inc/parserutils.php

   1 <?php
   2 /**
   3  * Utilities for accessing the parser
   4  *
   5  * @license    GPL 2 (http://www.gnu.org/licenses/gpl.html)
   6  * @author     Harry Fuecks <hfuecks@gmail.com>
   7  * @author     Andreas Gohr <andi@splitbrain.org>
   8  */
   9
  10 if(!defined('DOKU_INC')) die('meh.');
  11 require_once(DOKU_INC.'inc/confutils.php');
  12 require_once(DOKU_INC.'inc/pageutils.php');
  13 require_once(DOKU_INC.'inc/pluginutils.php');
  14 require_once(DOKU_INC.'inc/cache.php');
  15
  16 /**
  17  * Returns the parsed Wikitext in XHTML for the given id and revision.
  18  *
  19  * If $excuse is true an explanation is returned if the file
  20  * wasn't found
  21  *
  22  * @author Andreas Gohr <andi@splitbrain.org>
  23  */
  24 function p_wiki_xhtml($id, $rev='', $excuse=true){
  25   $file = wikiFN($id,$rev);
  26   $ret  = '';
  27
  28   //ensure $id is in global $ID (needed for parsing)
  29   global $ID;
  30   $keep = $ID;
  31   $ID   = $id;
  32
  33   if($rev){
  34     if(@file_exists($file)){
  35       $ret = p_render('xhtml',p_get_instructions(io_readWikiPage($file,$id,$rev)),$info); //no caching on old revisions
  36     }elseif($excuse){
  37       $ret = p_locale_xhtml('norev');
  38     }
  39   }else{
  40     if(@file_exists($file)){
  41       $ret = p_cached_output($file,'xhtml',$id);
  42     }elseif($excuse){
  43       $ret = p_locale_xhtml('newpage');
  44     }
  45   }
  46
  47   //restore ID (just in case)
  48   $ID = $keep;
  49
  50   return $ret;
  51 }
  52
  53 /**
  54  * Returns starting summary for a page (e.g. the first few
  55  * paragraphs), marked up in XHTML.
  56  *
  57  * If $excuse is true an explanation is returned if the file
  58  * wasn't found
  59  *
  60  * @param string wiki page id
  61  * @param reference populated with page title from heading or page id
  62  * @deprecated
  63  * @author Harry Fuecks <hfuecks@gmail.com>
  64  */
  65 function p_wiki_xhtml_summary($id, &$title, $rev='', $excuse=true){
  66   $file = wikiFN($id,$rev);
  67   $ret  = '';
  68
  69   //ensure $id is in global $ID (needed for parsing)
  70   global $ID;
  71   $keep = $ID;
  72   $ID   = $id;
  73
  74   if($rev){
  75     if(@file_exists($file)){
  76       //no caching on old revisions
  77       $ins = p_get_instructions(io_readWikiPage($file,$id,$rev));
  78     }elseif($excuse){
  79       $ret = p_locale_xhtml('norev');
  80       //restore ID (just in case)
  81       $ID = $keep;
  82       return $ret;
  83     }
  84
  85   }else{
  86
  87     if(@file_exists($file)){
  88       // The XHTML for a summary is not cached so use the instruction cache
  89       $ins = p_cached_instructions($file);
  90     }elseif($excuse){
  91       $ret = p_locale_xhtml('newpage');
  92       //restore ID (just in case)
  93       $ID = $keep;
  94       return $ret;
  95     }
  96   }
  97
  98   $ret = p_render('xhtmlsummary',$ins,$info);
  99
 100   if ( $info['sum_pagetitle'] ) {
 101     $title = $info['sum_pagetitle'];
 102   } else {
 103     $title = $id;
 104   }
 105
 106   $ID = $keep;
 107   return $ret;
 108 }
 109
 110 /**
 111  * Returns the specified local text in parsed format
 112  *
 113  * @author Andreas Gohr <andi@splitbrain.org>
 114  */
 115 function p_locale_xhtml($id){
 116   //fetch parsed locale
 117   $html = p_cached_output(localeFN($id));
 118   return $html;
 119 }
 120
 121 /**
 122  *     *** DEPRECATED ***
 123  *
 124  * use p_cached_output()
 125  *
 126  * Returns the given file parsed to XHTML
 127  *
 128  * Uses and creates a cachefile
 129  *
 130  * @deprecated
 131  * @author Andreas Gohr <andi@splitbrain.org>
 132  * @todo   rewrite to use mode instead of hardcoded XHTML
 133  */
 134 function p_cached_xhtml($file){
 135   return p_cached_output($file);
 136 }
 137
 138 /**
 139  * Returns the given file parsed into the requested output format
 140  *
 141  * @author Andreas Gohr <andi@splitbrain.org>
 142  * @author Chris Smith <chris@jalakai.co.uk>
 143  */
 144 function p_cached_output($file, $format='xhtml', $id='') {
 145   global $conf;
 146
 147   $cache = new cache_renderer($id, $file, $format);
 148   if ($cache->useCache()) {
 149     $parsed = $cache->retrieveCache(false);
 150     if($conf['allowdebug'] && $format=='xhtml') $parsed .= "\n<!-- cachefile {$cache->cache} used -->\n";
 151   } else {
 152     $parsed = p_render($format, p_cached_instructions($file,false,$id), $info);
 153
 154     if ($info['cache']) {
 155       $cache->storeCache($parsed);               //save cachefile
 156       if($conf['allowdebug'] && $format=='xhtml') $parsed .= "\n<!-- no cachefile used, but created {$cache->cache} -->\n";
 157     }else{
 158       $cache->removeCache();                     //try to delete cachefile
 159       if($conf['allowdebug'] && $format=='xhtml') $parsed .= "\n<!-- no cachefile used, caching forbidden -->\n";
 160     }
 161   }
 162
 163   return $parsed;
 164 }
 165
 166 /**
 167  * Returns the render instructions for a file
 168  *
 169  * Uses and creates a serialized cache file
 170  *
 171  * @author Andreas Gohr <andi@splitbrain.org>
 172  */
 173 function p_cached_instructions($file,$cacheonly=false,$id='') {
 174   global $conf;
 175   static $run = null;
 176   if(is_null($run)) $run = array();
 177
 178   $cache = new cache_instructions($id, $file);
 179
 180   if ($cacheonly || $cache->useCache() || isset($run[$file])) {
 181     return $cache->retrieveCache();
 182   } else if (@file_exists($file)) {
 183     // no cache - do some work
 184     $ins = p_get_instructions(io_readWikiPage($file,$id));
 185     if ($cache->storeCache($ins)) {
 186       $run[$file] = true; // we won't rebuild these instructions in the same run again
 187     } else {
 188       msg('Unable to save cache file. Hint: disk full; file permissions; safe_mode setting.',-1);
 189     }
 190     return $ins;
 191   }
 192
 193   return null;
 194 }
 195
 196 /**
 197  * turns a page into a list of instructions
 198  *
 199  * @author Harry Fuecks <hfuecks@gmail.com>
 200  * @author Andreas Gohr <andi@splitbrain.org>
 201  */
 202 function p_get_instructions($text){
 203
 204   $modes = p_get_parsermodes();
 205
 206   // Create the parser
 207   $Parser = & new Doku_Parser();
 208
 209   // Add the Handler
 210   $Parser->Handler = & new Doku_Handler();
 211
 212   //add modes to parser
 213   foreach($modes as $mode){
 214     $Parser->addMode($mode['mode'],$mode['obj']);
 215   }
 216
 217   // Do the parsing
 218   trigger_event('PARSER_WIKITEXT_PREPROCESS', $text);
 219   $p = $Parser->parse($text);
 220 //  dbg($p);
 221   return $p;
 222 }
 223
 224 /**
 225  * returns the metadata of a page
 226  *
 227  * @author Esther Brunner <esther@kaffeehaus.ch>
 228  */
 229 function p_get_metadata($id, $key=false, $render=false){
 230   global $ID, $INFO, $cache_metadata;
 231
 232   // cache the current page
 233   // Benchmarking shows the current page's metadata is generally the only page metadata
 234   // accessed several times. This may catch a few other pages, but that shouldn't be an issue.
 235   $cache = ($ID == $id);
 236   $meta = p_read_metadata($id, $cache);
 237
 238   // metadata has never been rendered before - do it! (but not for non-existent pages)
 239   if ($render && !$meta['current']['description']['abstract'] && page_exists($id)){
 240     $meta = p_render_metadata($id, $meta);
 241     io_saveFile(metaFN($id, '.meta'), serialize($meta));
 242
 243     // sync cached copies, including $INFO metadata
 244     if (!empty($cache_metadata[$id])) $cache_metadata[$id] = $meta;
 245     if (!empty($INFO) && ($id == $INFO['id'])) { $INFO['meta'] = $meta['current']; }
 246   }
 247
 248   // filter by $key
 249   if ($key){
 250     list($key, $subkey) = explode(' ', $key, 2);
 251     $subkey = trim($subkey);
 252
 253     if ($subkey) {
 254       return isset($meta['current'][$key][$subkey]) ? $meta['current'][$key][$subkey] : null;
 255     }
 256
 257     return isset($meta['current'][$key]) ? $meta['current'][$key] : null;
 258   }
 259
 260   return $meta['current'];
 261 }
 262
 263 /**
 264  * sets metadata elements of a page
 265  *
 266  * @author Esther Brunner <esther@kaffeehaus.ch>
 267  */
 268 function p_set_metadata($id, $data, $render=false, $persistent=true){
 269   if (!is_array($data)) return false;
 270
 271   global $ID;
 272
 273   // cache the current page
 274   $cache = ($ID == $id);
 275   $orig = p_read_metadata($id, $cache);
 276
 277   // render metadata first?
 278   $meta = $render ? p_render_metadata($id, $orig) : $orig;
 279
 280   // now add the passed metadata
 281   $protected = array('description', 'date', 'contributor');
 282   foreach ($data as $key => $value){
 283
 284     // be careful with sub-arrays of $meta['relation']
 285     if ($key == 'relation'){
 286
 287       foreach ($value as $subkey => $subvalue){
 288         $meta['current'][$key][$subkey] = !empty($meta['current'][$key][$subkey]) ? array_merge($meta['current'][$key][$subkey], $subvalue) : $subvalue;
 289         if ($persistent)
 290           $meta['persistent'][$key][$subkey] = !empty($meta['persistent'][$key][$subkey]) ? array_merge($meta['persistent'][$key][$subkey], $subvalue) : $subvalue;
 291       }
 292
 293     // be careful with some senisitive arrays of $meta
 294     } elseif (in_array($key, $protected)){
 295
 296       // these keys, must have subkeys - a legitimate value must be an array
 297       if (is_array($value)) {
 298         $meta['current'][$key] = !empty($meta['current'][$key]) ? array_merge($meta['current'][$key],$value) : $value;
 299
 300         if ($persistent) {
 301           $meta['persistent'][$key] = !empty($meta['persistent'][$key]) ? array_merge($meta['persistent'][$key],$value) : $value;
 302         }
 303       }
 304
 305     // no special treatment for the rest
 306     } else {
 307       $meta['current'][$key] = $value;
 308       if ($persistent) $meta['persistent'][$key] = $value;
 309     }
 310   }
 311
 312   // save only if metadata changed
 313   if ($meta == $orig) return true;
 314
 315   // sync cached copies, including $INFO metadata
 316   global $cache_metadata, $INFO;
 317
 318   if (!empty($cache_metadata[$id])) $cache_metadata[$id] = $meta;
 319   if (!empty($INFO) && ($id == $INFO['id'])) { $INFO['meta'] = $meta['current']; }
 320
 321   return io_saveFile(metaFN($id, '.meta'), serialize($meta));
 322 }
 323
 324 /**
 325  * Purges the non-persistant part of the meta data
 326  * used on page deletion
 327  *
 328  * @author Michael Klier <chi@chimeric.de>
 329  */
 330 function p_purge_metadata($id) {
 331     $metafn = metaFN('id', '.meta');
 332     $meta   = p_read_metadata($id);
 333     foreach($meta['current'] as $key => $value) {
 334         if(is_array($meta[$key])) {
 335             $meta['current'][$key] = array();
 336         } else {
 337             $meta['current'][$key] = '';
 338         }
 339     }
 340     return io_saveFile(metaFN($id, '.meta'), serialize($meta));
 341 }
 342
 343 /**
 344  * read the metadata from source/cache for $id
 345  * (internal use only - called by p_get_metadata & p_set_metadata)
 346  *
 347  * this function also converts the metadata from the original format to
 348  * the current format ('current' & 'persistent' arrays)
 349  *
 350  * @author   Christopher Smith <chris@jalakai.co.uk>
 351  *
 352  * @param    string   $id      absolute wiki page id
 353  * @param    bool     $cache   whether or not to cache metadata in memory
 354  *                             (only use for metadata likely to be accessed several times)
 355  *
 356  * @return   array             metadata
 357  */
 358 function p_read_metadata($id,$cache=false) {
 359   global $cache_metadata;
 360
 361   if (isset($cache_metadata[(string)$id])) return $cache_metadata[(string)$id];
 362
 363   $file = metaFN($id, '.meta');
 364   $meta = @file_exists($file) ? unserialize(io_readFile($file, false)) : array('current'=>array(),'persistent'=>array());
 365
 366   // convert $meta from old format to new (current+persistent) format
 367   if (!isset($meta['current'])) {
 368     $meta = array('current'=>$meta,'persistent'=>$meta);
 369
 370     // remove non-persistent keys
 371     unset($meta['persistent']['title']);
 372     unset($meta['persistent']['description']['abstract']);
 373     unset($meta['persistent']['description']['tableofcontents']);
 374     unset($meta['persistent']['relation']['haspart']);
 375     unset($meta['persistent']['relation']['references']);
 376     unset($meta['persistent']['date']['valid']);
 377
 378     if (empty($meta['persistent']['description'])) unset($meta['persistent']['description']);
 379     if (empty($meta['persistent']['relation'])) unset($meta['persistent']['relation']);
 380     if (empty($meta['persistent']['date'])) unset($meta['persistent']['date']);
 381
 382     // save converted metadata
 383     io_saveFile($file, serialize($meta));
 384   }
 385
 386   if ($cache) {
 387     $cache_metadata[(string)$id] = $meta;
 388   }
 389
 390   return $meta;
 391 }
 392
 393 /**
 394  * renders the metadata of a page
 395  *
 396  * @author Esther Brunner <esther@kaffeehaus.ch>
 397  */
 398 function p_render_metadata($id, $orig){
 399   // make sure the correct ID is in global ID
 400   global $ID;
 401   $keep = $ID;
 402   $ID   = $id;
 403
 404
 405   // add an extra key for the event - to tell event handlers the page whose metadata this is
 406   $orig['page'] = $id;
 407   $evt = new Doku_Event('PARSER_METADATA_RENDER', $orig);
 408   if ($evt->advise_before()) {
 409
 410     require_once DOKU_INC."inc/parser/metadata.php";
 411
 412     // get instructions
 413     $instructions = p_cached_instructions(wikiFN($id),false,$id);
 414     if(is_null($instructions)){
 415       $ID = $keep;
 416       return null; // something went wrong with the instructions
 417     }
 418
 419     // set up the renderer
 420     $renderer = & new Doku_Renderer_metadata();
 421     $renderer->meta = $orig['current'];
 422     $renderer->persistent = $orig['persistent'];
 423
 424     // loop through the instructions
 425     foreach ($instructions as $instruction){
 426       // execute the callback against the renderer
 427       call_user_func_array(array(&$renderer, $instruction[0]), $instruction[1]);
 428     }
 429
 430     $evt->result = array('current'=>$renderer->meta,'persistent'=>$renderer->persistent);
 431   }
 432   $evt->advise_after();
 433
 434   $ID = $keep;
 435   return $evt->result;
 436 }
 437
 438 /**
 439  * returns all available parser syntax modes in correct order
 440  *
 441  * @author Andreas Gohr <andi@splitbrain.org>
 442  */
 443 function p_get_parsermodes(){
 444   global $conf;
 445
 446   //reuse old data
 447   static $modes = null;
 448   if($modes != null){
 449     return $modes;
 450   }
 451
 452   //import parser classes and mode definitions
 453   require_once DOKU_INC . 'inc/parser/parser.php';
 454
 455   // we now collect all syntax modes and their objects, then they will
 456   // be sorted and added to the parser in correct order
 457   $modes = array();
 458
 459   // add syntax plugins
 460   $pluginlist = plugin_list('syntax');
 461   if(count($pluginlist)){
 462     global $PARSER_MODES;
 463     $obj = null;
 464     foreach($pluginlist as $p){
 465       if(!$obj =& plugin_load('syntax',$p)) continue; //attempt to load plugin into $obj
 466       $PARSER_MODES[$obj->getType()][] = "plugin_$p"; //register mode type
 467       //add to modes
 468       $modes[] = array(
 469                    'sort' => $obj->getSort(),
 470                    'mode' => "plugin_$p",
 471                    'obj'  => $obj,
 472                  );
 473       unset($obj); //remove the reference
 474     }
 475   }
 476
 477   // add default modes
 478   $std_modes = array('listblock','preformatted','notoc','nocache',
 479                      'header','table','linebreak','footnote','hr',
 480                      'unformatted','php','html','code','file','quote',
 481                      'internallink','rss','media','externallink',
 482                      'emaillink','windowssharelink','eol');
 483   if($conf['typography']){
 484     $std_modes[] = 'quotes';
 485     $std_modes[] = 'multiplyentity';
 486   }
 487   foreach($std_modes as $m){
 488     $class = "Doku_Parser_Mode_$m";
 489     $obj   = new $class();
 490     $modes[] = array(
 491                  'sort' => $obj->getSort(),
 492                  'mode' => $m,
 493                  'obj'  => $obj
 494                );
 495   }
 496
 497   // add formatting modes
 498   $fmt_modes = array('strong','emphasis','underline','monospace',
 499                      'subscript','superscript','deleted');
 500   foreach($fmt_modes as $m){
 501     $obj   = new Doku_Parser_Mode_formatting($m);
 502     $modes[] = array(
 503                  'sort' => $obj->getSort(),
 504                  'mode' => $m,
 505                  'obj'  => $obj
 506                );
 507   }
 508
 509   // add modes which need files
 510   $obj     = new Doku_Parser_Mode_smiley(array_keys(getSmileys()));
 511   $modes[] = array('sort' => $obj->getSort(), 'mode' => 'smiley','obj'  => $obj );
 512   $obj     = new Doku_Parser_Mode_acronym(array_keys(getAcronyms()));
 513   $modes[] = array('sort' => $obj->getSort(), 'mode' => 'acronym','obj'  => $obj );
 514   $obj     = new Doku_Parser_Mode_entity(array_keys(getEntities()));
 515   $modes[] = array('sort' => $obj->getSort(), 'mode' => 'entity','obj'  => $obj );
 516
 517
 518   // add optional camelcase mode
 519   if($conf['camelcase']){
 520     $obj     = new Doku_Parser_Mode_camelcaselink();
 521     $modes[] = array('sort' => $obj->getSort(), 'mode' => 'camelcaselink','obj'  => $obj );
 522   }
 523
 524   //sort modes
 525   usort($modes,'p_sort_modes');
 526
 527   return $modes;
 528 }
 529
 530 /**
 531  * Callback function for usort
 532  *
 533  * @author Andreas Gohr <andi@splitbrain.org>
 534  */
 535 function p_sort_modes($a, $b){
 536   if($a['sort'] == $b['sort']) return 0;
 537   return ($a['sort'] < $b['sort']) ? -1 : 1;
 538 }
 539
 540 /**
 541  * Renders a list of instruction to the specified output mode
 542  *
 543  * In the $info array are informations from the renderer returned
 544  *
 545  * @author Harry Fuecks <hfuecks@gmail.com>
 546  * @author Andreas Gohr <andi@splitbrain.org>
 547  */
 548 function p_render($mode,$instructions,&$info){
 549   if(is_null($instructions)) return '';
 550
 551   $Renderer =& p_get_renderer($mode);
 552   if (is_null($Renderer)) return null;
 553
 554   $Renderer->reset();
 555
 556   $Renderer->smileys = getSmileys();
 557   $Renderer->entities = getEntities();
 558   $Renderer->acronyms = getAcronyms();
 559   $Renderer->interwiki = getInterwiki();
 560   #$Renderer->badwords = getBadWords();
 561
 562   // Loop through the instructions
 563   foreach ( $instructions as $instruction ) {
 564       // Execute the callback against the Renderer
 565       call_user_func_array(array(&$Renderer, $instruction[0]),$instruction[1]);
 566   }
 567
 568   //set info array
 569   $info = $Renderer->info;
 570
 571   // Post process and return the output
 572   $data = array($mode,& $Renderer->doc);
 573   trigger_event('RENDERER_CONTENT_POSTPROCESS',$data);
 574   return $Renderer->doc;
 575 }
 576
 577 function & p_get_renderer($mode) {
 578   global $conf;
 579
 580   $rname = !empty($conf['renderer_'.$mode]) ? $conf['renderer_'.$mode] : $mode;
 581
 582   // try default renderer first:
 583   $file = DOKU_INC."inc/parser/$rname.php";
 584   if(@file_exists($file)){
 585     require_once $file;
 586     $rclass = "Doku_Renderer_$rname";
 587
 588     if ( !class_exists($rclass) ) {
 589       trigger_error("Unable to resolve render class $rclass",E_USER_WARNING);
 590       msg("Renderer '$rname' for $mode not valid",-1);
 591       return null;
 592     }
 593     $Renderer = & new $rclass();
 594   }else{
 595     // Maybe a plugin is available?
 596     $Renderer =& plugin_load('renderer',$rname, true);
 597     if(is_null($Renderer)){
 598       msg("No renderer '$rname' found for mode '$mode'",-1);
 599       return null;
 600     }
 601   }
 602
 603   return $Renderer;
 604 }
 605
 606 /**
 607  * Gets the first heading from a file
 608  *
 609  * @param   string   $id       dokuwiki page id
 610  * @param   bool     $render   rerender if first heading not known
 611  *                             default: true  -- must be set to false for calls from the metadata renderer to
 612  *                                               protects against loops and excessive resource usage when pages
 613  *                                               for which only a first heading is required will attempt to
 614  *                                               render metadata for all the pages for which they require first
 615  *                                               headings ... and so on.
 616  *
 617  * @author Andreas Gohr <andi@splitbrain.org>
 618  */
 619 function p_get_first_heading($id, $render=true){
 620   return p_get_metadata($id,'title',$render);
 621 }
 622
 623 /**
 624  * Wrapper for GeSHi Code Highlighter, provides caching of its output
 625  *
 626  * @param  string   $code       source code to be highlighted
 627  * @param  string   $language   language to provide highlighting
 628  * @param  string   $wrapper    html element to wrap the returned highlighted text
 629  *
 630  * @author Christopher Smith <chris@jalakai.co.uk>
 631  * @author Andreas Gohr <andi@splitbrain.org>
 632  */
 633 function p_xhtml_cached_geshi($code, $language, $wrapper='pre') {
 634   global $conf, $config_cascade;
 635   $language = strtolower($language);
 636
 637   // remove any leading or trailing blank lines
 638   $code = preg_replace('/^\s*?\n|\s*?\n$/','',$code);
 639
 640   $cache = getCacheName($language.$code,".code");
 641   $ctime = @filemtime($cache);
 642   if($ctime && !$_REQUEST['purge'] &&
 643      $ctime > filemtime(DOKU_INC.'inc/geshi.php') &&                 // geshi changed
 644      $ctime > @filemtime(DOKU_INC.'inc/geshi/'.$language.'.php') &&  // language syntax definition changed
 645      $ctime > filemtime(reset($config_cascade['main']['default']))){ // dokuwiki changed
 646     $highlighted_code = io_readFile($cache, false);
 647
 648   } else {
 649
 650     require_once(DOKU_INC . 'inc/geshi.php');
 651
 652     $geshi = new GeSHi($code, $language, DOKU_INC . 'inc/geshi');
 653     $geshi->set_encoding('utf-8');
 654     $geshi->enable_classes();
 655     $geshi->set_header_type(GESHI_HEADER_PRE);
 656     $geshi->set_link_target($conf['target']['extern']);
 657
 658     // remove GeSHi's wrapper element (we'll replace it with our own later)
 659     // we need to use a GeSHi wrapper to avoid <BR> throughout the highlighted text
 660     $highlighted_code = preg_replace('!^<pre[^>]*>|</pre>$!','',$geshi->parse_code());
 661     io_saveFile($cache,$highlighted_code);
 662   }
 663
 664   // add a wrapper element if required
 665   if ($wrapper) {
 666     return "<$wrapper class=\"code $language\">$highlighted_code</$wrapper>";
 667   } else {
 668     return $highlighted_code;
 669   }
 670 }
 671
 672 //Setup VIM: ex: et ts=2 enc=utf-8 :