1
/* Copyright (C) 2006 MySQL AB & MySQL Finland AB & TCX DataKonsult AB
3
This program is free software; you can redistribute it and/or modify
4
it under the terms of the GNU General Public License as published by
5
the Free Software Foundation; version 2 of the License.
7
This program is distributed in the hope that it will be useful,
8
but WITHOUT ANY WARRANTY; without even the implied warranty of
9
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
10
GNU General Public License for more details.
12
You should have received a copy of the GNU General Public License
13
along with this program; if not, write to the Free Software
14
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA */
16
/* Written by Sergei A. Golubchik, who has a shared copyright to this code */
18
/* functions to work with full-text indices */
20
#include "ma_ftdefs.h"
23
void _ma_ft_segiterator_init(MARIA_HA *info, uint keynr, const uchar *record,
24
FT_SEG_ITERATOR *ftsi)
26
DBUG_ENTER("_ma_ft_segiterator_init");
28
ftsi->num=info->s->keyinfo[keynr].keysegs;
29
ftsi->seg=info->s->keyinfo[keynr].seg;
34
void _ma_ft_segiterator_dummy_init(const uchar *record, uint len,
35
FT_SEG_ITERATOR *ftsi)
37
DBUG_ENTER("_ma_ft_segiterator_dummy_init");
47
This function breaks convention "return 0 in success"
48
but it's easier to use like this
50
while(_ma_ft_segiterator())
52
so "1" means "OK", "0" means "EOF"
55
uint _ma_ft_segiterator(register FT_SEG_ITERATOR *ftsi)
57
DBUG_ENTER("_ma_ft_segiterator");
68
if (ftsi->seg->null_bit &&
69
(ftsi->rec[ftsi->seg->null_pos] & ftsi->seg->null_bit))
74
ftsi->pos= ftsi->rec+ftsi->seg->start;
75
if (ftsi->seg->flag & HA_VAR_LENGTH_PART)
77
uint pack_length= (ftsi->seg->bit_start);
78
ftsi->len= (pack_length == 1 ? (uint) * ftsi->pos :
79
uint2korr(ftsi->pos));
80
ftsi->pos+= pack_length; /* Skip VARCHAR length */
83
if (ftsi->seg->flag & HA_BLOB_PART)
85
ftsi->len= _ma_calc_blob_length(ftsi->seg->bit_start,ftsi->pos);
86
memcpy_fixed((char*) &ftsi->pos, ftsi->pos+ftsi->seg->bit_start,
90
ftsi->len=ftsi->seg->length;
95
/* parses a document i.e. calls maria_ft_parse for every keyseg */
97
uint _ma_ft_parse(TREE *parsed, MARIA_HA *info, uint keynr, const uchar *record,
98
MYSQL_FTPARSER_PARAM *param, MEM_ROOT *mem_root)
100
FT_SEG_ITERATOR ftsi;
101
struct st_mysql_ftparser *parser;
102
DBUG_ENTER("_ma_ft_parse");
104
_ma_ft_segiterator_init(info, keynr, record, &ftsi);
106
maria_ft_parse_init(parsed, info->s->keyinfo[keynr].seg->charset);
107
parser= info->s->keyinfo[keynr].parser;
108
while (_ma_ft_segiterator(&ftsi))
110
/** @todo this casts ftsi.pos (const) to non-const */
112
if (maria_ft_parse(parsed, (uchar *)ftsi.pos, ftsi.len, parser, param,
119
FT_WORD * _ma_ft_parserecord(MARIA_HA *info, uint keynr, const uchar *record,
123
MYSQL_FTPARSER_PARAM *param;
124
DBUG_ENTER("_ma_ft_parserecord");
125
if (! (param= maria_ftparser_call_initializer(info, keynr, 0)))
127
bzero((char*) &ptree, sizeof(ptree));
129
if (_ma_ft_parse(&ptree, info, keynr, record, param, mem_root))
132
DBUG_RETURN(maria_ft_linearize(&ptree, mem_root));
135
static int _ma_ft_store(MARIA_HA *info, uint keynr, uchar *keybuf,
136
FT_WORD *wlist, my_off_t filepos)
138
DBUG_ENTER("_ma_ft_store");
140
for (; wlist->pos; wlist++)
143
_ma_ft_make_key(info, &key, keynr, keybuf, wlist, filepos);
144
if (_ma_ck_write(info, &key))
150
static int _ma_ft_erase(MARIA_HA *info, uint keynr, uchar *keybuf,
151
FT_WORD *wlist, my_off_t filepos)
154
DBUG_ENTER("_ma_ft_erase");
156
for (; wlist->pos; wlist++)
159
_ma_ft_make_key(info, &key, keynr, keybuf, wlist, filepos);
160
if (_ma_ck_delete(info, &key))
167
Compares an appropriate parts of two WORD_KEY keys directly out of records
168
returns 1 if they are different
171
#define THOSE_TWO_DAMN_KEYS_ARE_REALLY_DIFFERENT 1
172
#define GEE_THEY_ARE_ABSOLUTELY_IDENTICAL 0
174
int _ma_ft_cmp(MARIA_HA *info, uint keynr, const uchar *rec1, const uchar *rec2)
176
FT_SEG_ITERATOR ftsi1, ftsi2;
177
CHARSET_INFO *cs=info->s->keyinfo[keynr].seg->charset;
178
DBUG_ENTER("_ma_ft_cmp");
180
_ma_ft_segiterator_init(info, keynr, rec1, &ftsi1);
181
_ma_ft_segiterator_init(info, keynr, rec2, &ftsi2);
183
while (_ma_ft_segiterator(&ftsi1) && _ma_ft_segiterator(&ftsi2))
185
if ((ftsi1.pos != ftsi2.pos) &&
186
(!ftsi1.pos || !ftsi2.pos ||
187
ha_compare_text(cs, ftsi1.pos,ftsi1.len,
188
ftsi2.pos,ftsi2.len,0,0)))
189
DBUG_RETURN(THOSE_TWO_DAMN_KEYS_ARE_REALLY_DIFFERENT);
191
DBUG_RETURN(GEE_THEY_ARE_ABSOLUTELY_IDENTICAL);
195
/* update a document entry */
197
int _ma_ft_update(MARIA_HA *info, uint keynr, uchar *keybuf,
198
const uchar *oldrec, const uchar *newrec, my_off_t pos)
201
FT_WORD *oldlist,*newlist, *old_word, *new_word;
202
CHARSET_INFO *cs=info->s->keyinfo[keynr].seg->charset;
204
DBUG_ENTER("_ma_ft_update");
206
if (!(old_word=oldlist=_ma_ft_parserecord(info, keynr, oldrec,
207
&info->ft_memroot)) ||
208
!(new_word=newlist=_ma_ft_parserecord(info, keynr, newrec,
213
while(old_word->pos && new_word->pos)
215
cmp= ha_compare_text(cs, (uchar*) old_word->pos,old_word->len,
216
(uchar*) new_word->pos,new_word->len,0,0);
217
cmp2= cmp ? 0 : (fabs(old_word->weight - new_word->weight) > 1.e-5);
222
_ma_ft_make_key(info, &key, keynr, keybuf, old_word, pos);
223
if ((error= _ma_ck_delete(info, &key)))
229
_ma_ft_make_key(info, &key, keynr, keybuf, new_word,pos);
230
if ((error= _ma_ck_write(info, &key)))
233
if (cmp<=0) old_word++;
234
if (cmp>=0) new_word++;
237
error= _ma_ft_erase(info,keynr,keybuf,old_word,pos);
238
else if (new_word->pos)
239
error= _ma_ft_store(info,keynr,keybuf,new_word,pos);
242
free_root(&info->ft_memroot, MYF(MY_MARK_BLOCKS_FREE));
247
/* adds a document to the collection */
249
int _ma_ft_add(MARIA_HA *info, uint keynr, uchar *keybuf, const uchar *record,
254
DBUG_ENTER("_ma_ft_add");
255
DBUG_PRINT("enter",("keynr: %d",keynr));
257
if ((wlist= _ma_ft_parserecord(info, keynr, record, &info->ft_memroot)))
258
error= _ma_ft_store(info,keynr,keybuf,wlist,pos);
259
free_root(&info->ft_memroot, MYF(MY_MARK_BLOCKS_FREE));
260
DBUG_PRINT("exit",("Return: %d",error));
265
/* removes a document from the collection */
267
int _ma_ft_del(MARIA_HA *info, uint keynr, uchar *keybuf, const uchar *record,
272
DBUG_ENTER("_ma_ft_del");
273
DBUG_PRINT("enter",("keynr: %d",keynr));
275
if ((wlist= _ma_ft_parserecord(info, keynr, record, &info->ft_memroot)))
276
error= _ma_ft_erase(info,keynr,keybuf,wlist,pos);
277
free_root(&info->ft_memroot, MYF(MY_MARK_BLOCKS_FREE));
278
DBUG_PRINT("exit",("Return: %d",error));
283
MARIA_KEY *_ma_ft_make_key(MARIA_HA *info, MARIA_KEY *key, uint keynr,
285
FT_WORD *wptr, my_off_t filepos)
287
uchar buf[HA_FT_MAXBYTELEN+16];
288
DBUG_ENTER("_ma_ft_make_key");
290
#if HA_FT_WTYPE == HA_KEYTYPE_FLOAT
292
float weight=(float) ((filepos==HA_OFFSET_ERROR) ? 0 : wptr->weight);
293
mi_float4store(buf,weight);
299
int2store(buf+HA_FT_WLEN,wptr->len);
300
memcpy(buf+HA_FT_WLEN+2,wptr->pos,wptr->len);
301
/* Can't be spatial so it's ok to call _ma_make_key directly here */
302
DBUG_RETURN(_ma_make_key(info, key, keynr, keybuf, buf, filepos, 0));
307
convert key value to ft2
310
uint _ma_ft_convert_to_ft2(MARIA_HA *info, MARIA_KEY *key)
312
MARIA_SHARE *share= info->s;
314
DYNAMIC_ARRAY *da=info->ft1_to_ft2;
315
MARIA_KEYDEF *keyinfo=&share->ft2_keyinfo;
316
uchar *key_ptr= (uchar*) dynamic_array_ptr(da, 0), *end;
317
uint length, key_length;
318
MARIA_PINNED_PAGE tmp_page_link, *page_link= &tmp_page_link;
320
DBUG_ENTER("_ma_ft_convert_to_ft2");
322
/* we'll generate one pageful at once, and insert the rest one-by-one */
323
/* calculating the length of this page ...*/
324
length=(keyinfo->block_length-2) / keyinfo->keylength;
325
set_if_smaller(length, da->elements);
326
length=length * keyinfo->keylength;
328
get_key_full_length_rdonly(key_length, key->data);
329
while (_ma_ck_delete(info, key) == 0)
333
_ma_ck_delete() will populate info->ft1_to_ft2 with deleted keys
337
/* creating pageful of keys */
338
bzero(info->buff, share->keypage_header);
339
_ma_store_keynr(share, info->buff, keyinfo->key_nr);
340
_ma_store_page_used(share, info->buff, length + share->keypage_header);
341
memcpy(info->buff + share->keypage_header, key_ptr, length);
342
info->keyread_buff_used= info->page_changed=1; /* info->buff is used */
344
@todo RECOVERY BUG this is not logged yet. Ok as this code is never
345
called, but soon it will be.
347
if ((root= _ma_new(info, DFLT_INIT_HITS, &page_link)) == HA_OFFSET_ERROR ||
348
_ma_write_keypage(info, keyinfo, root, page_link->write_lock,
349
DFLT_INIT_HITS, info->buff))
352
/* inserting the rest of key values */
353
end= (uchar*) dynamic_array_ptr(da, da->elements);
354
tmp_key.keyinfo= keyinfo;
355
tmp_key.data_length= keyinfo->keylength;
356
tmp_key.ref_length= 0;
358
for (key_ptr+=length; key_ptr < end; key_ptr+=keyinfo->keylength)
360
tmp_key.data= key_ptr;
361
if (_ma_ck_real_write_btree(info, key, &root, SEARCH_SAME))
365
/* now, writing the word key entry */
366
ft_intXstore(key->data + key_length, - (int) da->elements);
367
_ma_dpointer(share, key->data + key_length + HA_FT_WLEN, root);
369
DBUG_RETURN(_ma_ck_real_write_btree(info, key,
370
&share->state.key_root[key->keyinfo->