Source

gd-websites / wiki / inc / fulltext.php

Full commit
  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
<?php
/**
 * DokuWiki fulltextsearch functions using the index
 *
 * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
 * @author     Andreas Gohr <andi@splitbrain.org>
 */

  if(!defined('DOKU_INC')) define('DOKU_INC',fullpath(dirname(__FILE__).'/../').'/');
  require_once(DOKU_INC.'inc/indexer.php');


/**
 * The fulltext search
 *
 * Returns a list of matching documents for the given query
 *
 */
function ft_pageSearch($query,&$highlight){
    $q = ft_queryParser($query);

    $highlight = array();

    // remember for hilighting later
    foreach($q['words'] as $wrd){
        $highlight[] =  str_replace('*','',$wrd);
    }

    // lookup all words found in the query
    $words  = array_merge($q['and'],$q['not']);
    if(!count($words)) return array();
    $result = idx_lookup($words);
    if(!count($result)) return array();

    // merge search results with query
    foreach($q['and'] as $pos => $w){
        $q['and'][$pos] = $result[$w];
    }
    // create a list of unwanted docs
    $not = array();
    foreach($q['not'] as $pos => $w){
        $not = array_merge($not,array_keys($result[$w]));
    }

    // combine and-words
    if(count($q['and']) > 1){
        $docs = ft_resultCombine($q['and']);
    }else{
        $docs = $q['and'][0];
    }
    if(!count($docs)) return array();

    // create a list of hidden pages in the result
    $hidden = array();
    $hidden = array_filter(array_keys($docs),'isHiddenPage');
    $not = array_merge($not,$hidden);

    // filter unmatched namespaces
    if(!empty($q['ns'])) {
        $pattern = implode('|^',$q['ns']);
        foreach($docs as $key => $val) {
            if(!preg_match('/^'.$pattern.'/',$key)) {
                unset($docs[$key]);
            }
        }
    }

    // remove negative matches
    foreach($not as $n){
        unset($docs[$n]);
    }

    if(!count($docs)) return array();
    // handle phrases
    if(count($q['phrases'])){
        $q['phrases'] = array_map('utf8_strtolower',$q['phrases']);
        // use this for higlighting later:
        $highlight = array_merge($highlight,$q['phrases']);
        $q['phrases'] = array_map('preg_quote_cb',$q['phrases']);
        // check the source of all documents for the exact phrases
        foreach(array_keys($docs) as $id){
            $text  = utf8_strtolower(rawWiki($id));
            foreach($q['phrases'] as $phrase){
                if(!preg_match('/'.$phrase.'/usi',$text)){
                    unset($docs[$id]); // no hit - remove
                    break;
                }
            }
        }
    }

    if(!count($docs)) return array();

    // check ACL permissions
    foreach(array_keys($docs) as $doc){
        if(auth_quickaclcheck($doc) < AUTH_READ){
            unset($docs[$doc]);
        }
    }

    if(!count($docs)) return array();

    // if there are any hits left, sort them by count
    arsort($docs);

    return $docs;
}

/**
 * Returns the backlinks for a given page
 *
 * Does a quick lookup with the fulltext index, then
 * evaluates the instructions of the found pages
 */
function ft_backlinks($id){
    global $conf;
    $swfile   = DOKU_INC.'inc/lang/'.$conf['lang'].'/stopwords.txt';
    $stopwords = @file_exists($swfile) ? file($swfile) : array();

    $result = array();

    // quick lookup of the pagename
    $page    = noNS($id);
    $matches = idx_lookup(idx_tokenizer($page,$stopwords));  // pagename may contain specials (_ or .)
    $docs    = array_keys(ft_resultCombine(array_values($matches)));
    $docs    = array_filter($docs,'isVisiblePage'); // discard hidden pages
    if(!count($docs)) return $result;
    require_once(DOKU_INC.'inc/parserutils.php');

    // check metadata for matching links
    foreach($docs as $match){
        // metadata relation reference links are already resolved
        $links = p_get_metadata($match,'relation references');
        if (isset($links[$id])) $result[] = $match;
    }

    if(!count($result)) return $result;

    // check ACL permissions
    foreach(array_keys($result) as $idx){
        if(auth_quickaclcheck($result[$idx]) < AUTH_READ){
            unset($result[$idx]);
        }
    }

    sort($result);
    return $result;
}

/**
 * Returns the pages that use a given media file
 *
 * Does a quick lookup with the fulltext index, then
 * evaluates the instructions of the found pages
 *
 * Aborts after $max found results
 */
function ft_mediause($id,$max){
    global $conf;
    $swfile   = DOKU_INC.'inc/lang/'.$conf['lang'].'/stopwords.txt';
    $stopwords = @file_exists($swfile) ? file($swfile) : array();

    if(!$max) $max = 1; // need to find at least one

    $result = array();

    // quick lookup of the mediafile
    $media   = noNS($id);
    $matches = idx_lookup(idx_tokenizer($media,$stopwords));
    $docs    = array_keys(ft_resultCombine(array_values($matches)));
    if(!count($docs)) return $result;

    // go through all found pages
    $found = 0;
    $pcre  = preg_quote($media,'/');
    foreach($docs as $doc){
        $ns = getNS($doc);
        preg_match_all('/\{\{([^|}]*'.$pcre.'[^|}]*)(|[^}]+)?\}\}/i',rawWiki($doc),$matches);
        foreach($matches[1] as $img){
            $img = trim($img);
            if(preg_match('/^https?:\/\//i',$img)) continue; // skip external images
            list($img) = explode('?',$img);                  // remove any parameters
            resolve_mediaid($ns,$img,$exists);               // resolve the possibly relative img

            if($img == $id){                                 // we have a match
                $result[] = $doc;
                $found++;
                break;
            }
        }
        if($found >= $max) break;
    }

    sort($result);
    return $result;
}



/**
 * Quicksearch for pagenames
 *
 * By default it only matches the pagename and ignores the
 * namespace. This can be changed with the second parameter
 *
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function ft_pageLookup($id,$pageonly=true){
    global $conf;
    $id    = preg_quote($id,'/');
    $pages = file($conf['indexdir'].'/page.idx');
    if($id) $pages = array_values(preg_grep('/'.$id.'/',$pages));

    $cnt = count($pages);
    for($i=0; $i<$cnt; $i++){
        if($pageonly){
            if(!preg_match('/'.$id.'/',noNS($pages[$i]))){
                unset($pages[$i]);
                continue;
            }
        }
        if(!page_exists($pages[$i])){
            unset($pages[$i]);
            continue;
        }
    }

    $pages = array_filter($pages,'isVisiblePage'); // discard hidden pages
    if(!count($pages)) return array();

    // check ACL permissions
    foreach(array_keys($pages) as $idx){
        if(auth_quickaclcheck($pages[$idx]) < AUTH_READ){
            unset($pages[$idx]);
        }
    }

    $pages = array_map('trim',$pages);
    sort($pages);
    return $pages;
}

/**
 * Creates a snippet extract
 *
 * @author Andreas Gohr <andi@splitbrain.org>
 */
function ft_snippet($id,$highlight){
    $text     = rawWiki($id);
    $match = array();
    $snippets = array();
    $utf8_offset = $offset = $end = 0;
    $len = utf8_strlen($text);

    // build a regexp from the phrases to highlight
    $re = join('|',array_map('preg_quote_cb',array_filter((array) $highlight)));

    for ($cnt=3; $cnt--;) {
      if (!preg_match('#('.$re.')#iu',$text,$match,PREG_OFFSET_CAPTURE,$offset)) break;

      list($str,$idx) = $match[0];

      // convert $idx (a byte offset) into a utf8 character offset
      $utf8_idx = utf8_strlen(substr($text,0,$idx));
      $utf8_len = utf8_strlen($str);

      // establish context, 100 bytes surrounding the match string
      // first look to see if we can go 100 either side,
      // then drop to 50 adding any excess if the other side can't go to 50,
      $pre = min($utf8_idx-$utf8_offset,100);
      $post = min($len-$utf8_idx-$utf8_len,100);

      if ($pre>50 && $post>50) {
        $pre = $post = 50;
      } else if ($pre>50) {
        $pre = min($pre,100-$post);
      } else if ($post>50) {
        $post = min($post, 100-$pre);
      } else {
        // both are less than 50, means the context is the whole string
        // make it so and break out of this loop - there is no need for the
        // complex snippet calculations
        $snippets = array($text);
        break;
      }

      // establish context start and end points, try to append to previous
      // context if possible
      $start = $utf8_idx - $pre;
      $append = ($start < $end) ? $end : false;  // still the end of the previous context snippet
      $end = $utf8_idx + $utf8_len + $post;      // now set it to the end of this context

      if ($append) {
        $snippets[count($snippets)-1] .= utf8_substr($text,$append,$end-$append);
      } else {
        $snippets[] = utf8_substr($text,$start,$end-$start);
      }

      // set $offset for next match attempt
      //   substract strlen to avoid splitting a potential search success,
      //   this is an approximation as the search pattern may match strings
      //   of varying length and it will fail if the context snippet
      //   boundary breaks a matching string longer than the current match
      $utf8_offset = $utf8_idx + $post;
      $offset = $idx + strlen(utf8_substr($text,$utf8_idx,$post));
      $offset = utf8_correctIdx($text,$offset);
    }

    $m = "\1";
    $snippets = preg_replace('#('.$re.')#iu',$m.'$1'.$m,$snippets);
    $snippet = preg_replace('#'.$m.'([^'.$m.']*?)'.$m.'#iu','<strong class="search_hit">$1</strong>',hsc(join('... ',$snippets)));

    return $snippet;
}

/**
 * Combine found documents and sum up their scores
 *
 * This function is used to combine searched words with a logical
 * AND. Only documents available in all arrays are returned.
 *
 * based upon PEAR's PHP_Compat function for array_intersect_key()
 *
 * @param array $args An array of page arrays
 */
function ft_resultCombine($args){
    $array_count = count($args);
    if($array_count == 1){
        return $args[0];
    }

    $result = array();
    if ($array_count > 1) {
      foreach ($args[0] as $key => $value) {
        $result[$key] = $value;
        for ($i = 1; $i !== $array_count; $i++) {
            if (!isset($args[$i][$key])) {
                unset($result[$key]);
                break;
            }
            $result[$key] += $args[$i][$key];
        }
      }
    }
    return $result;
}

/**
 * Builds an array of search words from a query
 *
 * @todo support OR and parenthesises?
 */
function ft_queryParser($query){
    global $conf;
    $swfile   = DOKU_INC.'inc/lang/'.$conf['lang'].'/stopwords.txt';
    if(@file_exists($swfile)){
        $stopwords = file($swfile);
    }else{
        $stopwords = array();
    }

    $q = array();
    $q['query']   = $query;
    $q['ns']      = array();
    $q['phrases'] = array();
    $q['words']   = array();
    $q['and']     = array();
    $q['not']     = array();

    // strip namespace from query
    if(preg_match('/([^@]*)@(.*)/',$query,$match))  {
        $query = $match[1];
        $q['ns'] = explode('@',preg_replace("/ /",'',$match[2]));
    }

    // handle phrase searches
    while(preg_match('/"(.*?)"/',$query,$match)){
        $q['phrases'][] = $match[1];
        $q['and'] = array_merge($q['and'], idx_tokenizer($match[0],$stopwords));
        $query = preg_replace('/"(.*?)"/','',$query,1);
    }

    $words = explode(' ',$query);
    foreach($words as $w){
        if($w{0} == '-'){
            $token = idx_tokenizer($w,$stopwords,true);
            if(count($token)) $q['not'] = array_merge($q['not'],$token);
        }else{
            // asian "words" need to be searched as phrases
            if(@preg_match_all('/(('.IDX_ASIAN.')+)/u',$w,$matches)){
                $q['phrases'] = array_merge($q['phrases'],$matches[1]);

            }
            $token = idx_tokenizer($w,$stopwords,true);
            if(count($token)){
                $q['and']   = array_merge($q['and'],$token);
                $q['words'] = array_merge($q['words'],$token);
            }
        }
    }

    return $q;
}

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