aboutsummaryrefslogtreecommitdiff
path: root/lib/dokuwiki/inc/indexer.php
blob: 14af579bbf168d5a75f2f6e6a915a5d03d8a1c3a (plain)
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
<?php
/**
 * Common DokuWiki functions
 *
 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
 * @author     Andreas Gohr <andi@splitbrain.org>
 */

if(!defined('DOKU_INC')) die('meh.');
require_once(DOKU_INC.'inc/io.php');
require_once(DOKU_INC.'inc/utf8.php');
require_once(DOKU_INC.'inc/parserutils.php');

// set the minimum token length to use in the index (note, this doesn't apply to numeric tokens)
if (!defined('IDX_MINWORDLENGTH')) define('IDX_MINWORDLENGTH',2);

// Asian characters are handled as words. The following regexp defines the
// Unicode-Ranges for Asian characters
// Ranges taken from http://en.wikipedia.org/wiki/Unicode_block
// I'm no language expert. If you think some ranges are wrongly chosen or
// a range is missing, please contact me
define('IDX_ASIAN1','[\x{0E00}-\x{0E7F}]'); // Thai
define('IDX_ASIAN2','['.
                   '\x{2E80}-\x{3040}'.  // CJK -> Hangul
                   '\x{309D}-\x{30A0}'.
                   '\x{30FD}-\x{31EF}\x{3200}-\x{D7AF}'.
                   '\x{F900}-\x{FAFF}'.  // CJK Compatibility Ideographs
                   '\x{FE30}-\x{FE4F}'.  // CJK Compatibility Forms
                   ']');
define('IDX_ASIAN3','['.                // Hiragana/Katakana (can be two characters)
                   '\x{3042}\x{3044}\x{3046}\x{3048}'.
                   '\x{304A}-\x{3062}\x{3064}-\x{3082}'.
                   '\x{3084}\x{3086}\x{3088}-\x{308D}'.
                   '\x{308F}-\x{3094}'.
                   '\x{30A2}\x{30A4}\x{30A6}\x{30A8}'.
                   '\x{30AA}-\x{30C2}\x{30C4}-\x{30E2}'.
                   '\x{30E4}\x{30E6}\x{30E8}-\x{30ED}'.
                   '\x{30EF}-\x{30F4}\x{30F7}-\x{30FA}'.
                   ']['.
                   '\x{3041}\x{3043}\x{3045}\x{3047}\x{3049}'.
                   '\x{3063}\x{3083}\x{3085}\x{3087}\x{308E}\x{3095}-\x{309C}'.
                   '\x{30A1}\x{30A3}\x{30A5}\x{30A7}\x{30A9}'.
                   '\x{30C3}\x{30E3}\x{30E5}\x{30E7}\x{30EE}\x{30F5}\x{30F6}\x{30FB}\x{30FC}'.
                   '\x{31F0}-\x{31FF}'.
                   ']?');
define('IDX_ASIAN', '(?:'.IDX_ASIAN1.'|'.IDX_ASIAN2.'|'.IDX_ASIAN3.')');

/**
 * Measure the length of a string.
 * Differs from strlen in handling of asian characters.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function wordlen($w){
    $l = strlen($w);
    // If left alone, all chinese "words" will get put into w3.idx
    // So the "length" of a "word" is faked
    if(preg_match('/'.IDX_ASIAN2.'/u',$w))
        $l += ord($w) - 0xE1;  // Lead bytes from 0xE2-0xEF
    return $l;
}

/**
 * Write a list of strings to an index file.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_saveIndex($pre, $wlen, &$idx){
    global $conf;
    $fn = $conf['indexdir'].'/'.$pre.$wlen;
    $fh = @fopen($fn.'.tmp','w');
    if(!$fh) return false;
    foreach ($idx as $line) {
        fwrite($fh,$line);
    }
    fclose($fh);
    if(isset($conf['fperm'])) chmod($fn.'.tmp', $conf['fperm']);
    io_rename($fn.'.tmp', $fn.'.idx');
    return true;
}

/**
 * Append a given line to an index file.
 *
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function idx_appendIndex($pre, $wlen, $line){
    global $conf;
    $fn = $conf['indexdir'].'/'.$pre.$wlen;
    $fh = @fopen($fn.'.idx','a');
    if(!$fh) return false;
    fwrite($fh,$line);
    fclose($fh);
    return true;
}

/**
 * Read the list of words in an index (if it exists).
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_getIndex($pre, $wlen){
    global $conf;
    $fn = $conf['indexdir'].'/'.$pre.$wlen.'.idx';
    if(!@file_exists($fn)) return array();
    return file($fn);
}

/**
 * Create an empty index file if it doesn't exist yet.
 *
 * FIXME: This function isn't currently used. It will probably be removed soon.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_touchIndex($pre, $wlen){
    global $conf;
    $fn = $conf['indexdir'].'/'.$pre.$wlen.'.idx';
    if(!@file_exists($fn)){
        touch($fn);
        if($conf['fperm']) chmod($fn, $conf['fperm']);
    }
}

/**
 * Read a line ending with \n.
 * Returns false on EOF.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function _freadline($fh) {
    if (feof($fh)) return false;
    $ln = '';
    while (($buf = fgets($fh,4096)) !== false) {
        $ln .= $buf;
        if (substr($buf,-1) == "\n") break;
    }
    if ($ln === '') return false;
    if (substr($ln,-1) != "\n") $ln .= "\n";
    return $ln;
}

/**
 * Write a line to an index file.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_saveIndexLine($pre, $wlen, $idx, $line){
    global $conf;
    if(substr($line,-1) != "\n") $line .= "\n";
    $fn = $conf['indexdir'].'/'.$pre.$wlen;
    $fh = @fopen($fn.'.tmp','w');
    if(!$fh) return false;
    $ih = @fopen($fn.'.idx','r');
    if ($ih) {
        $ln = -1;
        while (($curline = _freadline($ih)) !== false) {
            if (++$ln == $idx) {
                fwrite($fh, $line);
            } else {
                fwrite($fh, $curline);
            }
        }
        if ($idx > $ln) {
            fwrite($fh,$line);
        }
        fclose($ih);
    } else {
        fwrite($fh,$line);
    }
    fclose($fh);
    if($conf['fperm']) chmod($fn.'.tmp', $conf['fperm']);
    io_rename($fn.'.tmp', $fn.'.idx');
    return true;
}

/**
 * Read a single line from an index (if it exists).
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_getIndexLine($pre, $wlen, $idx){
    global $conf;
    $fn = $conf['indexdir'].'/'.$pre.$wlen.'.idx';
    if(!@file_exists($fn)) return '';
    $fh = @fopen($fn,'r');
    if(!$fh) return '';
    $ln = -1;
    while (($line = _freadline($fh)) !== false) {
        if (++$ln == $idx) break;
    }
    fclose($fh);
    return "$line";
}

/**
 * Split a page into words
 *
 * Returns an array of word counts, false if an error occurred.
 * Array is keyed on the word length, then the word index.
 *
 * @author Andreas Gohr <andi@splitbrain.org>
 * @author Christopher Smith <chris@jalakai.co.uk>
 */
function idx_getPageWords($page){
    global $conf;
    $swfile   = DOKU_INC.'inc/lang/'.$conf['lang'].'/stopwords.txt';
    if(@file_exists($swfile)){
        $stopwords = file($swfile);
    }else{
        $stopwords = array();
    }

    $body = '';
    $data = array($page, $body);
    $evt = new Doku_Event('INDEXER_PAGE_ADD', $data);
    if ($evt->advise_before()) $data[1] .= rawWiki($page);
    $evt->advise_after();
    unset($evt);

    list($page,$body) = $data;

    $body   = strtr($body, "\r\n\t", '   ');
    $tokens = explode(' ', $body);
    $tokens = array_count_values($tokens);   // count the frequency of each token

    // ensure the deaccented or romanised page names of internal links are added to the token array
    // (this is necessary for the backlink function -- there maybe a better way!)
    if ($conf['deaccent']) {
        $links = p_get_metadata($page,'relation references');

        if (!empty($links)) {
            $tmp = join(' ',array_keys($links));                // make a single string
            $tmp = strtr($tmp, ':', ' ');                       // replace namespace separator with a space
            $link_tokens = array_unique(explode(' ', $tmp));    // break into tokens

            foreach ($link_tokens as $link_token) {
                if (isset($tokens[$link_token])) continue;
                $tokens[$link_token] = 1;
            }
        }
    }

    $words = array();
    foreach ($tokens as $word => $count) {
        $arr = idx_tokenizer($word,$stopwords);
        $arr = array_count_values($arr);
        foreach ($arr as $w => $c) {
            $l = wordlen($w);
            if(isset($words[$l])){
                $words[$l][$w] = $c * $count + (isset($words[$l][$w]) ? $words[$l][$w] : 0);
            }else{
                $words[$l] = array($w => $c * $count);
            }
        }
    }

    // arrive here with $words = array(wordlen => array(word => frequency))

    $index = array(); //resulting index
    foreach (array_keys($words) as $wlen){
        $word_idx = idx_getIndex('w',$wlen);
        foreach ($words[$wlen] as $word => $freq) {
            $wid = array_search("$word\n",$word_idx);
            if(!is_int($wid)){
                $wid = count($word_idx);
                $word_idx[] = "$word\n";
            }
            if(!isset($index[$wlen]))
                $index[$wlen] = array();
            $index[$wlen][$wid] = $freq;
        }

        // save back word index
        if(!idx_saveIndex('w',$wlen,$word_idx)){
            trigger_error("Failed to write word index", E_USER_ERROR);
            return false;
        }
    }

    return $index;
}

/**
 * Adds/updates the search for the given page
 *
 * This is the core function of the indexer which does most
 * of the work. This function needs to be called with proper
 * locking!
 *
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function idx_addPage($page){
    global $conf;

    // load known documents
    $page_idx = idx_getIndex('page','');

    // get page id (this is the linenumber in page.idx)
    $pid = array_search("$page\n",$page_idx);
    if(!is_int($pid)){
        $pid = count($page_idx);
        // page was new - write back
        if (!idx_appendIndex('page','',"$page\n")){
            trigger_error("Failed to write page index", E_USER_ERROR);
            return false;
        }
    }
    unset($page_idx); // free memory

    $pagewords = array();
    // get word usage in page
    $words = idx_getPageWords($page);
    if($words === false) return false;

    if(!empty($words)) {
        foreach(array_keys($words) as $wlen){
            $index = idx_getIndex('i',$wlen);
            foreach($words[$wlen] as $wid => $freq){
                if($wid<count($index)){
                    $index[$wid] = idx_updateIndexLine($index[$wid],$pid,$freq);
                }else{
                    // New words **should** have been added in increasing order
                    // starting with the first unassigned index.
                    // If someone can show how this isn't true, then I'll need to sort
                    // or do something special.
                    $index[$wid] = idx_updateIndexLine('',$pid,$freq);
                }
                $pagewords[] = "$wlen*$wid";
            }
            // save back word index
            if(!idx_saveIndex('i',$wlen,$index)){
                trigger_error("Failed to write index", E_USER_ERROR);
                return false;
            }
        }
    }

    // Remove obsolete index entries
    $pageword_idx = trim(idx_getIndexLine('pageword','',$pid));
    if ($pageword_idx !== '') {
        $oldwords = explode(':',$pageword_idx);
        $delwords = array_diff($oldwords, $pagewords);
        $upwords = array();
        foreach ($delwords as $word) {
            if($word=='') continue;
            list($wlen,$wid) = explode('*',$word);
            $wid = (int)$wid;
            $upwords[$wlen][] = $wid;
        }
        foreach ($upwords as $wlen => $widx) {
            $index = idx_getIndex('i',$wlen);
            foreach ($widx as $wid) {
                $index[$wid] = idx_updateIndexLine($index[$wid],$pid,0);
            }
            idx_saveIndex('i',$wlen,$index);
        }
    }
    // Save the reverse index
    $pageword_idx = join(':',$pagewords)."\n";
    if(!idx_saveIndexLine('pageword','',$pid,$pageword_idx)){
        trigger_error("Failed to write word index", E_USER_ERROR);
        return false;
    }

    return true;
}

/**
 * Write a new index line to the filehandle
 *
 * This function writes an line for the index file to the
 * given filehandle. It removes the given document from
 * the given line and readds it when $count is >0.
 *
 * @deprecated - see idx_updateIndexLine
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function idx_writeIndexLine($fh,$line,$pid,$count){
    fwrite($fh,idx_updateIndexLine($line,$pid,$count));
}

/**
 * Modify an index line with new information
 *
 * This returns a line of the index. It removes the
 * given document from the line and readds it if
 * $count is >0.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function idx_updateIndexLine($line,$pid,$count){
    $line = trim($line);
    $updated = array();
    if($line != ''){
        $parts = explode(':',$line);
        // remove doc from given line
        foreach($parts as $part){
            if($part == '') continue;
            list($doc,$cnt) = explode('*',$part);
            if($doc != $pid){
                $updated[] = $part;
            }
        }
    }

    // add doc
    if ($count){
        $updated[] = "$pid*$count";
    }

    return join(':',$updated)."\n";
}

/**
 * Get the word lengths that have been indexed.
 *
 * Reads the index directory and returns an array of lengths
 * that there are indices for.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_indexLengths(&$filter){
    global $conf;
    $dir = @opendir($conf['indexdir']);
    if($dir===false)
        return array();
    $idx = array();
    if(is_array($filter)){
        while (($f = readdir($dir)) !== false) {
            if (substr($f,0,1) == 'i' && substr($f,-4) == '.idx'){
                $i = substr($f,1,-4);
                if (is_numeric($i) && isset($filter[(int)$i]))
                    $idx[] = (int)$i;
            }
        }
    }else{
        // Exact match first.
        if(@file_exists($conf['indexdir']."/i$filter.idx"))
            $idx[] = $filter;
        while (($f = readdir($dir)) !== false) {
            if (substr($f,0,1) == 'i' && substr($f,-4) == '.idx'){
                $i = substr($f,1,-4);
                if (is_numeric($i) && $i > $filter)
                    $idx[] = (int)$i;
            }
        }
    }
    closedir($dir);
    return $idx;
}

/**
 * Find the the index number of each search term.
 *
 * This will group together words that appear in the same index.
 * So it should perform better, because it only opens each index once.
 * Actually, it's not that great. (in my experience) Probably because of the disk cache.
 * And the sorted function does more work, making it slightly slower in some cases.
 *
 * @param array    $words   The query terms. Words should only contain valid characters,
 *                          with a '*' at either the beginning or end of the word (or both)
 * @param arrayref $result  Set to word => array("length*id" ...), use this to merge the
 *                          index locations with the appropriate query term.
 * @return array            Set to length => array(id ...)
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_getIndexWordsSorted($words,&$result){
    // parse and sort tokens
    $tokens = array();
    $tokenlength = array();
    $tokenwild = array();
    foreach($words as $word){
        $result[$word] = array();
        $wild = 0;
        $xword = $word;
        $wlen = wordlen($word);

        // check for wildcards
        if(substr($xword,0,1) == '*'){
            $xword = substr($xword,1);
            $wild |= 1;
            $wlen -= 1;
        }
        if(substr($xword,-1,1) == '*'){
            $xword = substr($xword,0,-1);
            $wild |= 2;
            $wlen -= 1;
        }
        if ($wlen < IDX_MINWORDLENGTH && $wild == 0 && !is_numeric($xword)) continue;
        if(!isset($tokens[$xword])){
            $tokenlength[$wlen][] = $xword;
        }
        if($wild){
            $ptn = preg_quote($xword,'/');
            if(($wild&1) == 0) $ptn = '^'.$ptn;
            if(($wild&2) == 0) $ptn = $ptn.'$';
            $tokens[$xword][] = array($word, '/'.$ptn.'/');
            if(!isset($tokenwild[$xword])) $tokenwild[$xword] = $wlen;
        }else
            $tokens[$xword][] = array($word, null);
    }
    asort($tokenwild);
    // $tokens = array( base word => array( [ query word , grep pattern ] ... ) ... )
    // $tokenlength = array( base word length => base word ... )
    // $tokenwild = array( base word => base word length ... )

    $length_filter = empty($tokenwild) ? $tokenlength : min(array_keys($tokenlength));
    $indexes_known = idx_indexLengths($length_filter);
    if(!empty($tokenwild)) sort($indexes_known);
    // get word IDs
    $wids = array();
    foreach($indexes_known as $ixlen){
        $word_idx = idx_getIndex('w',$ixlen);
        // handle exact search
        if(isset($tokenlength[$ixlen])){
            foreach($tokenlength[$ixlen] as $xword){
                $wid = array_search("$xword\n",$word_idx);
                if(is_int($wid)){
                    $wids[$ixlen][] = $wid;
                    foreach($tokens[$xword] as $w)
                        $result[$w[0]][] = "$ixlen*$wid";
                }
            }
        }
        // handle wildcard search
        foreach($tokenwild as $xword => $wlen){
            if($wlen >= $ixlen) break;
            foreach($tokens[$xword] as $w){
                if(is_null($w[1])) continue;
                foreach(array_keys(preg_grep($w[1],$word_idx)) as $wid){
                    $wids[$ixlen][] = $wid;
                    $result[$w[0]][] = "$ixlen*$wid";
                }
            }
        }
    }
    return $wids;
}

/**
 * Lookup words in index
 *
 * Takes an array of word and will return a list of matching
 * documents for each one.
 *
 * Important: No ACL checking is done here! All results are
 *            returned, regardless of permissions
 *
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function idx_lookup($words){
    global $conf;

    $result = array();

    $wids = idx_getIndexWordsSorted($words, $result);
    if(empty($wids)) return array();

    // load known words and documents
    $page_idx = idx_getIndex('page','');

    $docs = array();                          // hold docs found
    foreach(array_keys($wids) as $wlen){
        $wids[$wlen] = array_unique($wids[$wlen]);
        $index = idx_getIndex('i',$wlen);
        foreach($wids[$wlen] as $ixid){
            if($ixid < count($index))
                $docs["$wlen*$ixid"] = idx_parseIndexLine($page_idx,$index[$ixid]);
        }
    }

    // merge found pages into final result array
    $final = array();
    foreach($result as $word => $res){
        $final[$word] = array();
        foreach($res as $wid){
            $hits = &$docs[$wid];
            foreach ($hits as $hitkey => $hitcnt) {
                if (!isset($final[$word][$hitkey])) {
                    $final[$word][$hitkey] = $hitcnt;
                } else {
                    $final[$word][$hitkey] += $hitcnt;
                }
            }
        }
    }
    return $final;
}

/**
 * Returns a list of documents and counts from a index line
 *
 * It omits docs with a count of 0 and pages that no longer
 * exist.
 *
 * @param  array  $page_idx The list of known pages
 * @param  string $line     A line from the main index
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function idx_parseIndexLine(&$page_idx,$line){
    $result = array();

    $line = trim($line);
    if($line == '') return $result;

    $parts = explode(':',$line);
    foreach($parts as $part){
        if($part == '') continue;
        list($doc,$cnt) = explode('*',$part);
        if(!$cnt) continue;
        $doc = trim($page_idx[$doc]);
        if(!$doc) continue;
        // make sure the document still exists
        if(!page_exists($doc,'',false)) continue;

        $result[$doc] = $cnt;
    }
    return $result;
}

/**
 * Tokenizes a string into an array of search words
 *
 * Uses the same algorithm as idx_getPageWords()
 *
 * @param string   $string     the query as given by the user
 * @param arrayref $stopwords  array of stopwords
 * @param boolean  $wc         are wildcards allowed?
 */
function idx_tokenizer($string,&$stopwords,$wc=false){
    $words = array();
    $wc = ($wc) ? '' : $wc = '\*';

    if(preg_match('/[^0-9A-Za-z]/u', $string)){
        // handle asian chars as single words (may fail on older PHP version)
        $asia = @preg_replace('/('.IDX_ASIAN.')/u',' \1 ',$string);
        if(!is_null($asia)) $string = $asia; //recover from regexp failure

        $arr = explode(' ', utf8_stripspecials($string,' ','\._\-:'.$wc));
        foreach ($arr as $w) {
            if (!is_numeric($w) && strlen($w) < IDX_MINWORDLENGTH) continue;
            $w = utf8_strtolower($w);
            if($stopwords && is_int(array_search("$w\n",$stopwords))) continue;
            $words[] = $w;
        }
    }else{
        $w = $string;
        if (!is_numeric($w) && strlen($w) < IDX_MINWORDLENGTH) return $words;
        $w = strtolower($w);
        if(is_int(array_search("$w\n",$stopwords))) return $words;
        $words[] = $w;
    }

    return $words;
}

/**
 * Create a pagewords index from the existing index.
 *
 * @author Tom N Harris <tnharris@whoopdedo.org>
 */
function idx_upgradePageWords(){
    global $conf;
    $page_idx = idx_getIndex('page','');
    if (empty($page_idx)) return;
    $pagewords = array();
    $len = count($page_idx);
    for ($n=0;$n<$len;$n++){
        $pagewords[] = array();
    }
    unset($page_idx);

    $n=0;
    foreach (idx_indexLengths($n) as $wlen) {
        $lines = idx_getIndex('i',$wlen);
        $len = count($lines);
        for ($wid=0;$wid<$len;$wid++) {
            $wkey = "$wlen*$wid";
            foreach (explode(':',trim($lines[$wid])) as $part) {
                if($part == '') continue;
                list($doc,$cnt) = explode('*',$part);
                $pagewords[(int)$doc][] = $wkey;
            }
        }
    }

    $fn = $conf['indexdir'].'/pageword';
    $fh = @fopen($fn.'.tmp','w');
    if (!$fh){
        trigger_error("Failed to write word index", E_USER_ERROR);
        return false;
    }
    foreach ($pagewords as $line){
        fwrite($fh, join(':',$line)."\n");
    }
    fclose($fh);
    if($conf['fperm']) chmod($fn.'.tmp', $conf['fperm']);
    io_rename($fn.'.tmp', $fn.'.idx');
    return true;
}

//Setup VIM: ex: et ts=4 enc=utf-8 :