Fixed bug #12682: Section links coming from RTE broken in FE
[Packages/TYPO3.CMS.git] / t3lib / class.t3lib_parsehtml_proc.php
1 <?php
2 /***************************************************************
3 * Copyright notice
4 *
5 * (c) 1999-2009 Kasper Skaarhoj (kasperYYYY@typo3.com)
6 * All rights reserved
7 *
8 * This script is part of the TYPO3 project. The TYPO3 project is
9 * free software; you can redistribute it and/or modify
10 * it under the terms of the GNU General Public License as published by
11 * the Free Software Foundation; either version 2 of the License, or
12 * (at your option) any later version.
13 *
14 * The GNU General Public License can be found at
15 * http://www.gnu.org/copyleft/gpl.html.
16 * A copy is found in the textfile GPL.txt and important notices to the license
17 * from the author is found in LICENSE.txt distributed with these scripts.
18 *
19 *
20 * This script is distributed in the hope that it will be useful,
21 * but WITHOUT ANY WARRANTY; without even the implied warranty of
22 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
23 * GNU General Public License for more details.
24 *
25 * This copyright notice MUST APPEAR in all copies of the script!
26 ***************************************************************/
27 /**
28 * Functions for parsing HTML, specially for TYPO3 processing in relation to TCEmain and Rich Text Editor (RTE)
29 *
30 * $Id$
31 * Revised for TYPO3 3.6 December/2003 by Kasper Skaarhoj
32 * XHTML compatible.
33 *
34 * @author Kasper Skaarhoj <kasperYYYY@typo3.com>
35 * @internal
36 */
37 /**
38 * [CLASS/FUNCTION INDEX of SCRIPT]
39 *
40 *
41 *
42 * 103: class t3lib_parsehtml_proc extends t3lib_parsehtml
43 * 138: function init($elRef='',$recPid=0)
44 * 150: function setRelPath($path)
45 * 174: function evalWriteFile($pArr,$currentRecord)
46 *
47 * SECTION: Main function
48 * 232: function RTE_transform($value,$specConf,$direction='rte',$thisConfig=array())
49 *
50 * SECTION: Specific RTE TRANSFORMATION functions
51 * 398: function TS_images_db($value)
52 * 550: function TS_images_rte($value)
53 * 589: function TS_reglinks($value,$direction)
54 * 626: function TS_links_db($value)
55 * 675: function TS_links_rte($value)
56 * 760: function TS_preserve_db($value)
57 * 784: function TS_preserve_rte($value)
58 * 805: function TS_transform_db($value,$css=FALSE)
59 * 922: function transformStyledATags($value)
60 * 948: function TS_transform_rte($value,$css=0)
61 * 1019: function TS_strip_db($value)
62 *
63 * SECTION: Generic RTE transformation, analysis and helper functions
64 * 1050: function getURL($url)
65 * 1064: function HTMLcleaner_db($content,$tagList='')
66 * 1091: function getKeepTags($direction='rte',$tagList='')
67 * 1200: function divideIntoLines($value,$count=5,$returnArray=FALSE)
68 * 1304: function setDivTags($value,$dT='p')
69 * 1349: function internalizeFontTags($value)
70 * 1385: function siteUrl()
71 * 1395: function rteImageStorageDir()
72 * 1407: function removeTables($value,$breakChar='<br />')
73 * 1439: function defaultTStagMapping($code,$direction='rte')
74 * 1462: function getWHFromAttribs($attribArray)
75 * 1489: function urlInfoForLinkTags($url)
76 * 1548: function TS_AtagToAbs($value,$dontSetRTEKEEP=FALSE)
77 *
78 * TOTAL FUNCTIONS: 28
79 * (This index is automatically created/updated by the extension "extdeveval")
80 *
81 */
82
83
84
85
86
87
88
89
90
91
92
93
94
95 /**
96 * Class for parsing HTML for the Rich Text Editor. (also called transformations)
97 *
98 * @author Kasper Skaarhoj <kasperYYYY@typo3.com>
99 * @package TYPO3
100 * @subpackage t3lib
101 */
102 class t3lib_parsehtml_proc extends t3lib_parsehtml {
103
104 // Static:
105 var $blockElementList = 'PRE,UL,OL,H1,H2,H3,H4,H5,H6,HR,ADDRESS,DL,DD'; // List of tags for these elements
106
107 // Internal, static:
108 var $recPid = 0; // Set this to the pid of the record manipulated by the class.
109 var $elRef = ''; // Element reference [table]:[field], eg. "tt_content:bodytext"
110 var $relPath=''; // Relative path
111 var $relBackPath=''; // Relative back-path
112 public $tsConfig = array(); // Current Page TSConfig
113 var $procOptions = ''; // Set to the TSconfig options coming from Page TSconfig
114
115 // Internal, dynamic
116 var $TS_transform_db_safecounter=100; // Run-away brake for recursive calls.
117 var $rte_p=''; // Parameters from TCA types configuration related to the RTE
118 var $getKeepTags_cache=array(); // Data caching for processing function
119 var $allowedClasses=array(); // Storage of the allowed CSS class names in the RTE
120 var $preserveTags = ''; // Set to tags to preserve from Page TSconfig configuration
121
122
123
124
125
126
127
128
129
130
131 /**
132 * Initialize, setting element reference and record PID
133 *
134 * @param string Element reference, eg "tt_content:bodytext"
135 * @param integer PID of the record (page id)
136 * @return void
137 */
138 function init($elRef='',$recPid=0) {
139 $this->recPid = $recPid;
140 $this->elRef = $elRef;
141 }
142
143 /**
144 * Setting the ->relPath and ->relBackPath to proper values so absolute references to links and images can be converted to relative dittos.
145 * This is used when editing files with the RTE
146 *
147 * @param string The relative path from PATH_site to the place where the file being edited is. Eg. "fileadmin/static".
148 * @return void There is no output, it is set in internal variables. With the above example of "fileadmin/static" as input this will yield ->relPath to be "fileadmin/static/" and ->relBackPath to be "../../"
149 */
150 function setRelPath($path) {
151 $path = trim($path);
152 $path = preg_replace('/^\//','',$path);
153 $path = preg_replace('/\/$/','',$path);
154 if ($path) {
155 $this->relPath = $path;
156 $this->relBackPath = '';
157 $partsC=count(explode('/',$this->relPath));
158 for ($a=0;$a<$partsC;$a++) {
159 $this->relBackPath.='../';
160 }
161 $this->relPath.='/';
162 }
163 }
164
165 /**
166 * Evaluate the environment for editing a staticFileEdit file.
167 * Called for almost all fields being saved in the database. Is called without an instance of the object: t3lib_parsehtml_proc::evalWriteFile()
168 *
169 * @param array Parameters for the current field as found in types-config
170 * @param array Current record we are editing.
171 * @return mixed On success an array with various information is returned, otherwise a string with an error message
172 * @see t3lib_TCEmain, t3lib_transferData
173 */
174 function evalWriteFile($pArr,$currentRecord) {
175
176 // Write file configuration:
177 if (is_array($pArr)) {
178 if ($GLOBALS['TYPO3_CONF_VARS']['BE']['staticFileEditPath']
179 && substr($GLOBALS['TYPO3_CONF_VARS']['BE']['staticFileEditPath'],-1)=='/'
180 && @is_dir(PATH_site.$GLOBALS['TYPO3_CONF_VARS']['BE']['staticFileEditPath'])) {
181
182 $SW_p = $pArr['parameters'];
183 $SW_editFileField = trim($SW_p[0]);
184 $SW_editFile = $currentRecord[$SW_editFileField];
185 if ($SW_editFileField && $SW_editFile && t3lib_div::validPathStr($SW_editFile)) {
186 $SW_relpath = $GLOBALS['TYPO3_CONF_VARS']['BE']['staticFileEditPath'].$SW_editFile;
187 $SW_editFile = PATH_site.$SW_relpath;
188 if (@is_file($SW_editFile)) {
189 return array(
190 'editFile' => $SW_editFile,
191 'relEditFile' => $SW_relpath,
192 'contentField' => trim($SW_p[1]),
193 'markerField' => trim($SW_p[2]),
194 'loadFromFileField' => trim($SW_p[3]),
195 'statusField' => trim($SW_p[4])
196 );
197 } else return "ERROR: Editfile '".$SW_relpath."' did not exist";
198 } else return "ERROR: Edit file name could not be found or was bad.";
199 } else return "ERROR: staticFileEditPath was not set, not set correctly or did not exist!";
200 }
201 }
202
203
204
205
206
207
208
209
210
211
212
213
214
215 /**********************************************
216 *
217 * Main function
218 *
219 **********************************************/
220
221 /**
222 * Transform value for RTE based on specConf in the direction specified by $direction (rte/db)
223 * This is the main function called from tcemain and transfer data classes
224 *
225 * @param string Input value
226 * @param array Special configuration for a field; This is coming from the types-configuration of the field in the TCA. In the types-configuration you can setup features for the field rendering and in particular the RTE takes al its major configuration options from there!
227 * @param string Direction of the transformation. Two keywords are allowed; "db" or "rte". If "db" it means the transformation will clean up content coming from the Rich Text Editor and goes into the database. The other direction, "rte", is of course when content is coming from database and must be transformed to fit the RTE.
228 * @param array Parsed TypoScript content configuring the RTE, probably coming from Page TSconfig.
229 * @return string Output value
230 * @see t3lib_TCEmain::fillInFieldArray(), t3lib_transferData::renderRecord_typesProc()
231 */
232 function RTE_transform($value,$specConf,$direction='rte',$thisConfig=array()) {
233
234 // Init:
235 $this->tsConfig = $thisConfig;
236 $this->procOptions = $thisConfig['proc.'];
237 $this->preserveTags = strtoupper(implode(',',t3lib_div::trimExplode(',',$this->procOptions['preserveTags'])));
238
239 // dynamic configuration of blockElementList
240 if ($this->procOptions['blockElementList']) {
241 $this->blockElementList = $this->procOptions['blockElementList'];
242 }
243
244 // Get parameters for rte_transformation:
245 $p = $this->rte_p = t3lib_BEfunc::getSpecConfParametersFromArray($specConf['rte_transform']['parameters']);
246
247 // Setting modes:
248 if (strcmp($this->procOptions['overruleMode'],'')) {
249 $modes = array_unique(t3lib_div::trimExplode(',',$this->procOptions['overruleMode']));
250 } else {
251 $modes = array_unique(t3lib_div::trimExplode('-',$p['mode']));
252 }
253 $revmodes = array_flip($modes);
254
255 // Find special modes and extract them:
256 if (isset($revmodes['ts'])) {
257 $modes[$revmodes['ts']] = 'ts_transform,ts_preserve,ts_images,ts_links';
258 }
259 // Find special modes and extract them:
260 if (isset($revmodes['ts_css'])) {
261 $modes[$revmodes['ts_css']] = 'css_transform,ts_images,ts_links';
262 }
263
264 // Make list unique
265 $modes = array_unique(t3lib_div::trimExplode(',',implode(',',$modes),1));
266
267 // Reverse order if direction is "rte"
268 if ($direction=='rte') {
269 $modes = array_reverse($modes);
270 }
271
272 // Getting additional HTML cleaner configuration. These are applied either before or after the main transformation is done and is thus totally independant processing options you can set up:
273 $entry_HTMLparser = $this->procOptions['entryHTMLparser_'.$direction] ? $this->HTMLparserConfig($this->procOptions['entryHTMLparser_'.$direction.'.']) : '';
274 $exit_HTMLparser = $this->procOptions['exitHTMLparser_'.$direction] ? $this->HTMLparserConfig($this->procOptions['exitHTMLparser_'.$direction.'.']) : '';
275
276 // Line breaks of content is unified into char-10 only (removing char 13)
277 if (!$this->procOptions['disableUnifyLineBreaks']) {
278 $value = str_replace(chr(13).chr(10),chr(10),$value);
279 }
280
281 // In an entry-cleaner was configured, pass value through the HTMLcleaner with that:
282 if (is_array($entry_HTMLparser)) {
283 $value = $this->HTMLcleaner($value,$entry_HTMLparser[0],$entry_HTMLparser[1],$entry_HTMLparser[2],$entry_HTMLparser[3]);
284 }
285
286 // Traverse modes:
287 foreach($modes as $cmd) {
288 // ->DB
289 if ($direction=='db') {
290 // Checking for user defined transformation:
291 if ($_classRef = $GLOBALS['TYPO3_CONF_VARS']['SC_OPTIONS']['t3lib/class.t3lib_parsehtml_proc.php']['transformation'][$cmd]) {
292 $_procObj = t3lib_div::getUserObj($_classRef);
293 $_procObj->pObj = $this;
294 $_procObj->transformationKey = $cmd;
295 $value = $_procObj->transform_db($value,$this);
296 } else { // ... else use defaults:
297 switch($cmd) {
298 case 'ts_images':
299 $value = $this->TS_images_db($value);
300 break;
301 case 'ts_reglinks':
302 $value = $this->TS_reglinks($value,'db');
303 break;
304 case 'ts_links':
305 $value = $this->TS_links_db($value);
306 break;
307 case 'ts_preserve':
308 $value = $this->TS_preserve_db($value);
309 break;
310 case 'ts_transform':
311 case 'css_transform':
312 $value = str_replace(chr(13),'',$value); // Has a very disturbing effect, so just remove all '13' - depend on '10'
313 $this->allowedClasses = t3lib_div::trimExplode(',', $this->procOptions['allowedClasses'], 1);
314 $value = $this->TS_transform_db($value,$cmd=='css_transform');
315 break;
316 case 'ts_strip':
317 $value = $this->TS_strip_db($value);
318 break;
319 default:
320 break;
321 }
322 }
323 }
324 // ->RTE
325 if ($direction=='rte') {
326 // Checking for user defined transformation:
327 if ($_classRef = $GLOBALS['TYPO3_CONF_VARS']['SC_OPTIONS']['t3lib/class.t3lib_parsehtml_proc.php']['transformation'][$cmd]) {
328 $_procObj = t3lib_div::getUserObj($_classRef);
329 $_procObj->pObj = $this;
330 $value = $_procObj->transform_rte($value,$this);
331 } else { // ... else use defaults:
332 switch($cmd) {
333 case 'ts_images':
334 $value = $this->TS_images_rte($value);
335 break;
336 case 'ts_reglinks':
337 $value = $this->TS_reglinks($value,'rte');
338 break;
339 case 'ts_links':
340 $value = $this->TS_links_rte($value);
341 break;
342 case 'ts_preserve':
343 $value = $this->TS_preserve_rte($value);
344 break;
345 case 'ts_transform':
346 case 'css_transform':
347 $value = str_replace(chr(13),'',$value); // Has a very disturbing effect, so just remove all '13' - depend on '10'
348 $value = $this->TS_transform_rte($value,$cmd=='css_transform');
349 break;
350 default:
351 break;
352 }
353 }
354 }
355 }
356
357 // In an exit-cleaner was configured, pass value through the HTMLcleaner with that:
358 if (is_array($exit_HTMLparser)) {
359 $value = $this->HTMLcleaner($value,$exit_HTMLparser[0],$exit_HTMLparser[1],$exit_HTMLparser[2],$exit_HTMLparser[3]);
360 }
361
362 // Final clean up of linebreaks:
363 if (!$this->procOptions['disableUnifyLineBreaks']) {
364 $value = str_replace(chr(13).chr(10),chr(10),$value); // Make sure no \r\n sequences has entered in the meantime...
365 $value = str_replace(chr(10),chr(13).chr(10),$value); // ... and then change all \n into \r\n
366 }
367
368 // Return value:
369 return $value;
370 }
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387 /************************************
388 *
389 * Specific RTE TRANSFORMATION functions
390 *
391 *************************************/
392
393 /**
394 * Transformation handler: 'ts_images' / direction: "db"
395 * Processing images inserted in the RTE.
396 * This is used when content goes from the RTE to the database.
397 * Images inserted in the RTE has an absolute URL applied to the src attribute. This URL is converted to a relative URL
398 * If it turns out that the URL is from another website than the current the image is read from that external URL and moved to the local server.
399 * Also "magic" images are processed here.
400 *
401 * @param string The content from RTE going to Database
402 * @return string Processed content
403 */
404 function TS_images_db($value) {
405
406 // Split content by <img> tags and traverse the resulting array for processing:
407 $imgSplit = $this->splitTags('img',$value);
408 foreach($imgSplit as $k => $v) {
409 if ($k%2) { // image found, do processing:
410
411 // Init
412 $attribArray = $this->get_tag_attributes_classic($v,1);
413 $siteUrl = $this->siteUrl();
414 $sitePath = str_replace (t3lib_div::getIndpEnv('TYPO3_REQUEST_HOST'), '', $siteUrl);
415
416 $absRef = trim($attribArray['src']); // It's always a absolute URL coming from the RTE into the Database.
417
418 // make path absolute if it is relative and we have a site path wich is not '/'
419 $pI=pathinfo($absRef);
420 if($sitePath AND !$pI['scheme'] && t3lib_div::isFirstPartOfStr($absRef,$sitePath)) {
421 // if site is in a subpath (eg. /~user_jim/) this path needs to be removed because it will be added with $siteUrl
422 $absRef = substr($absRef,strlen($sitePath));
423 $absRef = $siteUrl.$absRef;
424 }
425
426 // External image from another URL? In that case, fetch image (unless disabled feature).
427 if (!t3lib_div::isFirstPartOfStr($absRef,$siteUrl) && !$this->procOptions['dontFetchExtPictures']) {
428 $externalFile = $this->getUrl($absRef); // Get it
429 if ($externalFile) {
430 $pU = parse_url($absRef);
431 $pI=pathinfo($pU['path']);
432
433 if (t3lib_div::inList('gif,png,jpeg,jpg',strtolower($pI['extension']))) {
434 $filename = t3lib_div::shortMD5($absRef).'.'.$pI['extension'];
435 $origFilePath = PATH_site.$this->rteImageStorageDir().'RTEmagicP_'.$filename;
436 $C_origFilePath = PATH_site.$this->rteImageStorageDir().'RTEmagicC_'.$filename.'.'.$pI['extension'];
437 if (!@is_file($origFilePath)) {
438 t3lib_div::writeFile($origFilePath,$externalFile);
439 t3lib_div::writeFile($C_origFilePath,$externalFile);
440 }
441 $absRef = $siteUrl.$this->rteImageStorageDir().'RTEmagicC_'.$filename.'.'.$pI['extension'];
442
443 $attribArray['src']=$absRef;
444 $params = t3lib_div::implodeAttributes($attribArray,1);
445 $imgSplit[$k] = '<img '.$params.' />';
446 }
447 }
448 }
449
450 // Check image as local file (siteURL equals the one of the image)
451 if (t3lib_div::isFirstPartOfStr($absRef,$siteUrl)) {
452 $path = rawurldecode(substr($absRef,strlen($siteUrl))); // Rel-path, rawurldecoded for special characters.
453 $filepath = t3lib_div::getFileAbsFileName($path); // Abs filepath, locked to relative path of this project.
454
455 // Check file existence (in relative dir to this installation!)
456 if ($filepath && @is_file($filepath)) {
457
458 // If "magic image":
459 $pathPre=$this->rteImageStorageDir().'RTEmagicC_';
460 if (t3lib_div::isFirstPartOfStr($path,$pathPre)) {
461 // Find original file:
462 $pI=pathinfo(substr($path,strlen($pathPre)));
463 $filename = substr($pI['basename'],0,-strlen('.'.$pI['extension']));
464 $origFilePath = PATH_site.$this->rteImageStorageDir().'RTEmagicP_'.$filename;
465 if (@is_file($origFilePath)) {
466 $imgObj = t3lib_div::makeInstance('t3lib_stdGraphic');
467 $imgObj->init();
468 $imgObj->mayScaleUp=0;
469 $imgObj->tempPath=PATH_site.$imgObj->tempPath;
470
471 $curInfo = $imgObj->getImageDimensions($filepath); // Image dimensions of the current image
472 $curWH = $this->getWHFromAttribs($attribArray); // Image dimensions as set in the image tag
473 // Compare dimensions:
474 if ($curWH[0]!=$curInfo[0] || $curWH[1]!=$curInfo[1]) {
475 $origImgInfo = $imgObj->getImageDimensions($origFilePath); // Image dimensions of the current image
476 $cW = $curWH[0];
477 $cH = $curWH[1];
478 $cH = 1000; // Make the image based on the width solely...
479 $imgI = $imgObj->imageMagickConvert($origFilePath,$pI['extension'],$cW.'m',$cH.'m');
480 if ($imgI[3]) {
481 $fI=pathinfo($imgI[3]);
482 @copy($imgI[3],$filepath); // Override the child file
483 // Removing width and heigth form style attribute
484 $attribArray['style'] = preg_replace('/((?:^|)\s*(?:width|height)\s*:[^;]*(?:$|;))/si', '', $attribArray['style']);
485 $attribArray['width']=$imgI[0];
486 $attribArray['height']=$imgI[1];
487 $params = t3lib_div::implodeAttributes($attribArray,1);
488 $imgSplit[$k]='<img '.$params.' />';
489 }
490 }
491 }
492
493 } elseif ($this->procOptions['plainImageMode']) { // If "plain image" has been configured:
494
495 // Image dimensions as set in the image tag, if any
496 $curWH = $this->getWHFromAttribs($attribArray);
497 if ($curWH[0]) $attribArray['width'] = $curWH[0];
498 if ($curWH[1]) $attribArray['height'] = $curWH[1];
499
500 // Removing width and heigth form style attribute
501 $attribArray['style'] = preg_replace('/((?:^|)\s*(?:width|height)\s*:[^;]*(?:$|;))/si', '', $attribArray['style']);
502
503 // Finding dimensions of image file:
504 $fI = @getimagesize($filepath);
505
506 // Perform corrections to aspect ratio based on configuration:
507 switch((string)$this->procOptions['plainImageMode']) {
508 case 'lockDimensions':
509 $attribArray['width']=$fI[0];
510 $attribArray['height']=$fI[1];
511 break;
512 case 'lockRatioWhenSmaller': // If the ratio has to be smaller, then first set the width...:
513 if ($attribArray['width']>$fI[0]) $attribArray['width'] = $fI[0];
514 case 'lockRatio':
515 if ($fI[0]>0) {
516 $attribArray['height']=round($attribArray['width']*($fI[1]/$fI[0]));
517 }
518 break;
519 }
520
521 // Compile the image tag again:
522 $params = t3lib_div::implodeAttributes($attribArray,1);
523 $imgSplit[$k]='<img '.$params.' />';
524 }
525 } else { // Remove image if it was not found in a proper position on the server!
526
527 // Commented out; removing the image tag might not be that logical...
528 #$imgSplit[$k]='';
529 }
530 }
531
532 // Convert abs to rel url
533 if ($imgSplit[$k]) {
534 $attribArray=$this->get_tag_attributes_classic($imgSplit[$k],1);
535 $absRef = trim($attribArray['src']);
536 if (t3lib_div::isFirstPartOfStr($absRef,$siteUrl)) {
537 $attribArray['src'] = $this->relBackPath.substr($absRef,strlen($siteUrl));
538 if (!isset($attribArray['alt'])) $attribArray['alt']=''; // Must have alt-attribute for XHTML compliance.
539 $imgSplit[$k]='<img '.t3lib_div::implodeAttributes($attribArray,1,1).' />';
540 }
541 }
542 }
543 }
544 return implode('',$imgSplit);
545 }
546
547 /**
548 * Transformation handler: 'ts_images' / direction: "rte"
549 * Processing images from database content going into the RTE.
550 * Processing includes converting the src attribute to an absolute URL.
551 *
552 * @param string Content input
553 * @return string Content output
554 */
555 function TS_images_rte($value) {
556
557 $siteUrl = $this->siteUrl();
558 $sitePath = str_replace (t3lib_div::getIndpEnv('TYPO3_REQUEST_HOST'), '', $siteUrl);
559
560 // Split content by <img> tags and traverse the resulting array for processing:
561 $imgSplit = $this->splitTags('img',$value);
562 foreach($imgSplit as $k => $v) {
563 if ($k%2) { // image found:
564
565 // Init
566 $attribArray=$this->get_tag_attributes_classic($v,1);
567 $absRef = trim($attribArray['src']);
568
569 // Unless the src attribute is already pointing to an external URL:
570 if (strtolower(substr($absRef,0,4))!='http') {
571 $attribArray['src'] = substr($attribArray['src'],strlen($this->relBackPath));
572 // if site is in a subpath (eg. /~user_jim/) this path needs to be removed because it will be added with $siteUrl
573 $attribArray['src'] = preg_replace('#^'.preg_quote($sitePath,'#').'#','',$attribArray['src']);
574 $attribArray['src'] = $siteUrl.$attribArray['src'];
575 if (!isset($attribArray['alt'])) $attribArray['alt']='';
576 $params = t3lib_div::implodeAttributes($attribArray);
577 $imgSplit[$k]='<img '.$params.' />';
578 }
579 }
580 }
581
582 // return processed content:
583 return implode('',$imgSplit);
584 }
585
586 /**
587 * Transformation handler: 'ts_reglinks' / direction: "db"+"rte" depending on $direction variable.
588 * Converting <A>-tags to/from abs/rel
589 *
590 * @param string Content input
591 * @param string Direction of conversion; "rte" (from database to RTE) or "db" (from RTE to database)
592 * @return string Content output
593 */
594 function TS_reglinks($value,$direction) {
595 $retVal = '';
596
597 switch($direction) {
598 case 'rte':
599 $retVal = $this->TS_AtagToAbs($value,1);
600 break;
601 case 'db':
602 $siteURL = $this->siteUrl();
603 $blockSplit = $this->splitIntoBlock('A',$value);
604 reset($blockSplit);
605 while(list($k,$v)=each($blockSplit)) {
606 if ($k%2) { // block:
607 $attribArray=$this->get_tag_attributes_classic($this->getFirstTag($v),1);
608 // If the url is local, remove url-prefix
609 if ($siteURL && substr($attribArray['href'],0,strlen($siteURL))==$siteURL) {
610 $attribArray['href']=$this->relBackPath.substr($attribArray['href'],strlen($siteURL));
611 }
612 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
613 $eTag='</a>';
614 $blockSplit[$k] = $bTag.$this->TS_reglinks($this->removeFirstAndLastTag($blockSplit[$k]),$direction).$eTag;
615 }
616 }
617 $retVal = implode('',$blockSplit);
618 break;
619 }
620 return $retVal;
621 }
622
623 /**
624 * Transformation handler: 'ts_links' / direction: "db"
625 * Converting <A>-tags to <link tags>
626 *
627 * @param string Content input
628 * @return string Content output
629 * @see TS_links_rte()
630 */
631 function TS_links_db($value) {
632
633 // Split content into <a> tag blocks and process:
634 $blockSplit = $this->splitIntoBlock('A',$value);
635 foreach($blockSplit as $k => $v) {
636 if ($k%2) { // If an A-tag was found:
637 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
638 $info = $this->urlInfoForLinkTags($attribArray['href']);
639
640 // Check options:
641 $attribArray_copy = $attribArray;
642 unset($attribArray_copy['href']);
643 unset($attribArray_copy['target']);
644 unset($attribArray_copy['class']);
645 unset($attribArray_copy['title']);
646 if ($attribArray_copy['rteerror']) { // Unset "rteerror" and "style" attributes if "rteerror" is set!
647 unset($attribArray_copy['style']);
648 unset($attribArray_copy['rteerror']);
649 }
650 if (!count($attribArray_copy)) { // Only if href, target and class are the only attributes, we can alter the link!
651 // Creating the TYPO3 pseudo-tag "<LINK>" for the link (includes href/url, target and class attributes):
652 $bTag='<link '.$info['url'].($info['query']?',0,'.$info['query']:'').($attribArray['target']?' '.$attribArray['target']:(($attribArray['class'] || $attribArray['title'])?' -':'')).($attribArray['class']?' '.$attribArray['class']:($attribArray['title']?' -':'')).($attribArray['title']?' "'.$attribArray['title'].'"':'').'>';
653 $eTag='</link>';
654 $blockSplit[$k] = $bTag.$this->TS_links_db($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
655 } else { // ... otherwise store the link as a-tag.
656 // Unsetting 'rtekeep' attribute if that had been set.
657 unset($attribArray['rtekeep']);
658 // If the url is local, remove url-prefix
659 $siteURL = $this->siteUrl();
660 if ($siteURL && substr($attribArray['href'],0,strlen($siteURL))==$siteURL) {
661 $attribArray['href']=$this->relBackPath.substr($attribArray['href'],strlen($siteURL));
662 }
663 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
664 $eTag='</a>';
665 $blockSplit[$k] = $bTag.$this->TS_links_db($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
666 }
667 }
668 }
669 return implode('',$blockSplit);
670 }
671
672 /**
673 * Transformation handler: 'ts_links' / direction: "rte"
674 * Converting <link tags> to <A>-tags
675 *
676 * @param string Content input
677 * @return string Content output
678 * @see TS_links_rte()
679 */
680 function TS_links_rte($value) {
681 $value = $this->TS_AtagToAbs($value);
682
683 // Split content by the TYPO3 pseudo tag "<link>":
684 $blockSplit = $this->splitIntoBlock('link',$value,1);
685 foreach($blockSplit as $k => $v) {
686 $error = '';
687 if ($k%2) { // block:
688 $tagCode = t3lib_div::unQuoteFilenames(trim(substr($this->getFirstTag($v),0,-1)),true);
689 $link_param = $tagCode[1];
690 $href = '';
691 $siteUrl = $this->siteUrl();
692 // Parsing the typolink data. This parsing is roughly done like in tslib_content->typolink()
693 if(strstr($link_param,'@')) { // mailadr
694 $href = 'mailto:'.preg_replace('/^mailto:/i','',$link_param);
695 } elseif (substr($link_param,0,1)=='#') { // check if anchor
696 $href = $siteUrl.$link_param;
697 } else {
698 $fileChar=intval(strpos($link_param, '/'));
699 $urlChar=intval(strpos($link_param, '.'));
700
701 // Detects if a file is found in site-root OR is a simulateStaticDocument.
702 list($rootFileDat) = explode('?',$link_param);
703 $rFD_fI = pathinfo($rootFileDat);
704 if (trim($rootFileDat) && !strstr($link_param,'/') && (@is_file(PATH_site.$rootFileDat) || t3lib_div::inList('php,html,htm',strtolower($rFD_fI['extension'])))) {
705 $href = $siteUrl.$link_param;
706 } elseif($urlChar && (strstr($link_param,'//') || !$fileChar || $urlChar<$fileChar)) { // url (external): If doubleSlash or if a '.' comes before a '/'.
707 if (!preg_match('/^[a-z]*:\/\//',trim(strtolower($link_param)))) {$scheme='http://';} else {$scheme='';}
708 $href = $scheme.$link_param;
709 } elseif($fileChar) { // file (internal)
710 $href = $siteUrl.$link_param;
711 } else { // integer or alias (alias is without slashes or periods or commas, that is 'nospace,alphanum_x,lower,unique' according to tables.php!!)
712 // Splitting the parameter by ',' and if the array counts more than 1 element it's a id/type/parameters triplet
713 $pairParts = t3lib_div::trimExplode(',', $link_param, TRUE);
714 $idPart = $pairParts[0];
715 $link_params_parts = explode('#', $idPart);
716 $idPart = trim($link_params_parts[0]);
717 $sectionMark = trim($link_params_parts[1]);
718 if (!strcmp($idPart,'')) { $idPart=$this->recPid; } // If no id or alias is given, set it to class record pid
719 // Checking if the id-parameter is an alias.
720 if (!t3lib_div::testInt($idPart)) {
721 list($idPartR) = t3lib_BEfunc::getRecordsByField('pages','alias',$idPart);
722 $idPart = intval($idPartR['uid']);
723 }
724 $page = t3lib_BEfunc::getRecord('pages', $idPart);
725 if (is_array($page)) { // Page must exist...
726 $href = $siteUrl .'?id=' . $idPart . ($pairParts[2] ? $pairParts[2] : '') . ($sectionMark ? '#' . $sectionMark : '');
727 // linkHandler - allowing links to start with registerd linkHandler e.g.. "record:"
728 } elseif (isset($GLOBALS['TYPO3_CONF_VARS']['SC_OPTIONS']['tslib/class.tslib_content.php']['typolinkLinkHandler'][array_shift(explode(':', $link_param))])) {
729 $href = $link_param;
730 } else {
731 #$href = '';
732 $href = $siteUrl.'?id='.$link_param;
733 $error = 'No page found: '.$idPart;
734 }
735 }
736 }
737
738 // Setting the A-tag:
739 $bTag = '<a href="'.htmlspecialchars($href).'"'.
740 ($tagCode[2]&&$tagCode[2]!='-' ? ' target="'.htmlspecialchars($tagCode[2]).'"' : '').
741 ($tagCode[3]&&$tagCode[3]!='-' ? ' class="'.htmlspecialchars($tagCode[3]).'"' : '').
742 ($tagCode[4] ? ' title="'.htmlspecialchars($tagCode[4]).'"' : '').
743 ($error ? ' rteerror="'.htmlspecialchars($error).'" style="background-color: yellow; border:2px red solid; color: black;"' : ''). // Should be OK to add the style; the transformation back to databsae will remove it...
744 '>';
745 $eTag = '</a>';
746 $blockSplit[$k] = $bTag.$this->TS_links_rte($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
747 }
748 }
749
750 // Return content:
751 return implode('',$blockSplit);
752 }
753
754 /**
755 * Preserve special tags
756 *
757 * @param string Content input
758 * @return string Content output
759 */
760 function TS_preserve_db($value) {
761 if (!$this->preserveTags) return $value;
762
763 // Splitting into blocks for processing (span-tags are used for special tags)
764 $blockSplit = $this->splitIntoBlock('span',$value);
765 foreach($blockSplit as $k => $v) {
766 if ($k%2) { // block:
767 $attribArray=$this->get_tag_attributes_classic($this->getFirstTag($v));
768 if ($attribArray['specialtag']) {
769 $theTag = rawurldecode($attribArray['specialtag']);
770 $theTagName = $this->getFirstTagName($theTag);
771 $blockSplit[$k] = $theTag.$this->removeFirstAndLastTag($blockSplit[$k]).'</'.$theTagName.'>';
772 }
773 }
774 }
775 return implode('',$blockSplit);
776 }
777
778 /**
779 * Preserve special tags
780 *
781 * @param string Content input
782 * @return string Content output
783 */
784 function TS_preserve_rte($value) {
785 if (!$this->preserveTags) return $value;
786
787 $blockSplit = $this->splitIntoBlock($this->preserveTags,$value);
788 foreach($blockSplit as $k => $v) {
789 if ($k%2) { // block:
790 $blockSplit[$k] = '<span specialtag="'.rawurlencode($this->getFirstTag($v)).'">'.$this->removeFirstAndLastTag($blockSplit[$k]).'</span>';
791 }
792 }
793 return implode('',$blockSplit);
794 }
795
796 /**
797 * Transformation handler: 'ts_transform' + 'css_transform' / direction: "db"
798 * Cleaning (->db) for standard content elements (ts)
799 *
800 * @param string Content input
801 * @param boolean If true, the transformation was "css_transform", otherwise "ts_transform"
802 * @return string Content output
803 * @see TS_transform_rte()
804 */
805 function TS_transform_db($value,$css=FALSE) {
806
807 // safety... so forever loops are avoided (they should not occur, but an error would potentially do this...)
808 $this->TS_transform_db_safecounter--;
809 if ($this->TS_transform_db_safecounter<0) return $value;
810
811 // Split the content from RTE by the occurence of these blocks:
812 $blockSplit = $this->splitIntoBlock('TABLE,BLOCKQUOTE,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$value);
813
814 $cc=0;
815 $aC = count($blockSplit);
816
817 // Avoid superfluous linebreaks by transform_db after ending headListTag
818 while($aC && !strcmp(trim($blockSplit[$aC-1]),'')) {
819 unset($blockSplit[$aC-1]);
820 $aC = count($blockSplit);
821 }
822
823 // Traverse the blocks
824 foreach($blockSplit as $k => $v) {
825 $cc++;
826 $lastBR = $cc==$aC ? '' : chr(10);
827
828 if ($k%2) { // Inside block:
829
830 // Init:
831 $tag=$this->getFirstTag($v);
832 $tagName=strtolower($this->getFirstTagName($v));
833
834 // Process based on the tag:
835 switch($tagName) {
836 case 'blockquote': // Keep blockquotes, but clean the inside recursively in the same manner as the main code
837 case 'dd' : // Do the same on dd elements
838 case 'div': // Do the same on div sections, if they were splitted
839 $blockSplit[$k]=$tag.$this->TS_transform_db($this->removeFirstAndLastTag($blockSplit[$k]),$css).'</'.$tagName.'>'.$lastBR;
840 break;
841 case 'ol':
842 case 'ul': // Transform lists into <typolist>-tags:
843 if (!$css) {
844 if (!isset($this->procOptions['typolist']) || $this->procOptions['typolist']) {
845 $parts = $this->getAllParts($this->splitIntoBlock('LI',$this->removeFirstAndLastTag($blockSplit[$k])),1,0);
846 while(list($k2)=each($parts)) {
847 $parts[$k2]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/','',$parts[$k2]); // remove all linesbreaks!
848 $parts[$k2]=$this->defaultTStagMapping($parts[$k2],'db');
849 $parts[$k2]=$this->cleanFontTags($parts[$k2],0,0,0);
850 $parts[$k2] = $this->HTMLcleaner_db($parts[$k2],strtolower($this->procOptions['allowTagsInTypolists']?$this->procOptions['allowTagsInTypolists']:'br,font,b,i,u,a,img,span,strong,em'));
851 }
852 if ($tagName=='ol') { $params=' type="1"'; } else { $params=''; }
853 $blockSplit[$k]='<typolist'.$params.'>'.chr(10).implode(chr(10),$parts).chr(10).'</typolist>'.$lastBR;
854 }
855 } else {
856 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
857 }
858 break;
859 case 'table': // Tables are NOT allowed in any form (unless preserveTables is set or CSS is the mode)
860 if (!$this->procOptions['preserveTables'] && !$css) {
861 $blockSplit[$k]=$this->TS_transform_db($this->removeTables($blockSplit[$k]));
862 } else {
863 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
864 }
865 break;
866 case 'h1':
867 case 'h2':
868 case 'h3':
869 case 'h4':
870 case 'h5':
871 case 'h6':
872 if (!$css) {
873 $attribArray=$this->get_tag_attributes_classic($tag);
874 // Processing inner content here:
875 $innerContent = $this->HTMLcleaner_db($this->removeFirstAndLastTag($blockSplit[$k]));
876
877 if (!isset($this->procOptions['typohead']) || $this->procOptions['typohead']) {
878 $type = intval(substr($tagName,1));
879 $blockSplit[$k]='<typohead'.
880 ($type!=6?' type="'.$type.'"':'').
881 ($attribArray['align']?' align="'.$attribArray['align'].'"':'').
882 ($attribArray['class']?' class="'.$attribArray['class'].'"':'').
883 '>'.
884 $innerContent.
885 '</typohead>'.
886 $lastBR;
887 } else {
888 $blockSplit[$k]='<'.$tagName.
889 ($attribArray['align']?' align="'.htmlspecialchars($attribArray['align']).'"':'').
890 ($attribArray['class']?' class="'.htmlspecialchars($attribArray['class']).'"':'').
891 '>'.
892 $innerContent.
893 '</'.$tagName.'>'.
894 $lastBR;
895 }
896 } else {
897 // Eliminate true linebreaks inside Hx tags
898 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
899 }
900 break;
901 default:
902 // Eliminate true linebreaks inside other headlist tags and after hr tag
903 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
904 break;
905 }
906 } else { // NON-block:
907 if (strcmp(trim($blockSplit[$k]),'')) {
908 $blockSplit[$k]=$this->divideIntoLines(preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$blockSplit[$k])).$lastBR;
909 $blockSplit[$k]=$this->transformStyledATags($blockSplit[$k]);
910 } else unset($blockSplit[$k]);
911 }
912 }
913 $this->TS_transform_db_safecounter++;
914
915 return implode('',$blockSplit);
916 }
917
918 /**
919 * Wraps a-tags that contain a style attribute with a span-tag
920 *
921 * @param string Content input
922 * @return string Content output
923 */
924 function transformStyledATags($value) {
925 $blockSplit = $this->splitIntoBlock('A',$value);
926 foreach($blockSplit as $k => $v) {
927 if ($k%2) { // If an A-tag was found:
928 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
929 if ($attribArray['style']) { // If "style" attribute is set!
930 $attribArray_copy['style'] = $attribArray['style'];
931 unset($attribArray['style']);
932 $bTag='<span '.t3lib_div::implodeAttributes($attribArray_copy,1).'><a '.t3lib_div::implodeAttributes($attribArray,1).'>';
933 $eTag='</a></span>';
934 $blockSplit[$k] = $bTag.$this->removeFirstAndLastTag($blockSplit[$k]).$eTag;
935 }
936 }
937 }
938 return implode('',$blockSplit);
939 }
940
941 /**
942 * Transformation handler: 'ts_transform' + 'css_transform' / direction: "rte"
943 * Set (->rte) for standard content elements (ts)
944 *
945 * @param string Content input
946 * @param boolean If true, the transformation was "css_transform", otherwise "ts_transform"
947 * @return string Content output
948 * @see TS_transform_db()
949 */
950 function TS_transform_rte($value,$css=0) {
951
952 // Split the content from Database by the occurence of these blocks:
953 $blockSplit = $this->splitIntoBlock('TABLE,BLOCKQUOTE,TYPOLIST,TYPOHEAD,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$value);
954
955 // Traverse the blocks
956 foreach($blockSplit as $k => $v) {
957 if ($k%2) { // Inside one of the blocks:
958
959 // Init:
960 $tag = $this->getFirstTag($v);
961 $tagName = strtolower($this->getFirstTagName($v));
962 $attribArray = $this->get_tag_attributes_classic($tag);
963
964 // Based on tagname, we do transformations:
965 switch($tagName) {
966 case 'blockquote': // Keep blockquotes
967 case 'dd': // Keep definitions
968 case 'div': // Keep div sections, if they were splitted
969 $blockSplit[$k] = $tag.
970 $this->TS_transform_rte($this->removeFirstAndLastTag($blockSplit[$k]),$css).
971 '</'.$tagName.'>';
972 break;
973 case 'typolist': // Transform typolist blocks into OL/UL lists. Type 1 is expected to be numerical block
974 if (!isset($this->procOptions['typolist']) || $this->procOptions['typolist']) {
975 $tListContent = $this->removeFirstAndLastTag($blockSplit[$k]);
976 $tListContent = preg_replace('/^[ ]*'.chr(10).'/','',$tListContent);
977 $tListContent = preg_replace('/'.chr(10).'[ ]*$/','',$tListContent);
978 $lines = explode(chr(10),$tListContent);
979 $typ = $attribArray['type']==1 ? 'ol' : 'ul';
980 $blockSplit[$k] = '<'.$typ.'>'.chr(10).
981 '<li>'.implode('</li>'.chr(10).'<li>',$lines).'</li>'.
982 '</'.$typ.'>';
983 }
984 break;
985 case 'typohead': // Transform typohead into Hx tags.
986 if (!isset($this->procOptions['typohead']) || $this->procOptions['typohead']) {
987 $tC = $this->removeFirstAndLastTag($blockSplit[$k]);
988 $typ = t3lib_div::intInRange($attribArray['type'],0,6);
989 if (!$typ) $typ=6;
990 $align = $attribArray['align']?' align="'.$attribArray['align'].'"': '';
991 $class = $attribArray['class']?' class="'.$attribArray['class'].'"': '';
992 $blockSplit[$k] = '<h'.$typ.$align.$class.'>'.
993 $tC.
994 '</h'.$typ.'>';
995 }
996 break;
997 }
998 $blockSplit[$k+1] = preg_replace('/^[ ]*'.chr(10).'/','',$blockSplit[$k+1]); // Removing linebreak if typohead
999 } else { // NON-block:
1000 $nextFTN = $this->getFirstTagName($blockSplit[$k+1]);
1001 $singleLineBreak = $blockSplit[$k]==chr(10);
1002 if (t3lib_div::inList('TABLE,BLOCKQUOTE,TYPOLIST,TYPOHEAD,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$nextFTN)) { // Removing linebreak if typolist/typohead
1003 $blockSplit[$k] = preg_replace('/'.chr(10).'[ ]*$/','',$blockSplit[$k]);
1004 }
1005 // If $blockSplit[$k] is blank then unset the line. UNLESS the line happend to be a single line break.
1006 if (!strcmp($blockSplit[$k],'') && !$singleLineBreak) {
1007 unset($blockSplit[$k]);
1008 } else {
1009 $blockSplit[$k] = $this->setDivTags($blockSplit[$k],($this->procOptions['useDIVasParagraphTagForRTE']?'div':'p'));
1010 }
1011 }
1012 }
1013 return implode(chr(10),$blockSplit);
1014 }
1015
1016 /**
1017 * Transformation handler: 'ts_strip' / direction: "db"
1018 * Removing all non-allowed tags
1019 *
1020 * @param string Content input
1021 * @return string Content output
1022 */
1023 function TS_strip_db($value) {
1024 $value = strip_tags($value,'<'.implode('><',explode(',','b,i,u,a,img,br,div,center,pre,font,hr,sub,sup,p,strong,em,li,ul,ol,blockquote')).'>');
1025 return $value;
1026 }
1027
1028
1029
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041 /***************************************************************
1042 *
1043 * Generic RTE transformation, analysis and helper functions
1044 *
1045 **************************************************************/
1046
1047 /**
1048 * Reads the file or url $url and returns the content
1049 *
1050 * @param string Filepath/URL to read
1051 * @return string The content from the resource given as input.
1052 * @see t3lib_div::getURL()
1053 */
1054 function getURL($url) {
1055 return t3lib_div::getURL($url);
1056 }
1057
1058 /**
1059 * Function for cleaning content going into the database.
1060 * Content is cleaned eg. by removing unallowed HTML and ds-HSC content
1061 * It is basically calling HTMLcleaner from the parent class with some preset configuration specifically set up for cleaning content going from the RTE into the db
1062 *
1063 * @param string Content to clean up
1064 * @param string Comma list of tags to specifically allow. Default comes from getKeepTags and is ""
1065 * @return string Clean content
1066 * @see getKeepTags()
1067 */
1068 function HTMLcleaner_db($content,$tagList='') {
1069 if (!$tagList) {
1070 $keepTags = $this->getKeepTags('db');
1071 } else {
1072 $keepTags = $this->getKeepTags('db',$tagList);
1073 }
1074 $kUknown = $this->procOptions['dontRemoveUnknownTags_db'] ? 1 : 0; // Default: remove unknown tags.
1075 $hSC = $this->procOptions['dontUndoHSC_db'] ? 0 : -1; // Default: re-convert literals to characters (that is &lt; to <)
1076
1077 // Create additional configuration in order to honor the setting RTE.default.proc.HTMLparser_db.xhtml_cleaning=1
1078 $addConfig=array();
1079 if ((is_array($this->procOptions['HTMLparser_db.']) && $this->procOptions['HTMLparser_db.']['xhtml_cleaning']) || (is_array($this->procOptions['entryHTMLparser_db.']) && $this->procOptions['entryHTMLparser_db.']['xhtml_cleaning']) || (is_array($this->procOptions['exitHTMLparser_db.']) && $this->procOptions['exitHTMLparser_db.']['xhtml_cleaning'])) {
1080 $addConfig['xhtml']=1;
1081 }
1082
1083 return $this->HTMLcleaner($content,$keepTags,$kUknown,$hSC,$addConfig);
1084 }
1085
1086 /**
1087 * Creates an array of configuration for the HTMLcleaner function based on whether content go TO or FROM the Rich Text Editor ($direction)
1088 * Unless "tagList" is given, the function will cache the configuration for next time processing goes on. (In this class that is the case only if we are processing a bulletlist)
1089 *
1090 * @param string The direction of the content being processed by the output configuration; "db" (content going into the database FROM the rte) or "rte" (content going into the form)
1091 * @param string Comma list of tags to keep (overriding default which is to keep all + take notice of internal configuration)
1092 * @return array Configuration array
1093 * @see HTMLcleaner_db()
1094 */
1095 function getKeepTags($direction='rte',$tagList='') {
1096 if (!is_array($this->getKeepTags_cache[$direction]) || $tagList) {
1097
1098 // Setting up allowed tags:
1099 if (strcmp($tagList,'')) { // If the $tagList input var is set, this will take precedence
1100 $keepTags = array_flip(t3lib_div::trimExplode(',',$tagList,1));
1101 } else { // Default is to get allowed/denied tags from internal array of processing options:
1102 // Construct default list of tags to keep:
1103 $typoScript_list = 'b,i,u,a,img,br,div,center,pre,font,hr,sub,sup,p,strong,em,li,ul,ol,blockquote,strike,span';
1104 $keepTags = array_flip(t3lib_div::trimExplode(',',$typoScript_list.','.strtolower($this->procOptions['allowTags']),1));
1105
1106 // For tags to deny, remove them from $keepTags array:
1107 $denyTags = t3lib_div::trimExplode(',',$this->procOptions['denyTags'],1);
1108 foreach($denyTags as $dKe) {
1109 unset($keepTags[$dKe]);
1110 }
1111 }
1112
1113 // Based on the direction of content, set further options:
1114 switch ($direction) {
1115
1116 // GOING from database to Rich Text Editor:
1117 case 'rte':
1118
1119 if (!isset($this->procOptions['transformBoldAndItalicTags']) || $this->procOptions['transformBoldAndItalicTags']) {
1120 // Transform bold/italics tags to strong/em
1121 if (isset($keepTags['b'])) {$keepTags['b']=array('remap'=>'STRONG');}
1122 if (isset($keepTags['i'])) {$keepTags['i']=array('remap'=>'EM');}
1123 }
1124
1125 // Transforming keepTags array so it can be understood by the HTMLcleaner function. This basically converts the format of the array from TypoScript (having .'s) to plain multi-dimensional array.
1126 list($keepTags) = $this->HTMLparserConfig($this->procOptions['HTMLparser_rte.'],$keepTags);
1127 break;
1128
1129 // GOING from RTE to database:
1130 case 'db':
1131
1132 if (!isset($this->procOptions['transformBoldAndItalicTags']) || $this->procOptions['transformBoldAndItalicTags']) {
1133 // Transform strong/em back to bold/italics:
1134 if (isset($keepTags['strong'])) { $keepTags['strong']=array('remap'=>'b'); }
1135 if (isset($keepTags['em'])) { $keepTags['em']=array('remap'=>'i'); }
1136 }
1137
1138 // Setting up span tags if they are allowed:
1139 if (isset($keepTags['span'])) {
1140 $classes=array_merge(array(''),$this->allowedClasses);
1141 $keepTags['span']=array(
1142 'allowedAttribs' => 'id,class,style,title,lang,xml:lang,dir',
1143 'fixAttrib' => Array(
1144 'class' => Array (
1145 'list' => $classes,
1146 'removeIfFalse' => 1
1147 ),
1148 ),
1149 'rmTagIfNoAttrib' => 1
1150 );
1151 if (!$this->procOptions['allowedClasses']) unset($keepTags['span']['fixAttrib']['class']['list']);
1152 }
1153
1154 // Setting up font tags if they are allowed:
1155 if (isset($keepTags['font'])) {
1156 $colors=array_merge(array(''),t3lib_div::trimExplode(',',$this->procOptions['allowedFontColors'],1));
1157 $keepTags['font']=array(
1158 'allowedAttribs'=>'face,color,size',
1159 'fixAttrib' => Array(
1160 'face' => Array (
1161 'removeIfFalse' => 1
1162 ),
1163 'color' => Array (
1164 'removeIfFalse' => 1,
1165 'list'=>$colors
1166 ),
1167 'size' => Array (
1168 'removeIfFalse' => 1,
1169 )
1170 ),
1171 'rmTagIfNoAttrib' => 1
1172 );
1173 if (!$this->procOptions['allowedFontColors']) unset($keepTags['font']['fixAttrib']['color']['list']);
1174 }
1175
1176 // Setting further options, getting them from the processiong options:
1177 $TSc = $this->procOptions['HTMLparser_db.'];
1178 if (!$TSc['globalNesting']) $TSc['globalNesting']='b,i,u,a,center,font,sub,sup,strong,em,strike,span';
1179 if (!$TSc['noAttrib']) $TSc['noAttrib']='b,i,u,br,center,hr,sub,sup,strong,em,li,ul,ol,blockquote,strike';
1180
1181 // Transforming the array from TypoScript to regular array:
1182 list($keepTags) = $this->HTMLparserConfig($TSc,$keepTags);
1183 break;
1184 }
1185
1186 // Caching (internally, in object memory) the result unless tagList is set:
1187 if (!$tagList) {
1188 $this->getKeepTags_cache[$direction] = $keepTags;
1189 } else {
1190 return $keepTags;
1191 }
1192 }
1193
1194 // Return result:
1195 return $this->getKeepTags_cache[$direction];
1196 }
1197
1198 /**
1199 * This resolves the $value into parts based on <div></div>-sections and <p>-sections and <br />-tags. These are returned as lines separated by chr(10).
1200 * This point is to resolve the HTML-code returned from RTE into ordinary lines so it's 'human-readable'
1201 * The function ->setDivTags does the opposite.
1202 * This function processes content to go into the database.
1203 *
1204 * @param string Value to process.
1205 * @param integer Recursion brake. Decremented on each recursion down to zero. Default is 5 (which equals the allowed nesting levels of p/div tags).
1206 * @param boolean If true, an array with the lines is returned, otherwise a string of the processed input value.
1207 * @return string Processed input value.
1208 * @see setDivTags()
1209 */
1210 function divideIntoLines($value,$count=5,$returnArray=FALSE) {
1211
1212 // Internalize font tags (move them from OUTSIDE p/div to inside it that is the case):
1213 if ($this->procOptions['internalizeFontTags']) {$value = $this->internalizeFontTags($value);}
1214
1215 // Setting configuration for processing:
1216 $allowTagsOutside = t3lib_div::trimExplode(',',strtolower($this->procOptions['allowTagsOutside']?$this->procOptions['allowTagsOutside']:'img'),1);
1217 $remapParagraphTag = strtoupper($this->procOptions['remapParagraphTag']);
1218 $divSplit = $this->splitIntoBlock('div,p',$value,1); // Setting the third param to 1 will eliminate false end-tags. Maybe this is a good thing to do...?
1219
1220 if ($this->procOptions['keepPDIVattribs']) {
1221 $keepAttribListArr = t3lib_div::trimExplode(',',strtolower($this->procOptions['keepPDIVattribs']),1);
1222 } else {
1223 $keepAttribListArr = array();
1224 }
1225
1226 // Returns plainly the value if there was no div/p sections in it
1227 if (count($divSplit)<=1 || $count<=0) {
1228 return $value;
1229 }
1230
1231 // Traverse the splitted sections:
1232 foreach($divSplit as $k => $v) {
1233 if ($k%2) { // Inside
1234 $v=$this->removeFirstAndLastTag($v);
1235
1236 // Fetching 'sub-lines' - which will explode any further p/div nesting...
1237 $subLines = $this->divideIntoLines($v,$count-1,1);
1238 if (is_array($subLines)) { // So, if there happend to be sub-nesting of p/div, this is written directly as the new content of THIS section. (This would be considered 'an error')
1239 // No noting.
1240 } else { //... but if NO subsection was found, we process it as a TRUE line without erronous content:
1241 $subLines = array($subLines);
1242 if (!$this->procOptions['dontConvBRtoParagraph']) { // process break-tags, if configured for. Simply, the breaktags will here be treated like if each was a line of content...
1243 $subLines = spliti('<br[[:space:]]*[\/]?>',$v);
1244 }
1245
1246 // Traverse sublines (there is typically one, except if <br/> has been converted to lines as well!)
1247 reset($subLines);
1248 while(list($sk)=each($subLines)) {
1249
1250 // Clear up the subline for DB.
1251 $subLines[$sk]=$this->HTMLcleaner_db($subLines[$sk]);
1252
1253 // Get first tag, attributes etc:
1254 $fTag = $this->getFirstTag($divSplit[$k]);
1255 $tagName=strtolower($this->getFirstTagName($divSplit[$k]));
1256 $attribs=$this->get_tag_attributes($fTag);
1257
1258 // Keep attributes (lowercase)
1259 $newAttribs=array();
1260 if (count($keepAttribListArr)) {
1261 foreach($keepAttribListArr as $keepA) {
1262 if (isset($attribs[0][$keepA])) { $newAttribs[$keepA] = $attribs[0][$keepA]; }
1263 }
1264 }
1265
1266 // ALIGN attribute:
1267 if (!$this->procOptions['skipAlign'] && strcmp(trim($attribs[0]['align']),'') && strtolower($attribs[0]['align'])!='left') { // Set to value, but not 'left'
1268 $newAttribs['align']=strtolower($attribs[0]['align']);
1269 }
1270
1271 // CLASS attribute:
1272 if (!$this->procOptions['skipClass'] && strcmp(trim($attribs[0]['class']),'')) { // Set to whatever value
1273 if (!count($this->allowedClasses) || in_array($attribs[0]['class'], $this->allowedClasses)) {
1274 $newAttribs['class'] = $attribs[0]['class'];
1275 } else {
1276 $classes = t3lib_div::trimExplode(' ', $attribs[0]['class'], true);
1277 $newClasses = array();
1278 foreach ($classes as $class) {
1279 if (in_array($class, $this->allowedClasses)) {
1280 $newClasses[] = $class;
1281 }
1282 }
1283 if (count($newClasses)) {
1284 $newAttribs['class'] = implode(' ', $newClasses);
1285 }
1286 }
1287 }
1288
1289 // Remove any line break char (10 or 13)
1290 $subLines[$sk]=preg_replace('/'.chr(10).'|'.chr(13).'/','',$subLines[$sk]);
1291
1292 // If there are any attributes or if we are supposed to remap the tag, then do so:
1293 if (count($newAttribs) && strcmp($remapParagraphTag,'1')) {
1294 if ($remapParagraphTag=='P') $tagName='p';
1295 if ($remapParagraphTag=='DIV') $tagName='div';
1296 $subLines[$sk]='<'.trim($tagName.' '.$this->compileTagAttribs($newAttribs)).'>'.$subLines[$sk].'</'.$tagName.'>';
1297 }
1298 }
1299 }
1300 // Add the processed line(s)
1301 $divSplit[$k] = implode(chr(10),$subLines);
1302
1303 // If it turns out the line is just blank (containing a &nbsp; possibly) then just make it pure blank.
1304 // But, prevent filtering of lines that are blank in sense above, but whose tags contain attributes.
1305 // Those attributes should have been filtered before; if they are still there they must be considered as possible content.
1306 if (trim(strip_tags($divSplit[$k]))=='&nbsp;' && !preg_match('/\<(img)(\s[^>]*)?\/?>/si', $divSplit[$k]) && !preg_match('/\<([^>]*)?( align| class| style| id| title| dir| lang| xml:lang)([^>]*)?>/si', trim($divSplit[$k]))) {
1307 $divSplit[$k]='';
1308 }
1309 } else { // outside div:
1310 // Remove positions which are outside div/p tags and without content
1311 $divSplit[$k]=trim(strip_tags($divSplit[$k],'<'.implode('><',$allowTagsOutside).'>'));
1312 if (!strcmp($divSplit[$k],'')) unset($divSplit[$k]); // Remove part if it's empty
1313 }
1314 }
1315
1316 // Return value:
1317 return $returnArray ? $divSplit : implode(chr(10),$divSplit);
1318 }
1319
1320 /**
1321 * Converts all lines into <div></div>/<p></p>-sections (unless the line is a div-section already)
1322 * For processing of content going FROM database TO RTE.
1323 *
1324 * @param string Value to convert
1325 * @param string Tag to wrap with. Either "p" or "div" should it be. Lowercase preferably.
1326 * @return string Processed value.
1327 * @see divideIntoLines()
1328 */
1329 function setDivTags($value,$dT='p') {
1330
1331 // First, setting configuration for the HTMLcleaner function. This will process each line between the <div>/<p> section on their way to the RTE
1332 $keepTags = $this->getKeepTags('rte');
1333 $kUknown = $this->procOptions['dontProtectUnknownTags_rte'] ? 0 : 'protect'; // Default: remove unknown tags.
1334 $hSC = $this->procOptions['dontHSC_rte'] ? 0 : 1; // Default: re-convert literals to characters (that is &lt; to <)
1335 $convNBSP = !$this->procOptions['dontConvAmpInNBSP_rte']?1:0;
1336
1337 // Divide the content into lines, based on chr(10):
1338 $parts = explode(chr(10),$value);
1339 foreach($parts as $k => $v) {
1340
1341 // Processing of line content:
1342 if (!strcmp(trim($parts[$k]),'')) { // If the line is blank, set it to &nbsp;
1343 $parts[$k]='&nbsp;';
1344 } else { // Clean the line content:
1345 $parts[$k]=$this->HTMLcleaner($parts[$k],$keepTags,$kUknown,$hSC);
1346 if ($convNBSP) $parts[$k]=str_replace('&amp;nbsp;','&nbsp;',$parts[$k]);
1347 }
1348
1349 // Wrapping the line in <$dT> is not already wrapped:
1350 $testStr = strtolower(trim($parts[$k]));
1351 if (substr($testStr,0,4)!='<div' || substr($testStr,-6)!='</div>') {
1352 if (substr($testStr,0,2)!='<p' || substr($testStr,-4)!='</p>') {
1353 // Only set p-tags if there is not already div or p tags:
1354 $parts[$k]='<'.$dT.'>'.$parts[$k].'</'.$dT.'>';
1355 }
1356 }
1357 }
1358
1359 // Implode result:
1360 return implode(chr(10),$parts);
1361 }
1362
1363 /**
1364 * This splits the $value in font-tag chunks.
1365 * If there are any <P>/<DIV> sections inside of them, the font-tag is wrapped AROUND the content INSIDE of the P/DIV sections and the outer font-tag is removed.
1366 * This functions seems to be a good choice for pre-processing content if it has been pasted into the RTE from eg. star-office.
1367 * In that case the font-tags are normally on the OUTSIDE of the sections.
1368 * This function is used by eg. divideIntoLines() if the procesing option 'internalizeFontTags' is set.
1369 *
1370 * @param string Input content
1371 * @return string Output content
1372 * @see divideIntoLines()
1373 */
1374 function internalizeFontTags($value) {
1375
1376 // Splitting into font tag blocks:
1377 $fontSplit = $this->splitIntoBlock('font',$value);
1378
1379 foreach($fontSplit as $k => $v) {
1380 if ($k%2) { // Inside
1381 $fTag = $this->getFirstTag($v); // Fint font-tag
1382
1383 $divSplit_sub = $this->splitIntoBlock('div,p',$this->removeFirstAndLastTag($v),1);
1384 if (count($divSplit_sub)>1) { // If there were div/p sections inside the font-tag, do something about it...
1385 // traverse those sections:
1386 foreach($divSplit_sub as $k2 => $v2) {
1387 if ($k2%2) { // Inside
1388 $div_p = $this->getFirstTag($v2); // Fint font-tag
1389 $div_p_tagname = $this->getFirstTagName($v2); // Fint font-tag
1390 $v2=$this->removeFirstAndLastTag($v2); // ... and remove it from original.
1391 $divSplit_sub[$k2]=$div_p.$fTag.$v2.'</font>'.'</'.$div_p_tagname.'>';
1392 } elseif (trim(strip_tags($v2))) {
1393 $divSplit_sub[$k2]=$fTag.$v2.'</font>';
1394 }
1395 }
1396 $fontSplit[$k]=implode('',$divSplit_sub);
1397 }
1398 }
1399 }
1400
1401 return implode('',$fontSplit);
1402 }
1403
1404 /**
1405 * Returns SiteURL based on thisScript.
1406 *
1407 * @return string Value of t3lib_div::getIndpEnv('TYPO3_SITE_URL');
1408 * @see t3lib_div::getIndpEnv()
1409 */
1410 function siteUrl() {
1411 return t3lib_div::getIndpEnv('TYPO3_SITE_URL');
1412 }
1413
1414 /**
1415 * Return the storage folder of RTE image files.
1416 * Default is $GLOBALS['TYPO3_CONF_VARS']['BE']['RTE_imageStorageDir'] unless something else is configured in the types configuration for the RTE.
1417 *
1418 * @return string
1419 */
1420 function rteImageStorageDir() {
1421 return $this->rte_p['imgpath'] ? $this->rte_p['imgpath'] : $GLOBALS['TYPO3_CONF_VARS']['BE']['RTE_imageStorageDir'];
1422 }
1423
1424 /**
1425 * Remove all tables from incoming code
1426 * The function is trying to to this is some more or less respectfull way. The approach is to resolve each table cells content and implode it all by <br /> chars. Thus at least the content is preserved in some way.
1427 *
1428 * @param string Input value
1429 * @param string Break character to use for linebreaks.
1430 * @return string Output value
1431 */
1432 function removeTables($value,$breakChar='<br />') {
1433
1434 // Splitting value into table blocks:
1435 $tableSplit = $this->splitIntoBlock('table',$value);
1436
1437 // Traverse blocks of tables:
1438 foreach($tableSplit as $k => $v) {
1439 if ($k%2) {
1440 $tableSplit[$k]='';
1441 $rowSplit = $this->splitIntoBlock('tr',$v);
1442 foreach($rowSplit as $k2 => $v2) {
1443 if ($k2%2) {
1444 $cellSplit = $this->getAllParts($this->splitIntoBlock('td',$v2),1,0);
1445 foreach($cellSplit as $k3 => $v3) {
1446 $tableSplit[$k].=$v3.$breakChar;
1447 }
1448 }
1449 }
1450 }
1451 }
1452
1453 // Implode it all again:
1454 return implode($breakChar,$tableSplit);
1455 }
1456
1457 /**
1458 * Default tag mapping for TS
1459 *
1460 * @param string Input code to process
1461 * @param string Direction To databsae (db) or from database to RTE (rte)
1462 * @return string Processed value
1463 */
1464 function defaultTStagMapping($code,$direction='rte') {
1465 if ($direction=='db') {
1466 $code=$this->mapTags($code,array( // Map tags
1467 'strong' => 'b',
1468 'em' => 'i'
1469 ));
1470 }
1471 if ($direction=='rte') {
1472 $code=$this->mapTags($code,array( // Map tags
1473 'b' => 'strong',
1474 'i' => 'em'
1475 ));
1476 }
1477 return $code;
1478 }
1479
1480 /**
1481 * Finds width and height from attrib-array
1482 * If the width and height is found in the style-attribute, use that!
1483 *
1484 * @param array Array of attributes from tag in which to search. More specifically the content of the key "style" is used to extract "width:xxx / height:xxx" information
1485 * @return array Integer w/h in key 0/1. Zero is returned if not found.
1486 */
1487 function getWHFromAttribs($attribArray) {
1488 $style =trim($attribArray['style']);
1489 if ($style) {
1490 $regex='[[:space:]]*:[[:space:]]*([0-9]*)[[:space:]]*px';
1491 // Width
1492 $reg = array();
1493 preg_match('/width'.$regex.'/i',$style,$reg);
1494 $w = intval($reg[1]);
1495 // Height
1496 preg_match('/height'.$regex.'/i',$style,$reg);
1497 $h = intval($reg[1]);
1498 }
1499 if (!$w) {
1500 $w = $attribArray['width'];
1501 }
1502 if (!$h) {
1503 $h = $attribArray['height'];
1504 }
1505 return array(intval($w),intval($h));
1506 }
1507
1508 /**
1509 * Parse <A>-tag href and return status of email,external,file or page
1510 *
1511 * @param string URL to analyse.
1512 * @return array Information in an array about the URL
1513 */
1514 function urlInfoForLinkTags($url) {
1515 $info = array();
1516 $url = trim($url);
1517 if (substr(strtolower($url),0,7)=='mailto:') {
1518 $info['url']=trim(substr($url,7));
1519 $info['type']='email';
1520 } else {
1521 $curURL = $this->siteUrl(); // 100502, removed this: 'http://'.t3lib_div::getThisUrl(); Reason: The url returned had typo3/ in the end - should be only the site's url as far as I see...
1522 for($a=0;$a<strlen($url);$a++) {
1523 if ($url{$a}!=$curURL{$a}) {
1524 break;
1525 }
1526 }
1527
1528 $info['relScriptPath']=substr($curURL,$a);
1529 $info['relUrl']=substr($url,$a);
1530 $info['url']=$url;
1531 $info['type']='ext';
1532
1533 $siteUrl_parts = parse_url($url);
1534 $curUrl_parts = parse_url($curURL);
1535
1536 if ($siteUrl_parts['host']==$curUrl_parts['host'] // Hosts should match
1537 && (!$info['relScriptPath'] || (defined('TYPO3_mainDir') && substr($info['relScriptPath'],0,strlen(TYPO3_mainDir))==TYPO3_mainDir))) { // If the script path seems to match or is empty (FE-EDIT)
1538
1539 // New processing order 100502
1540 $uP=parse_url($info['relUrl']);
1541
1542 if (!strcmp('#'.$siteUrl_parts['fragment'],$info['relUrl'])) {
1543 $info['url']=$info['relUrl'];
1544 $info['type']='anchor';
1545 } elseif (!trim($uP['path']) || !strcmp($uP['path'],'index.php')) {
1546 $pp = explode('id=',$uP['query']);
1547 $parameters = explode('&', $pp[1]);
1548 $id = array_shift($parameters);
1549 if ($id) {
1550 $info['pageid']=$id;
1551 $info['cElement']=$uP['fragment'];
1552 $info['url']=$id.($info['cElement']?'#'.$info['cElement']:'');
1553 $info['type']='page';
1554 $info['query'] = $parameters[0]?'&'.implode('&', $parameters):'';
1555 }
1556 } else {
1557 $info['url']=$info['relUrl'];
1558 $info['type']='file';
1559 }
1560 } else {
1561 unset($info['relScriptPath']);
1562 unset($info['relUrl']);
1563 }
1564 }
1565 return $info;
1566 }
1567
1568 /**
1569 * Converting <A>-tags to absolute URLs (+ setting rtekeep attribute)
1570 *
1571 * @param string Content input
1572 * @param boolean If true, then the "rtekeep" attribute will not be set.
1573 * @return string Content output
1574 */
1575 function TS_AtagToAbs($value,$dontSetRTEKEEP=FALSE) {
1576 $blockSplit = $this->splitIntoBlock('A',$value);
1577 reset($blockSplit);
1578 while(list($k,$v)=each($blockSplit)) {
1579 if ($k%2) { // block:
1580 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
1581
1582 // Checking if there is a scheme, and if not, prepend the current url.
1583 if (strlen($attribArray['href'])) { // ONLY do this if href has content - the <a> tag COULD be an anchor and if so, it should be preserved...
1584 $uP = parse_url(strtolower($attribArray['href']));
1585 if (!$uP['scheme']) {
1586 $attribArray['href'] = $this->siteUrl().substr($attribArray['href'],strlen($this->relBackPath));
1587 }
1588 } else {
1589 $attribArray['rtekeep'] = 1;
1590 }
1591 if (!$dontSetRTEKEEP) $attribArray['rtekeep'] = 1;
1592
1593 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
1594 $eTag='</a>';
1595 $blockSplit[$k] = $bTag.$this->TS_AtagToAbs($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
1596 }
1597 }
1598 return implode('',$blockSplit);
1599 }
1600 }
1601
1602
1603 if (defined('TYPO3_MODE') && $TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['t3lib/class.t3lib_parsehtml_proc.php']) {
1604 include_once($TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['t3lib/class.t3lib_parsehtml_proc.php']);
1605 }
1606
1607 ?>