT_WORD *word)
{
byte *doc=*start;
int mwc;
DBUG_ENTER("ft_simple_get_word");
while (doc<end)
{
for (;doc<end;doc++)
{
if (true_word_char(*doc)) break;
}
mwc=0;
for(word->pos=doc; doc<end; doc++)
if (true_word_char(*doc))
mwc=0;
else if (!misc_word_char(*doc) || mwc++)
break;
word->len= (uint)(doc-word->pos) - mwc;
if (word->len >= ft_min_word_len && word->len < ft_max_word_len &&
!is_stopword(word->pos, word->len))
{
*start=doc;
DBUG_RETURN(1);
}
}
DBUG_RETURN(0);
}
void ft_parse_init(TREE *wtree, CHARSET_INFO *cs)
{
DBUG_ENTER("ft_parse_init");
if (!is_tree_inited(wtree))
init_tree(wtree,0,0,sizeof(FT_WORD),(qsort_cmp2)&FT_WORD_cmp,0,NULL, cs);
DBUG_VOID_RETURN;
}
int ft_parse(TREE *wtree, byte *doc, int doclen)
{
byte *end=doc+doclen;
FT_WORD w;
DBUG_ENTER("ft_parse");
while (ft_simple_get_word(&doc,end,&w))
{
if (!tree_insert(wtree, &w, 0))
goto err;
}
DBUG_RETURN(0);
err:
delete_tree(wtree);
DBUG_RETURN(1);
}