Added Feature #12066: Added constants throughout the core for LF, CR, TAB (Thanks...
[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 $value = str_replace(CRLF,LF,$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 $value = str_replace(CR,'',$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 $value = str_replace(CR,'',$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 $value = str_replace(CRLF,LF,$value); // Make sure no \r\n sequences has entered in the meantime...
365 $value = str_replace(LF,CRLF,$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 // Quoting class and title attributes if they contain spaces
652 $attribArray['class'] = preg_match('/ /', $attribArray['class']) ? '"' . $attribArray['class'] . '"' : $attribArray['class'];
653 $attribArray['title'] = preg_match('/ /', $attribArray['title']) ? '"' . $attribArray['title'] . '"' : $attribArray['title'];
654 // Creating the TYPO3 pseudo-tag "<LINK>" for the link (includes href/url, target and class attributes):
655 $bTag='<link '.$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']:'').'>';
656 $eTag='</link>';
657 $blockSplit[$k] = $bTag.$this->TS_links_db($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
658 } else { // ... otherwise store the link as a-tag.
659 // Unsetting 'rtekeep' attribute if that had been set.
660 unset($attribArray['rtekeep']);
661 // If the url is local, remove url-prefix
662 $siteURL = $this->siteUrl();
663 if ($siteURL && substr($attribArray['href'],0,strlen($siteURL))==$siteURL) {
664 $attribArray['href']=$this->relBackPath.substr($attribArray['href'],strlen($siteURL));
665 }
666 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
667 $eTag='</a>';
668 $blockSplit[$k] = $bTag.$this->TS_links_db($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
669 }
670 }
671 }
672 return implode('',$blockSplit);
673 }
674
675 /**
676 * Transformation handler: 'ts_links' / direction: "rte"
677 * Converting <link tags> to <A>-tags
678 *
679 * @param string Content input
680 * @return string Content output
681 * @see TS_links_rte()
682 */
683 function TS_links_rte($value) {
684 $value = $this->TS_AtagToAbs($value);
685
686 // Split content by the TYPO3 pseudo tag "<link>":
687 $blockSplit = $this->splitIntoBlock('link',$value,1);
688 foreach($blockSplit as $k => $v) {
689 $error = '';
690 if ($k%2) { // block:
691 $tagCode = t3lib_div::unQuoteFilenames(trim(substr($this->getFirstTag($v),0,-1)),true);
692 $link_param = $tagCode[1];
693 $href = '';
694 $siteUrl = $this->siteUrl();
695 // Parsing the typolink data. This parsing is roughly done like in tslib_content->typolink()
696 if(strstr($link_param,'@')) { // mailadr
697 $href = 'mailto:'.preg_replace('/^mailto:/i','',$link_param);
698 } elseif (substr($link_param,0,1)=='#') { // check if anchor
699 $href = $siteUrl.$link_param;
700 } else {
701 $fileChar=intval(strpos($link_param, '/'));
702 $urlChar=intval(strpos($link_param, '.'));
703
704 // Detects if a file is found in site-root OR is a simulateStaticDocument.
705 list($rootFileDat) = explode('?',$link_param);
706 $rFD_fI = pathinfo($rootFileDat);
707 if (trim($rootFileDat) && !strstr($link_param,'/') && (@is_file(PATH_site.$rootFileDat) || t3lib_div::inList('php,html,htm',strtolower($rFD_fI['extension'])))) {
708 $href = $siteUrl.$link_param;
709 } elseif($urlChar && (strstr($link_param,'//') || !$fileChar || $urlChar<$fileChar)) { // url (external): If doubleSlash or if a '.' comes before a '/'.
710 if (!preg_match('/^[a-z]*:\/\//',trim(strtolower($link_param)))) {$scheme='http://';} else {$scheme='';}
711 $href = $scheme.$link_param;
712 } elseif($fileChar) { // file (internal)
713 $href = $siteUrl.$link_param;
714 } else { // integer or alias (alias is without slashes or periods or commas, that is 'nospace,alphanum_x,lower,unique' according to tables.php!!)
715 // Splitting the parameter by ',' and if the array counts more than 1 element it's a id/type/parameters triplet
716 $pairParts = t3lib_div::trimExplode(',', $link_param, TRUE);
717 $idPart = $pairParts[0];
718 $link_params_parts = explode('#', $idPart);
719 $idPart = trim($link_params_parts[0]);
720 $sectionMark = trim($link_params_parts[1]);
721 if (!strcmp($idPart,'')) { $idPart=$this->recPid; } // If no id or alias is given, set it to class record pid
722 // Checking if the id-parameter is an alias.
723 if (!t3lib_div::testInt($idPart)) {
724 list($idPartR) = t3lib_BEfunc::getRecordsByField('pages','alias',$idPart);
725 $idPart = intval($idPartR['uid']);
726 }
727 $page = t3lib_BEfunc::getRecord('pages', $idPart);
728 if (is_array($page)) { // Page must exist...
729 $href = $siteUrl .'?id=' . $idPart . ($pairParts[2] ? $pairParts[2] : '') . ($sectionMark ? '#' . $sectionMark : '');
730 // linkHandler - allowing links to start with registerd linkHandler e.g.. "record:"
731 } elseif (isset($GLOBALS['TYPO3_CONF_VARS']['SC_OPTIONS']['tslib/class.tslib_content.php']['typolinkLinkHandler'][array_shift(explode(':', $link_param))])) {
732 $href = $link_param;
733 } else {
734 #$href = '';
735 $href = $siteUrl.'?id='.$link_param;
736 $error = 'No page found: '.$idPart;
737 }
738 }
739 }
740
741 // Setting the A-tag:
742 $bTag = '<a href="'.htmlspecialchars($href).'"'.
743 ($tagCode[2]&&$tagCode[2]!='-' ? ' target="'.htmlspecialchars($tagCode[2]).'"' : '').
744 ($tagCode[3]&&$tagCode[3]!='-' ? ' class="'.htmlspecialchars($tagCode[3]).'"' : '').
745 ($tagCode[4] ? ' title="'.htmlspecialchars($tagCode[4]).'"' : '').
746 ($error ? ' rteerror="'.htmlspecialchars($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...
747 '>';
748 $eTag = '</a>';
749 $blockSplit[$k] = $bTag.$this->TS_links_rte($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
750 }
751 }
752
753 // Return content:
754 return implode('',$blockSplit);
755 }
756
757 /**
758 * Preserve special tags
759 *
760 * @param string Content input
761 * @return string Content output
762 */
763 function TS_preserve_db($value) {
764 if (!$this->preserveTags) return $value;
765
766 // Splitting into blocks for processing (span-tags are used for special tags)
767 $blockSplit = $this->splitIntoBlock('span',$value);
768 foreach($blockSplit as $k => $v) {
769 if ($k%2) { // block:
770 $attribArray=$this->get_tag_attributes_classic($this->getFirstTag($v));
771 if ($attribArray['specialtag']) {
772 $theTag = rawurldecode($attribArray['specialtag']);
773 $theTagName = $this->getFirstTagName($theTag);
774 $blockSplit[$k] = $theTag.$this->removeFirstAndLastTag($blockSplit[$k]).'</'.$theTagName.'>';
775 }
776 }
777 }
778 return implode('',$blockSplit);
779 }
780
781 /**
782 * Preserve special tags
783 *
784 * @param string Content input
785 * @return string Content output
786 */
787 function TS_preserve_rte($value) {
788 if (!$this->preserveTags) return $value;
789
790 $blockSplit = $this->splitIntoBlock($this->preserveTags,$value);
791 foreach($blockSplit as $k => $v) {
792 if ($k%2) { // block:
793 $blockSplit[$k] = '<span specialtag="'.rawurlencode($this->getFirstTag($v)).'">'.$this->removeFirstAndLastTag($blockSplit[$k]).'</span>';
794 }
795 }
796 return implode('',$blockSplit);
797 }
798
799 /**
800 * Transformation handler: 'ts_transform' + 'css_transform' / direction: "db"
801 * Cleaning (->db) for standard content elements (ts)
802 *
803 * @param string Content input
804 * @param boolean If true, the transformation was "css_transform", otherwise "ts_transform"
805 * @return string Content output
806 * @see TS_transform_rte()
807 */
808 function TS_transform_db($value,$css=FALSE) {
809
810 // safety... so forever loops are avoided (they should not occur, but an error would potentially do this...)
811 $this->TS_transform_db_safecounter--;
812 if ($this->TS_transform_db_safecounter<0) return $value;
813
814 // Split the content from RTE by the occurence of these blocks:
815 $blockSplit = $this->splitIntoBlock('TABLE,BLOCKQUOTE,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$value);
816
817 $cc=0;
818 $aC = count($blockSplit);
819
820 // Avoid superfluous linebreaks by transform_db after ending headListTag
821 while($aC && !strcmp(trim($blockSplit[$aC-1]),'')) {
822 unset($blockSplit[$aC-1]);
823 $aC = count($blockSplit);
824 }
825
826 // Traverse the blocks
827 foreach($blockSplit as $k => $v) {
828 $cc++;
829 $lastBR = $cc==$aC ? '' : LF;
830
831 if ($k%2) { // Inside block:
832
833 // Init:
834 $tag=$this->getFirstTag($v);
835 $tagName=strtolower($this->getFirstTagName($v));
836
837 // Process based on the tag:
838 switch($tagName) {
839 case 'blockquote': // Keep blockquotes, but clean the inside recursively in the same manner as the main code
840 case 'dd' : // Do the same on dd elements
841 case 'div': // Do the same on div sections, if they were splitted
842 $blockSplit[$k]=$tag.$this->TS_transform_db($this->removeFirstAndLastTag($blockSplit[$k]),$css).'</'.$tagName.'>'.$lastBR;
843 break;
844 case 'ol':
845 case 'ul': // Transform lists into <typolist>-tags:
846 if (!$css) {
847 if (!isset($this->procOptions['typolist']) || $this->procOptions['typolist']) {
848 $parts = $this->getAllParts($this->splitIntoBlock('LI',$this->removeFirstAndLastTag($blockSplit[$k])),1,0);
849 while(list($k2)=each($parts)) {
850 $parts[$k2]=preg_replace('/['.preg_quote(LF.CR).']+/','',$parts[$k2]); // remove all linesbreaks!
851 $parts[$k2]=$this->defaultTStagMapping($parts[$k2],'db');
852 $parts[$k2]=$this->cleanFontTags($parts[$k2],0,0,0);
853 $parts[$k2] = $this->HTMLcleaner_db($parts[$k2],strtolower($this->procOptions['allowTagsInTypolists']?$this->procOptions['allowTagsInTypolists']:'br,font,b,i,u,a,img,span,strong,em'));
854 }
855 if ($tagName=='ol') { $params=' type="1"'; } else { $params=''; }
856 $blockSplit[$k]='<typolist'.$params.'>'.LF.implode(LF,$parts).LF.'</typolist>'.$lastBR;
857 }
858 } else {
859 $blockSplit[$k]=preg_replace('/['.preg_quote(LF.CR).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
860 }
861 break;
862 case 'table': // Tables are NOT allowed in any form (unless preserveTables is set or CSS is the mode)
863 if (!$this->procOptions['preserveTables'] && !$css) {
864 $blockSplit[$k]=$this->TS_transform_db($this->removeTables($blockSplit[$k]));
865 } else {
866 $blockSplit[$k]=preg_replace('/['.preg_quote(LF.CR).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
867 }
868 break;
869 case 'h1':
870 case 'h2':
871 case 'h3':
872 case 'h4':
873 case 'h5':
874 case 'h6':
875 if (!$css) {
876 $attribArray=$this->get_tag_attributes_classic($tag);
877 // Processing inner content here:
878 $innerContent = $this->HTMLcleaner_db($this->removeFirstAndLastTag($blockSplit[$k]));
879
880 if (!isset($this->procOptions['typohead']) || $this->procOptions['typohead']) {
881 $type = intval(substr($tagName,1));
882 $blockSplit[$k]='<typohead'.
883 ($type!=6?' type="'.$type.'"':'').
884 ($attribArray['align']?' align="'.$attribArray['align'].'"':'').
885 ($attribArray['class']?' class="'.$attribArray['class'].'"':'').
886 '>'.
887 $innerContent.
888 '</typohead>'.
889 $lastBR;
890 } else {
891 $blockSplit[$k]='<'.$tagName.
892 ($attribArray['align']?' align="'.htmlspecialchars($attribArray['align']).'"':'').
893 ($attribArray['class']?' class="'.htmlspecialchars($attribArray['class']).'"':'').
894 '>'.
895 $innerContent.
896 '</'.$tagName.'>'.
897 $lastBR;
898 }
899 } else {
900 // Eliminate true linebreaks inside Hx tags
901 $blockSplit[$k]=preg_replace('/['.preg_quote(LF.CR).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
902 }
903 break;
904 default:
905 // Eliminate true linebreaks inside other headlist tags and after hr tag
906 $blockSplit[$k]=preg_replace('/['.preg_quote(LF.CR).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
907 break;
908 }
909 } else { // NON-block:
910 if (strcmp(trim($blockSplit[$k]),'')) {
911 $blockSplit[$k]=$this->divideIntoLines(preg_replace('/['.preg_quote(LF.CR).']+/',' ',$blockSplit[$k])).$lastBR;
912 $blockSplit[$k]=$this->transformStyledATags($blockSplit[$k]);
913 } else unset($blockSplit[$k]);
914 }
915 }
916 $this->TS_transform_db_safecounter++;
917
918 return implode('',$blockSplit);
919 }
920
921 /**
922 * Wraps a-tags that contain a style attribute with a span-tag
923 *
924 * @param string Content input
925 * @return string Content output
926 */
927 function transformStyledATags($value) {
928 $blockSplit = $this->splitIntoBlock('A',$value);
929 foreach($blockSplit as $k => $v) {
930 if ($k%2) { // If an A-tag was found:
931 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
932 if ($attribArray['style']) { // If "style" attribute is set!
933 $attribArray_copy['style'] = $attribArray['style'];
934 unset($attribArray['style']);
935 $bTag='<span '.t3lib_div::implodeAttributes($attribArray_copy,1).'><a '.t3lib_div::implodeAttributes($attribArray,1).'>';
936 $eTag='</a></span>';
937 $blockSplit[$k] = $bTag.$this->removeFirstAndLastTag($blockSplit[$k]).$eTag;
938 }
939 }
940 }
941 return implode('',$blockSplit);
942 }
943
944 /**
945 * Transformation handler: 'ts_transform' + 'css_transform' / direction: "rte"
946 * Set (->rte) for standard content elements (ts)
947 *
948 * @param string Content input
949 * @param boolean If true, the transformation was "css_transform", otherwise "ts_transform"
950 * @return string Content output
951 * @see TS_transform_db()
952 */
953 function TS_transform_rte($value,$css=0) {
954
955 // Split the content from Database by the occurence of these blocks:
956 $blockSplit = $this->splitIntoBlock('TABLE,BLOCKQUOTE,TYPOLIST,TYPOHEAD,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$value);
957
958 // Traverse the blocks
959 foreach($blockSplit as $k => $v) {
960 if ($k%2) { // Inside one of the blocks:
961
962 // Init:
963 $tag = $this->getFirstTag($v);
964 $tagName = strtolower($this->getFirstTagName($v));
965 $attribArray = $this->get_tag_attributes_classic($tag);
966
967 // Based on tagname, we do transformations:
968 switch($tagName) {
969 case 'blockquote': // Keep blockquotes
970 case 'dd': // Keep definitions
971 case 'div': // Keep div sections, if they were splitted
972 $blockSplit[$k] = $tag.
973 $this->TS_transform_rte($this->removeFirstAndLastTag($blockSplit[$k]),$css).
974 '</'.$tagName.'>';
975 break;
976 case 'typolist': // Transform typolist blocks into OL/UL lists. Type 1 is expected to be numerical block
977 if (!isset($this->procOptions['typolist']) || $this->procOptions['typolist']) {
978 $tListContent = $this->removeFirstAndLastTag($blockSplit[$k]);
979 $tListContent = preg_replace('/^[ ]*'.LF.'/','',$tListContent);
980 $tListContent = preg_replace('/'.LF.'[ ]*$/','',$tListContent);
981 $lines = explode(LF,$tListContent);
982 $typ = $attribArray['type']==1 ? 'ol' : 'ul';
983 $blockSplit[$k] = '<'.$typ.'>'.LF.
984 '<li>'.implode('</li>'.LF.'<li>',$lines).'</li>'.
985 '</'.$typ.'>';
986 }
987 break;
988 case 'typohead': // Transform typohead into Hx tags.
989 if (!isset($this->procOptions['typohead']) || $this->procOptions['typohead']) {
990 $tC = $this->removeFirstAndLastTag($blockSplit[$k]);
991 $typ = t3lib_div::intInRange($attribArray['type'],0,6);
992 if (!$typ) $typ=6;
993 $align = $attribArray['align']?' align="'.$attribArray['align'].'"': '';
994 $class = $attribArray['class']?' class="'.$attribArray['class'].'"': '';
995 $blockSplit[$k] = '<h'.$typ.$align.$class.'>'.
996 $tC.
997 '</h'.$typ.'>';
998 }
999 break;
1000 }
1001 $blockSplit[$k+1] = preg_replace('/^[ ]*'.LF.'/','',$blockSplit[$k+1]); // Removing linebreak if typohead
1002 } else { // NON-block:
1003 $nextFTN = $this->getFirstTagName($blockSplit[$k+1]);
1004 $singleLineBreak = $blockSplit[$k]==LF;
1005 if (t3lib_div::inList('TABLE,BLOCKQUOTE,TYPOLIST,TYPOHEAD,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$nextFTN)) { // Removing linebreak if typolist/typohead
1006 $blockSplit[$k] = preg_replace('/'.LF.'[ ]*$/','',$blockSplit[$k]);
1007 }
1008 // If $blockSplit[$k] is blank then unset the line. UNLESS the line happend to be a single line break.
1009 if (!strcmp($blockSplit[$k],'') && !$singleLineBreak) {
1010 unset($blockSplit[$k]);
1011 } else {
1012 $blockSplit[$k] = $this->setDivTags($blockSplit[$k],($this->procOptions['useDIVasParagraphTagForRTE']?'div':'p'));
1013 }
1014 }
1015 }
1016 return implode(LF,$blockSplit);
1017 }
1018
1019 /**
1020 * Transformation handler: 'ts_strip' / direction: "db"
1021 * Removing all non-allowed tags
1022 *
1023 * @param string Content input
1024 * @return string Content output
1025 */
1026 function TS_strip_db($value) {
1027 $value = strip_tags($value,'<'.implode('><',explode(',','b,i,u,a,img,br,div,center,pre,font,hr,sub,sup,p,strong,em,li,ul,ol,blockquote')).'>');
1028 return $value;
1029 }
1030
1031
1032
1033
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044 /***************************************************************
1045 *
1046 * Generic RTE transformation, analysis and helper functions
1047 *
1048 **************************************************************/
1049
1050 /**
1051 * Reads the file or url $url and returns the content
1052 *
1053 * @param string Filepath/URL to read
1054 * @return string The content from the resource given as input.
1055 * @see t3lib_div::getURL()
1056 */
1057 function getURL($url) {
1058 return t3lib_div::getURL($url);
1059 }
1060
1061 /**
1062 * Function for cleaning content going into the database.
1063 * Content is cleaned eg. by removing unallowed HTML and ds-HSC content
1064 * It is basically calling HTMLcleaner from the parent class with some preset configuration specifically set up for cleaning content going from the RTE into the db
1065 *
1066 * @param string Content to clean up
1067 * @param string Comma list of tags to specifically allow. Default comes from getKeepTags and is ""
1068 * @return string Clean content
1069 * @see getKeepTags()
1070 */
1071 function HTMLcleaner_db($content,$tagList='') {
1072 if (!$tagList) {
1073 $keepTags = $this->getKeepTags('db');
1074 } else {
1075 $keepTags = $this->getKeepTags('db',$tagList);
1076 }
1077 $kUknown = $this->procOptions['dontRemoveUnknownTags_db'] ? 1 : 0; // Default: remove unknown tags.
1078 $hSC = $this->procOptions['dontUndoHSC_db'] ? 0 : -1; // Default: re-convert literals to characters (that is &lt; to <)
1079
1080 // Create additional configuration in order to honor the setting RTE.default.proc.HTMLparser_db.xhtml_cleaning=1
1081 $addConfig=array();
1082 if ((is_array($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'])) {
1083 $addConfig['xhtml']=1;
1084 }
1085
1086 return $this->HTMLcleaner($content,$keepTags,$kUknown,$hSC,$addConfig);
1087 }
1088
1089 /**
1090 * Creates an array of configuration for the HTMLcleaner function based on whether content go TO or FROM the Rich Text Editor ($direction)
1091 * Unless "tagList" is given, the function will cache the configuration for next time processing goes on. (In this class that is the case only if we are processing a bulletlist)
1092 *
1093 * @param string The direction of the content being processed by the output configuration; "db" (content going into the database FROM the rte) or "rte" (content going into the form)
1094 * @param string Comma list of tags to keep (overriding default which is to keep all + take notice of internal configuration)
1095 * @return array Configuration array
1096 * @see HTMLcleaner_db()
1097 */
1098 function getKeepTags($direction='rte',$tagList='') {
1099 if (!is_array($this->getKeepTags_cache[$direction]) || $tagList) {
1100
1101 // Setting up allowed tags:
1102 if (strcmp($tagList,'')) { // If the $tagList input var is set, this will take precedence
1103 $keepTags = array_flip(t3lib_div::trimExplode(',',$tagList,1));
1104 } else { // Default is to get allowed/denied tags from internal array of processing options:
1105 // Construct default list of tags to keep:
1106 $typoScript_list = 'b,i,u,a,img,br,div,center,pre,font,hr,sub,sup,p,strong,em,li,ul,ol,blockquote,strike,span';
1107 $keepTags = array_flip(t3lib_div::trimExplode(',',$typoScript_list.','.strtolower($this->procOptions['allowTags']),1));
1108
1109 // For tags to deny, remove them from $keepTags array:
1110 $denyTags = t3lib_div::trimExplode(',',$this->procOptions['denyTags'],1);
1111 foreach($denyTags as $dKe) {
1112 unset($keepTags[$dKe]);
1113 }
1114 }
1115
1116 // Based on the direction of content, set further options:
1117 switch ($direction) {
1118
1119 // GOING from database to Rich Text Editor:
1120 case 'rte':
1121
1122 if (!isset($this->procOptions['transformBoldAndItalicTags']) || $this->procOptions['transformBoldAndItalicTags']) {
1123 // Transform bold/italics tags to strong/em
1124 if (isset($keepTags['b'])) {$keepTags['b']=array('remap'=>'STRONG');}
1125 if (isset($keepTags['i'])) {$keepTags['i']=array('remap'=>'EM');}
1126 }
1127
1128 // Transforming keepTags array so it can be understood by the HTMLcleaner function. This basically converts the format of the array from TypoScript (having .'s) to plain multi-dimensional array.
1129 list($keepTags) = $this->HTMLparserConfig($this->procOptions['HTMLparser_rte.'],$keepTags);
1130 break;
1131
1132 // GOING from RTE to database:
1133 case 'db':
1134
1135 if (!isset($this->procOptions['transformBoldAndItalicTags']) || $this->procOptions['transformBoldAndItalicTags']) {
1136 // Transform strong/em back to bold/italics:
1137 if (isset($keepTags['strong'])) { $keepTags['strong']=array('remap'=>'b'); }
1138 if (isset($keepTags['em'])) { $keepTags['em']=array('remap'=>'i'); }
1139 }
1140
1141 // Setting up span tags if they are allowed:
1142 if (isset($keepTags['span'])) {
1143 $classes=array_merge(array(''),$this->allowedClasses);
1144 $keepTags['span']=array(
1145 'allowedAttribs' => 'id,class,style,title,lang,xml:lang,dir',
1146 'fixAttrib' => Array(
1147 'class' => Array (
1148 'list' => $classes,
1149 'removeIfFalse' => 1
1150 ),
1151 ),
1152 'rmTagIfNoAttrib' => 1
1153 );
1154 if (!$this->procOptions['allowedClasses']) unset($keepTags['span']['fixAttrib']['class']['list']);
1155 }
1156
1157 // Setting up font tags if they are allowed:
1158 if (isset($keepTags['font'])) {
1159 $colors=array_merge(array(''),t3lib_div::trimExplode(',',$this->procOptions['allowedFontColors'],1));
1160 $keepTags['font']=array(
1161 'allowedAttribs'=>'face,color,size',
1162 'fixAttrib' => Array(
1163 'face' => Array (
1164 'removeIfFalse' => 1
1165 ),
1166 'color' => Array (
1167 'removeIfFalse' => 1,
1168 'list'=>$colors
1169 ),
1170 'size' => Array (
1171 'removeIfFalse' => 1,
1172 )
1173 ),
1174 'rmTagIfNoAttrib' => 1
1175 );
1176 if (!$this->procOptions['allowedFontColors']) unset($keepTags['font']['fixAttrib']['color']['list']);
1177 }
1178
1179 // Setting further options, getting them from the processiong options:
1180 $TSc = $this->procOptions['HTMLparser_db.'];
1181 if (!$TSc['globalNesting']) $TSc['globalNesting']='b,i,u,a,center,font,sub,sup,strong,em,strike,span';
1182 if (!$TSc['noAttrib']) $TSc['noAttrib']='b,i,u,br,center,hr,sub,sup,strong,em,li,ul,ol,blockquote,strike';
1183
1184 // Transforming the array from TypoScript to regular array:
1185 list($keepTags) = $this->HTMLparserConfig($TSc,$keepTags);
1186 break;
1187 }
1188
1189 // Caching (internally, in object memory) the result unless tagList is set:
1190 if (!$tagList) {
1191 $this->getKeepTags_cache[$direction] = $keepTags;
1192 } else {
1193 return $keepTags;
1194 }
1195 }
1196
1197 // Return result:
1198 return $this->getKeepTags_cache[$direction];
1199 }
1200
1201 /**
1202 * This resolves the $value into parts based on <div></div>-sections and <p>-sections and <br />-tags. These are returned as lines separated by LF.
1203 * This point is to resolve the HTML-code returned from RTE into ordinary lines so it's 'human-readable'
1204 * The function ->setDivTags does the opposite.
1205 * This function processes content to go into the database.
1206 *
1207 * @param string Value to process.
1208 * @param integer Recursion brake. Decremented on each recursion down to zero. Default is 5 (which equals the allowed nesting levels of p/div tags).
1209 * @param boolean If true, an array with the lines is returned, otherwise a string of the processed input value.
1210 * @return string Processed input value.
1211 * @see setDivTags()
1212 */
1213 function divideIntoLines($value,$count=5,$returnArray=FALSE) {
1214
1215 // Internalize font tags (move them from OUTSIDE p/div to inside it that is the case):
1216 if ($this->procOptions['internalizeFontTags']) {$value = $this->internalizeFontTags($value);}
1217
1218 // Setting configuration for processing:
1219 $allowTagsOutside = t3lib_div::trimExplode(',',strtolower($this->procOptions['allowTagsOutside']?$this->procOptions['allowTagsOutside']:'img'),1);
1220 $remapParagraphTag = strtoupper($this->procOptions['remapParagraphTag']);
1221 $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...?
1222
1223 if ($this->procOptions['keepPDIVattribs']) {
1224 $keepAttribListArr = t3lib_div::trimExplode(',',strtolower($this->procOptions['keepPDIVattribs']),1);
1225 } else {
1226 $keepAttribListArr = array();
1227 }
1228
1229 // Returns plainly the value if there was no div/p sections in it
1230 if (count($divSplit)<=1 || $count<=0) {
1231 return $value;
1232 }
1233
1234 // Traverse the splitted sections:
1235 foreach($divSplit as $k => $v) {
1236 if ($k%2) { // Inside
1237 $v=$this->removeFirstAndLastTag($v);
1238
1239 // Fetching 'sub-lines' - which will explode any further p/div nesting...
1240 $subLines = $this->divideIntoLines($v,$count-1,1);
1241 if (is_array($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')
1242 // No noting.
1243 } else { //... but if NO subsection was found, we process it as a TRUE line without erronous content:
1244 $subLines = array($subLines);
1245 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...
1246 $subLines = preg_split('/<br[[:space:]]*[\/]?>/i', $v);
1247 }
1248
1249 // Traverse sublines (there is typically one, except if <br/> has been converted to lines as well!)
1250 reset($subLines);
1251 while(list($sk)=each($subLines)) {
1252
1253 // Clear up the subline for DB.
1254 $subLines[$sk]=$this->HTMLcleaner_db($subLines[$sk]);
1255
1256 // Get first tag, attributes etc:
1257 $fTag = $this->getFirstTag($divSplit[$k]);
1258 $tagName=strtolower($this->getFirstTagName($divSplit[$k]));
1259 $attribs=$this->get_tag_attributes($fTag);
1260
1261 // Keep attributes (lowercase)
1262 $newAttribs=array();
1263 if (count($keepAttribListArr)) {
1264 foreach($keepAttribListArr as $keepA) {
1265 if (isset($attribs[0][$keepA])) { $newAttribs[$keepA] = $attribs[0][$keepA]; }
1266 }
1267 }
1268
1269 // ALIGN attribute:
1270 if (!$this->procOptions['skipAlign'] && strcmp(trim($attribs[0]['align']),'') && strtolower($attribs[0]['align'])!='left') { // Set to value, but not 'left'
1271 $newAttribs['align']=strtolower($attribs[0]['align']);
1272 }
1273
1274 // CLASS attribute:
1275 if (!$this->procOptions['skipClass'] && strcmp(trim($attribs[0]['class']),'')) { // Set to whatever value
1276 if (!count($this->allowedClasses) || in_array($attribs[0]['class'], $this->allowedClasses)) {
1277 $newAttribs['class'] = $attribs[0]['class'];
1278 } else {
1279 $classes = t3lib_div::trimExplode(' ', $attribs[0]['class'], true);
1280 $newClasses = array();
1281 foreach ($classes as $class) {
1282 if (in_array($class, $this->allowedClasses)) {
1283 $newClasses[] = $class;
1284 }
1285 }
1286 if (count($newClasses)) {
1287 $newAttribs['class'] = implode(' ', $newClasses);
1288 }
1289 }
1290 }
1291
1292 // Remove any line break char (10 or 13)
1293 $subLines[$sk]=preg_replace('/'.LF.'|'.CR.'/','',$subLines[$sk]);
1294
1295 // If there are any attributes or if we are supposed to remap the tag, then do so:
1296 if (count($newAttribs) && strcmp($remapParagraphTag,'1')) {
1297 if ($remapParagraphTag=='P') $tagName='p';
1298 if ($remapParagraphTag=='DIV') $tagName='div';
1299 $subLines[$sk]='<'.trim($tagName.' '.$this->compileTagAttribs($newAttribs)).'>'.$subLines[$sk].'</'.$tagName.'>';
1300 }
1301 }
1302 }
1303 // Add the processed line(s)
1304 $divSplit[$k] = implode(LF,$subLines);
1305
1306 // If it turns out the line is just blank (containing a &nbsp; possibly) then just make it pure blank.
1307 // But, prevent filtering of lines that are blank in sense above, but whose tags contain attributes.
1308 // Those attributes should have been filtered before; if they are still there they must be considered as possible content.
1309 if (trim(strip_tags($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]))) {
1310 $divSplit[$k]='';
1311 }
1312 } else { // outside div:
1313 // Remove positions which are outside div/p tags and without content
1314 $divSplit[$k]=trim(strip_tags($divSplit[$k],'<'.implode('><',$allowTagsOutside).'>'));
1315 if (!strcmp($divSplit[$k],'')) unset($divSplit[$k]); // Remove part if it's empty
1316 }
1317 }
1318
1319 // Return value:
1320 return $returnArray ? $divSplit : implode(LF,$divSplit);
1321 }
1322
1323 /**
1324 * Converts all lines into <div></div>/<p></p>-sections (unless the line is a div-section already)
1325 * For processing of content going FROM database TO RTE.
1326 *
1327 * @param string Value to convert
1328 * @param string Tag to wrap with. Either "p" or "div" should it be. Lowercase preferably.
1329 * @return string Processed value.
1330 * @see divideIntoLines()
1331 */
1332 function setDivTags($value,$dT='p') {
1333
1334 // First, setting configuration for the HTMLcleaner function. This will process each line between the <div>/<p> section on their way to the RTE
1335 $keepTags = $this->getKeepTags('rte');
1336 $kUknown = $this->procOptions['dontProtectUnknownTags_rte'] ? 0 : 'protect'; // Default: remove unknown tags.
1337 $hSC = $this->procOptions['dontHSC_rte'] ? 0 : 1; // Default: re-convert literals to characters (that is &lt; to <)
1338 $convNBSP = !$this->procOptions['dontConvAmpInNBSP_rte']?1:0;
1339
1340 // Divide the content into lines, based on LF:
1341 $parts = explode(LF,$value);
1342 foreach($parts as $k => $v) {
1343
1344 // Processing of line content:
1345 if (!strcmp(trim($parts[$k]),'')) { // If the line is blank, set it to &nbsp;
1346 $parts[$k]='&nbsp;';
1347 } else { // Clean the line content:
1348 $parts[$k]=$this->HTMLcleaner($parts[$k],$keepTags,$kUknown,$hSC);
1349 if ($convNBSP) $parts[$k]=str_replace('&amp;nbsp;','&nbsp;',$parts[$k]);
1350 }
1351
1352 // Wrapping the line in <$dT> is not already wrapped:
1353 $testStr = strtolower(trim($parts[$k]));
1354 if (substr($testStr,0,4)!='<div' || substr($testStr,-6)!='</div>') {
1355 if (substr($testStr,0,2)!='<p' || substr($testStr,-4)!='</p>') {
1356 // Only set p-tags if there is not already div or p tags:
1357 $parts[$k]='<'.$dT.'>'.$parts[$k].'</'.$dT.'>';
1358 }
1359 }
1360 }
1361
1362 // Implode result:
1363 return implode(LF,$parts);
1364 }
1365
1366 /**
1367 * This splits the $value in font-tag chunks.
1368 * If there are any <P>/<DIV> sections inside of them, the font-tag is wrapped AROUND the content INSIDE of the P/DIV sections and the outer font-tag is removed.
1369 * This functions seems to be a good choice for pre-processing content if it has been pasted into the RTE from eg. star-office.
1370 * In that case the font-tags are normally on the OUTSIDE of the sections.
1371 * This function is used by eg. divideIntoLines() if the procesing option 'internalizeFontTags' is set.
1372 *
1373 * @param string Input content
1374 * @return string Output content
1375 * @see divideIntoLines()
1376 */
1377 function internalizeFontTags($value) {
1378
1379 // Splitting into font tag blocks:
1380 $fontSplit = $this->splitIntoBlock('font',$value);
1381
1382 foreach($fontSplit as $k => $v) {
1383 if ($k%2) { // Inside
1384 $fTag = $this->getFirstTag($v); // Fint font-tag
1385
1386 $divSplit_sub = $this->splitIntoBlock('div,p',$this->removeFirstAndLastTag($v),1);
1387 if (count($divSplit_sub)>1) { // If there were div/p sections inside the font-tag, do something about it...
1388 // traverse those sections:
1389 foreach($divSplit_sub as $k2 => $v2) {
1390 if ($k2%2) { // Inside
1391 $div_p = $this->getFirstTag($v2); // Fint font-tag
1392 $div_p_tagname = $this->getFirstTagName($v2); // Fint font-tag
1393 $v2=$this->removeFirstAndLastTag($v2); // ... and remove it from original.
1394 $divSplit_sub[$k2]=$div_p.$fTag.$v2.'</font>'.'</'.$div_p_tagname.'>';
1395 } elseif (trim(strip_tags($v2))) {
1396 $divSplit_sub[$k2]=$fTag.$v2.'</font>';
1397 }
1398 }
1399 $fontSplit[$k]=implode('',$divSplit_sub);
1400 }
1401 }
1402 }
1403
1404 return implode('',$fontSplit);
1405 }
1406
1407 /**
1408 * Returns SiteURL based on thisScript.
1409 *
1410 * @return string Value of t3lib_div::getIndpEnv('TYPO3_SITE_URL');
1411 * @see t3lib_div::getIndpEnv()
1412 */
1413 function siteUrl() {
1414 return t3lib_div::getIndpEnv('TYPO3_SITE_URL');
1415 }
1416
1417 /**
1418 * Return the storage folder of RTE image files.
1419 * Default is $GLOBALS['TYPO3_CONF_VARS']['BE']['RTE_imageStorageDir'] unless something else is configured in the types configuration for the RTE.
1420 *
1421 * @return string
1422 */
1423 function rteImageStorageDir() {
1424 return $this->rte_p['imgpath'] ? $this->rte_p['imgpath'] : $GLOBALS['TYPO3_CONF_VARS']['BE']['RTE_imageStorageDir'];
1425 }
1426
1427 /**
1428 * Remove all tables from incoming code
1429 * The function is trying to to this is some more or less respectfull way. The approach is to resolve each table cells content and implode it all by <br /> chars. Thus at least the content is preserved in some way.
1430 *
1431 * @param string Input value
1432 * @param string Break character to use for linebreaks.
1433 * @return string Output value
1434 */
1435 function removeTables($value,$breakChar='<br />') {
1436
1437 // Splitting value into table blocks:
1438 $tableSplit = $this->splitIntoBlock('table',$value);
1439
1440 // Traverse blocks of tables:
1441 foreach($tableSplit as $k => $v) {
1442 if ($k%2) {
1443 $tableSplit[$k]='';
1444 $rowSplit = $this->splitIntoBlock('tr',$v);
1445 foreach($rowSplit as $k2 => $v2) {
1446 if ($k2%2) {
1447 $cellSplit = $this->getAllParts($this->splitIntoBlock('td',$v2),1,0);
1448 foreach($cellSplit as $k3 => $v3) {
1449 $tableSplit[$k].=$v3.$breakChar;
1450 }
1451 }
1452 }
1453 }
1454 }
1455
1456 // Implode it all again:
1457 return implode($breakChar,$tableSplit);
1458 }
1459
1460 /**
1461 * Default tag mapping for TS
1462 *
1463 * @param string Input code to process
1464 * @param string Direction To databsae (db) or from database to RTE (rte)
1465 * @return string Processed value
1466 */
1467 function defaultTStagMapping($code,$direction='rte') {
1468 if ($direction=='db') {
1469 $code=$this->mapTags($code,array( // Map tags
1470 'strong' => 'b',
1471 'em' => 'i'
1472 ));
1473 }
1474 if ($direction=='rte') {
1475 $code=$this->mapTags($code,array( // Map tags
1476 'b' => 'strong',
1477 'i' => 'em'
1478 ));
1479 }
1480 return $code;
1481 }
1482
1483 /**
1484 * Finds width and height from attrib-array
1485 * If the width and height is found in the style-attribute, use that!
1486 *
1487 * @param array Array of attributes from tag in which to search. More specifically the content of the key "style" is used to extract "width:xxx / height:xxx" information
1488 * @return array Integer w/h in key 0/1. Zero is returned if not found.
1489 */
1490 function getWHFromAttribs($attribArray) {
1491 $style =trim($attribArray['style']);
1492 if ($style) {
1493 $regex='[[:space:]]*:[[:space:]]*([0-9]*)[[:space:]]*px';
1494 // Width
1495 $reg = array();
1496 preg_match('/width'.$regex.'/i',$style,$reg);
1497 $w = intval($reg[1]);
1498 // Height
1499 preg_match('/height'.$regex.'/i',$style,$reg);
1500 $h = intval($reg[1]);
1501 }
1502 if (!$w) {
1503 $w = $attribArray['width'];
1504 }
1505 if (!$h) {
1506 $h = $attribArray['height'];
1507 }
1508 return array(intval($w),intval($h));
1509 }
1510
1511 /**
1512 * Parse <A>-tag href and return status of email,external,file or page
1513 *
1514 * @param string URL to analyse.
1515 * @return array Information in an array about the URL
1516 */
1517 function urlInfoForLinkTags($url) {
1518 $info = array();
1519 $url = trim($url);
1520 if (substr(strtolower($url),0,7)=='mailto:') {
1521 $info['url']=trim(substr($url,7));
1522 $info['type']='email';
1523 } else {
1524 $curURL = $this->siteUrl(); // 100502, removed this: 'http://'.t3lib_div::getThisUrl(); Reason: The url returned had typo3/ in the end - should be only the site's url as far as I see...
1525 for($a=0;$a<strlen($url);$a++) {
1526 if ($url{$a}!=$curURL{$a}) {
1527 break;
1528 }
1529 }
1530
1531 $info['relScriptPath']=substr($curURL,$a);
1532 $info['relUrl']=substr($url,$a);
1533 $info['url']=$url;
1534 $info['type']='ext';
1535
1536 $siteUrl_parts = parse_url($url);
1537 $curUrl_parts = parse_url($curURL);
1538
1539 if ($siteUrl_parts['host']==$curUrl_parts['host'] // Hosts should match
1540 && (!$info['relScriptPath'] || (defined('TYPO3_mainDir') && substr($info['relScriptPath'],0,strlen(TYPO3_mainDir))==TYPO3_mainDir))) { // If the script path seems to match or is empty (FE-EDIT)
1541
1542 // New processing order 100502
1543 $uP=parse_url($info['relUrl']);
1544
1545 if (!strcmp('#'.$siteUrl_parts['fragment'],$info['relUrl'])) {
1546 $info['url']=$info['relUrl'];
1547 $info['type']='anchor';
1548 } elseif (!trim($uP['path']) || !strcmp($uP['path'],'index.php')) {
1549 // URL is a page (id parameter)
1550 $pp = preg_split('/^id=/', $uP['query']);
1551 $pp[1] = preg_replace( '/&id=[^&]*/', '', $pp[1]);
1552 $parameters = explode('&', $pp[1]);
1553 $id = array_shift($parameters);
1554 if ($id) {
1555 $info['pageid']=$id;
1556 $info['cElement']=$uP['fragment'];
1557 $info['url']=$id.($info['cElement']?'#'.$info['cElement']:'');
1558 $info['type']='page';
1559 $info['query'] = $parameters[0]?'&'.implode('&', $parameters):'';
1560 }
1561 } else {
1562 $info['url']=$info['relUrl'];
1563 $info['type']='file';
1564 }
1565 } else {
1566 unset($info['relScriptPath']);
1567 unset($info['relUrl']);
1568 }
1569 }
1570 return $info;
1571 }
1572
1573 /**
1574 * Converting <A>-tags to absolute URLs (+ setting rtekeep attribute)
1575 *
1576 * @param string Content input
1577 * @param boolean If true, then the "rtekeep" attribute will not be set.
1578 * @return string Content output
1579 */
1580 function TS_AtagToAbs($value,$dontSetRTEKEEP=FALSE) {
1581 $blockSplit = $this->splitIntoBlock('A',$value);
1582 reset($blockSplit);
1583 while(list($k,$v)=each($blockSplit)) {
1584 if ($k%2) { // block:
1585 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
1586
1587 // Checking if there is a scheme, and if not, prepend the current url.
1588 if (strlen($attribArray['href'])) { // ONLY do this if href has content - the <a> tag COULD be an anchor and if so, it should be preserved...
1589 $uP = parse_url(strtolower($attribArray['href']));
1590 if (!$uP['scheme']) {
1591 $attribArray['href'] = $this->siteUrl().substr($attribArray['href'],strlen($this->relBackPath));
1592 }
1593 } else {
1594 $attribArray['rtekeep'] = 1;
1595 }
1596 if (!$dontSetRTEKEEP) $attribArray['rtekeep'] = 1;
1597
1598 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
1599 $eTag='</a>';
1600 $blockSplit[$k] = $bTag.$this->TS_AtagToAbs($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
1601 }
1602 }
1603 return implode('',$blockSplit);
1604 }
1605 }
1606
1607
1608 if (defined('TYPO3_MODE') && $TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['t3lib/class.t3lib_parsehtml_proc.php']) {
1609 include_once($TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['t3lib/class.t3lib_parsehtml_proc.php']);
1610 }
1611
1612 ?>