Fixed bug #7961: headListTags in class.t3lib_parsehtml_proc.php are hardcoded (Thanks...
[Packages/TYPO3.CMS.git] / t3lib / class.t3lib_parsehtml_proc.php
1 <?php
2 /***************************************************************
3 * Copyright notice
4 *
5 * (c) 1999-2008 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 require_once (PATH_t3lib.'class.t3lib_parsehtml.php');
84
85
86
87
88
89
90
91
92
93
94
95
96 /**
97 * Class for parsing HTML for the Rich Text Editor. (also called transformations)
98 *
99 * @author Kasper Skaarhoj <kasperYYYY@typo3.com>
100 * @package TYPO3
101 * @subpackage t3lib
102 */
103 class t3lib_parsehtml_proc extends t3lib_parsehtml {
104
105 // Static:
106 var $blockElementList = 'PRE,UL,OL,H1,H2,H3,H4,H5,H6,HR,ADDRESS,DL,DD'; // List of tags for these elements
107
108 // Internal, static:
109 var $recPid = 0; // Set this to the pid of the record manipulated by the class.
110 var $elRef = ''; // Element reference [table]:[field], eg. "tt_content:bodytext"
111 var $relPath=''; // Relative path
112 var $relBackPath=''; // Relative back-path
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 $path = ereg_replace('^/','',$path);
153 $path = ereg_replace('/$','',$path);
154 if ($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->procOptions = $thisConfig['proc.'];
236 $this->preserveTags = strtoupper(implode(',',t3lib_div::trimExplode(',',$this->procOptions['preserveTags'])));
237
238 // dynamic configuration of blockElementList
239 if ($this->procOptions['blockElementList']) {
240 $this->blockElementList = $this->procOptions['blockElementList'];
241 }
242
243 // Get parameters for rte_transformation:
244 $p = $this->rte_p = t3lib_BEfunc::getSpecConfParametersFromArray($specConf['rte_transform']['parameters']);
245
246 // Setting modes:
247 if (strcmp($this->procOptions['overruleMode'],'')) {
248 $modes = array_unique(t3lib_div::trimExplode(',',$this->procOptions['overruleMode']));
249 } else {
250 $modes = array_unique(t3lib_div::trimExplode('-',$p['mode']));
251 }
252 $revmodes = array_flip($modes);
253
254 // Find special modes and extract them:
255 if (isset($revmodes['ts'])) {
256 $modes[$revmodes['ts']] = 'ts_transform,ts_preserve,ts_images,ts_links';
257 }
258 // Find special modes and extract them:
259 if (isset($revmodes['ts_css'])) {
260 $modes[$revmodes['ts_css']] = 'css_transform,ts_images,ts_links';
261 }
262
263 // Make list unique
264 $modes = array_unique(t3lib_div::trimExplode(',',implode(',',$modes),1));
265
266 // Reverse order if direction is "rte"
267 if ($direction=='rte') {
268 $modes = array_reverse($modes);
269 }
270
271 // 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:
272 $entry_HTMLparser = $this->procOptions['entryHTMLparser_'.$direction] ? $this->HTMLparserConfig($this->procOptions['entryHTMLparser_'.$direction.'.']) : '';
273 $exit_HTMLparser = $this->procOptions['exitHTMLparser_'.$direction] ? $this->HTMLparserConfig($this->procOptions['exitHTMLparser_'.$direction.'.']) : '';
274
275 // Line breaks of content is unified into char-10 only (removing char 13)
276 if (!$this->procOptions['disableUnifyLineBreaks']) {
277 $value = str_replace(chr(13).chr(10),chr(10),$value);
278 }
279
280 // In an entry-cleaner was configured, pass value through the HTMLcleaner with that:
281 if (is_array($entry_HTMLparser)) {
282 $value = $this->HTMLcleaner($value,$entry_HTMLparser[0],$entry_HTMLparser[1],$entry_HTMLparser[2],$entry_HTMLparser[3]);
283 }
284
285 // Traverse modes:
286 foreach($modes as $cmd) {
287 // ->DB
288 if ($direction=='db') {
289 // Checking for user defined transformation:
290 if ($_classRef = $GLOBALS['TYPO3_CONF_VARS']['SC_OPTIONS']['t3lib/class.t3lib_parsehtml_proc.php']['transformation'][$cmd]) {
291 $_procObj = &t3lib_div::getUserObj($_classRef);
292 $_procObj->pObj = &$this;
293 $_procObj->transformationKey = $cmd;
294 $value = $_procObj->transform_db($value,$this);
295 } else { // ... else use defaults:
296 switch($cmd) {
297 case 'ts_images':
298 $value = $this->TS_images_db($value);
299 break;
300 case 'ts_reglinks':
301 $value = $this->TS_reglinks($value,'db');
302 break;
303 case 'ts_links':
304 $value = $this->TS_links_db($value);
305 break;
306 case 'ts_preserve':
307 $value = $this->TS_preserve_db($value);
308 break;
309 case 'ts_transform':
310 case 'css_transform':
311 $value = str_replace(chr(13),'',$value); // Has a very disturbing effect, so just remove all '13' - depend on '10'
312 $this->allowedClasses = t3lib_div::trimExplode(',',strtoupper($this->procOptions['allowedClasses']),1);
313 $value = $this->TS_transform_db($value,$cmd=='css_transform');
314 break;
315 case 'ts_strip':
316 $value = $this->TS_strip_db($value);
317 break;
318 default:
319 break;
320 }
321 }
322 }
323 // ->RTE
324 if ($direction=='rte') {
325 // Checking for user defined transformation:
326 if ($_classRef = $GLOBALS['TYPO3_CONF_VARS']['SC_OPTIONS']['t3lib/class.t3lib_parsehtml_proc.php']['transformation'][$cmd]) {
327 $_procObj = &t3lib_div::getUserObj($_classRef);
328 $_procObj->pObj = &$this;
329 $value = $_procObj->transform_rte($value,$this);
330 } else { // ... else use defaults:
331 switch($cmd) {
332 case 'ts_images':
333 $value = $this->TS_images_rte($value);
334 break;
335 case 'ts_reglinks':
336 $value = $this->TS_reglinks($value,'rte');
337 break;
338 case 'ts_links':
339 $value = $this->TS_links_rte($value);
340 break;
341 case 'ts_preserve':
342 $value = $this->TS_preserve_rte($value);
343 break;
344 case 'ts_transform':
345 case 'css_transform':
346 $value = str_replace(chr(13),'',$value); // Has a very disturbing effect, so just remove all '13' - depend on '10'
347 $value = $this->TS_transform_rte($value,$cmd=='css_transform');
348 break;
349 default:
350 break;
351 }
352 }
353 }
354 }
355
356 // In an exit-cleaner was configured, pass value through the HTMLcleaner with that:
357 if (is_array($exit_HTMLparser)) {
358 $value = $this->HTMLcleaner($value,$exit_HTMLparser[0],$exit_HTMLparser[1],$exit_HTMLparser[2],$exit_HTMLparser[3]);
359 }
360
361 // Final clean up of linebreaks:
362 if (!$this->procOptions['disableUnifyLineBreaks']) {
363 $value = str_replace(chr(13).chr(10),chr(10),$value); // Make sure no \r\n sequences has entered in the meantime...
364 $value = str_replace(chr(10),chr(13).chr(10),$value); // ... and then change all \n into \r\n
365 }
366
367 // Return value:
368 return $value;
369 }
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386 /************************************
387 *
388 * Specific RTE TRANSFORMATION functions
389 *
390 *************************************/
391
392 /**
393 * Transformation handler: 'ts_images' / direction: "db"
394 * Processing images inserted in the RTE.
395 * This is used when content goes from the RTE to the database.
396 * Images inserted in the RTE has an absolute URL applied to the src attribute. This URL is converted to a relative URL
397 * If it turns out that the URL is from another website than the current the image is read from that external URL and moved to the local server.
398 * Also "magic" images are processed here.
399 *
400 * @param string The content from RTE going to Database
401 * @return string Processed content
402 */
403 function TS_images_db($value) {
404
405 // Split content by <img> tags and traverse the resulting array for processing:
406 $imgSplit = $this->splitTags('img',$value);
407 foreach($imgSplit as $k => $v) {
408 if ($k%2) { // image found, do processing:
409
410 // Init
411 $attribArray = $this->get_tag_attributes_classic($v,1);
412 $siteUrl = $this->siteUrl();
413 $sitePath = str_replace (t3lib_div::getIndpEnv('TYPO3_REQUEST_HOST'), '', $siteUrl);
414
415 $absRef = trim($attribArray['src']); // It's always a absolute URL coming from the RTE into the Database.
416
417 // make path absolute if it is relative and we have a site path wich is not '/'
418 $pI=pathinfo($absRef);
419 if($sitePath AND !$pI['scheme'] && t3lib_div::isFirstPartOfStr($absRef,$sitePath)) {
420 // if site is in a subpath (eg. /~user_jim/) this path needs to be removed because it will be added with $siteUrl
421 $absRef = substr($absRef,strlen($sitePath));
422 $absRef = $siteUrl.$absRef;
423 }
424
425 // External image from another URL? In that case, fetch image (unless disabled feature).
426 if (!t3lib_div::isFirstPartOfStr($absRef,$siteUrl) && !$this->procOptions['dontFetchExtPictures']) {
427 $externalFile = $this->getUrl($absRef); // Get it
428 if ($externalFile) {
429 $pU = parse_url($absRef);
430 $pI=pathinfo($pU['path']);
431
432 if (t3lib_div::inList('gif,png,jpeg,jpg',strtolower($pI['extension']))) {
433 $filename = t3lib_div::shortMD5($absRef).'.'.$pI['extension'];
434 $origFilePath = PATH_site.$this->rteImageStorageDir().'RTEmagicP_'.$filename;
435 $C_origFilePath = PATH_site.$this->rteImageStorageDir().'RTEmagicC_'.$filename.'.'.$pI['extension'];
436 if (!@is_file($origFilePath)) {
437 t3lib_div::writeFile($origFilePath,$externalFile);
438 t3lib_div::writeFile($C_origFilePath,$externalFile);
439 }
440 $absRef = $siteUrl.$this->rteImageStorageDir().'RTEmagicC_'.$filename.'.'.$pI['extension'];
441
442 $attribArray['src']=$absRef;
443 $params = t3lib_div::implodeAttributes($attribArray,1);
444 $imgSplit[$k] = '<img '.$params.' />';
445 }
446 }
447 }
448
449 // Check image as local file (siteURL equals the one of the image)
450 if (t3lib_div::isFirstPartOfStr($absRef,$siteUrl)) {
451 $path = rawurldecode(substr($absRef,strlen($siteUrl))); // Rel-path, rawurldecoded for special characters.
452 $filepath = t3lib_div::getFileAbsFileName($path); // Abs filepath, locked to relative path of this project.
453
454 // Check file existence (in relative dir to this installation!)
455 if ($filepath && @is_file($filepath)) {
456
457 // If "magic image":
458 $pathPre=$this->rteImageStorageDir().'RTEmagicC_';
459 if (t3lib_div::isFirstPartOfStr($path,$pathPre)) {
460 // Find original file:
461 $pI=pathinfo(substr($path,strlen($pathPre)));
462 $filename = substr($pI['basename'],0,-strlen('.'.$pI['extension']));
463 $origFilePath = PATH_site.$this->rteImageStorageDir().'RTEmagicP_'.$filename;
464 if (@is_file($origFilePath)) {
465 $imgObj = t3lib_div::makeInstance('t3lib_stdGraphic');
466 $imgObj->init();
467 $imgObj->mayScaleUp=0;
468 $imgObj->tempPath=PATH_site.$imgObj->tempPath;
469
470 $curInfo = $imgObj->getImageDimensions($filepath); // Image dimensions of the current image
471 $curWH = $this->getWHFromAttribs($attribArray); // Image dimensions as set in the image tag
472 // Compare dimensions:
473 if ($curWH[0]!=$curInfo[0] || $curWH[1]!=$curInfo[1]) {
474 $origImgInfo = $imgObj->getImageDimensions($origFilePath); // Image dimensions of the current image
475 $cW = $curWH[0];
476 $cH = $curWH[1];
477 $cH = 1000; // Make the image based on the width solely...
478 $imgI = $imgObj->imageMagickConvert($origFilePath,$pI['extension'],$cW.'m',$cH.'m');
479 if ($imgI[3]) {
480 $fI=pathinfo($imgI[3]);
481 @copy($imgI[3],$filepath); // Override the child file
482 // Removing width and heigth form style attribute
483 $attribArray['style'] = preg_replace('/((?:^|)\s*(?:width|height)\s*:[^;]*(?:$|;))/si', '', $attribArray['style']);
484 $attribArray['width']=$imgI[0];
485 $attribArray['height']=$imgI[1];
486 $params = t3lib_div::implodeAttributes($attribArray,1);
487 $imgSplit[$k]='<img '.$params.' />';
488 }
489 }
490 }
491
492 } elseif ($this->procOptions['plainImageMode']) { // If "plain image" has been configured:
493
494 // Image dimensions as set in the image tag, if any
495 $curWH = $this->getWHFromAttribs($attribArray);
496 if ($curWH[0]) $attribArray['width'] = $curWH[0];
497 if ($curWH[1]) $attribArray['height'] = $curWH[1];
498
499 // Removing width and heigth form style attribute
500 $attribArray['style'] = preg_replace('/((?:^|)\s*(?:width|height)\s*:[^;]*(?:$|;))/si', '', $attribArray['style']);
501
502 // Finding dimensions of image file:
503 $fI = @getimagesize($filepath);
504
505 // Perform corrections to aspect ratio based on configuration:
506 switch((string)$this->procOptions['plainImageMode']) {
507 case 'lockDimensions':
508 $attribArray['width']=$fI[0];
509 $attribArray['height']=$fI[1];
510 break;
511 case 'lockRatioWhenSmaller': // If the ratio has to be smaller, then first set the width...:
512 if ($attribArray['width']>$fI[0]) $attribArray['width'] = $fI[0];
513 case 'lockRatio':
514 if ($fI[0]>0) {
515 $attribArray['height']=round($attribArray['width']*($fI[1]/$fI[0]));
516 }
517 break;
518 }
519
520 // Compile the image tag again:
521 $params = t3lib_div::implodeAttributes($attribArray,1);
522 $imgSplit[$k]='<img '.$params.' />';
523 }
524 } else { // Remove image if it was not found in a proper position on the server!
525
526 // Commented out; removing the image tag might not be that logical...
527 #$imgSplit[$k]='';
528 }
529 }
530
531 // Convert abs to rel url
532 if ($imgSplit[$k]) {
533 $attribArray=$this->get_tag_attributes_classic($imgSplit[$k],1);
534 $absRef = trim($attribArray['src']);
535 if (t3lib_div::isFirstPartOfStr($absRef,$siteUrl)) {
536 $attribArray['src'] = $this->relBackPath.substr($absRef,strlen($siteUrl));
537 if (!isset($attribArray['alt'])) $attribArray['alt']=''; // Must have alt-attribute for XHTML compliance.
538 $imgSplit[$k]='<img '.t3lib_div::implodeAttributes($attribArray,1,1).' />';
539 }
540 }
541 }
542 }
543 return implode('',$imgSplit);
544 }
545
546 /**
547 * Transformation handler: 'ts_images' / direction: "rte"
548 * Processing images from database content going into the RTE.
549 * Processing includes converting the src attribute to an absolute URL.
550 *
551 * @param string Content input
552 * @return string Content output
553 */
554 function TS_images_rte($value) {
555
556 $siteUrl = $this->siteUrl();
557 $sitePath = str_replace (t3lib_div::getIndpEnv('TYPO3_REQUEST_HOST'), '', $siteUrl);
558
559 // Split content by <img> tags and traverse the resulting array for processing:
560 $imgSplit = $this->splitTags('img',$value);
561 foreach($imgSplit as $k => $v) {
562 if ($k%2) { // image found:
563
564 // Init
565 $attribArray=$this->get_tag_attributes_classic($v,1);
566 $absRef = trim($attribArray['src']);
567
568 // Unless the src attribute is already pointing to an external URL:
569 if (strtolower(substr($absRef,0,4))!='http') {
570 $attribArray['src'] = substr($attribArray['src'],strlen($this->relBackPath));
571 // if site is in a subpath (eg. /~user_jim/) this path needs to be removed because it will be added with $siteUrl
572 $attribArray['src'] = preg_replace('#^'.preg_quote($sitePath,'#').'#','',$attribArray['src']);
573 $attribArray['src'] = $siteUrl.$attribArray['src'];
574 if (!isset($attribArray['alt'])) $attribArray['alt']='';
575 $params = t3lib_div::implodeAttributes($attribArray);
576 $imgSplit[$k]='<img '.$params.' />';
577 }
578 }
579 }
580
581 // return processed content:
582 return implode('',$imgSplit);
583 }
584
585 /**
586 * Transformation handler: 'ts_reglinks' / direction: "db"+"rte" depending on $direction variable.
587 * Converting <A>-tags to/from abs/rel
588 *
589 * @param string Content input
590 * @param string Direction of conversion; "rte" (from database to RTE) or "db" (from RTE to database)
591 * @return string Content output
592 */
593 function TS_reglinks($value,$direction) {
594 $retVal = '';
595
596 switch($direction) {
597 case 'rte':
598 $retVal = $this->TS_AtagToAbs($value,1);
599 break;
600 case 'db':
601 $siteURL = $this->siteUrl();
602 $blockSplit = $this->splitIntoBlock('A',$value);
603 reset($blockSplit);
604 while(list($k,$v)=each($blockSplit)) {
605 if ($k%2) { // block:
606 $attribArray=$this->get_tag_attributes_classic($this->getFirstTag($v),1);
607 // If the url is local, remove url-prefix
608 if ($siteURL && substr($attribArray['href'],0,strlen($siteURL))==$siteURL) {
609 $attribArray['href']=$this->relBackPath.substr($attribArray['href'],strlen($siteURL));
610 }
611 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
612 $eTag='</a>';
613 $blockSplit[$k] = $bTag.$this->TS_reglinks($this->removeFirstAndLastTag($blockSplit[$k]),$direction).$eTag;
614 }
615 }
616 $retVal = implode('',$blockSplit);
617 break;
618 }
619 return $retVal;
620 }
621
622 /**
623 * Transformation handler: 'ts_links' / direction: "db"
624 * Converting <A>-tags to <link tags>
625 *
626 * @param string Content input
627 * @return string Content output
628 * @see TS_links_rte()
629 */
630 function TS_links_db($value) {
631
632 // Split content into <a> tag blocks and process:
633 $blockSplit = $this->splitIntoBlock('A',$value);
634 foreach($blockSplit as $k => $v) {
635 if ($k%2) { // If an A-tag was found:
636 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
637 $info = $this->urlInfoForLinkTags($attribArray['href']);
638
639 // Check options:
640 $attribArray_copy = $attribArray;
641 unset($attribArray_copy['href']);
642 unset($attribArray_copy['target']);
643 unset($attribArray_copy['class']);
644 unset($attribArray_copy['title']);
645 if ($attribArray_copy['rteerror']) { // Unset "rteerror" and "style" attributes if "rteerror" is set!
646 unset($attribArray_copy['style']);
647 unset($attribArray_copy['rteerror']);
648 }
649 if (!count($attribArray_copy)) { // Only if href, target and class are the only attributes, we can alter the link!
650 // Creating the TYPO3 pseudo-tag "<LINK>" for the link (includes href/url, target and class attributes):
651 $bTag='<link '.$info['url'].($attribArray['target']?' '.$attribArray['target']:(($attribArray['class'] || $attribArray['title'])?' -':'')).($attribArray['class']?' '.$attribArray['class']:($attribArray['title']?' -':'')).($attribArray['title']?' "'.$attribArray['title'].'"':'').'>';
652 $eTag='</link>';
653 $blockSplit[$k] = $bTag.$this->TS_links_db($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
654 } else { // ... otherwise store the link as a-tag.
655 // Unsetting 'rtekeep' attribute if that had been set.
656 unset($attribArray['rtekeep']);
657 // If the url is local, remove url-prefix
658 $siteURL = $this->siteUrl();
659 if ($siteURL && substr($attribArray['href'],0,strlen($siteURL))==$siteURL) {
660 $attribArray['href']=$this->relBackPath.substr($attribArray['href'],strlen($siteURL));
661 }
662 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
663 $eTag='</a>';
664 $blockSplit[$k] = $bTag.$this->TS_links_db($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
665 }
666 }
667 }
668 return implode('',$blockSplit);
669 }
670
671 /**
672 * Transformation handler: 'ts_links' / direction: "rte"
673 * Converting <link tags> to <A>-tags
674 *
675 * @param string Content input
676 * @return string Content output
677 * @see TS_links_rte()
678 */
679 function TS_links_rte($value) {
680 $value = $this->TS_AtagToAbs($value);
681
682 // Split content by the TYPO3 pseudo tag "<link>":
683 $blockSplit = $this->splitIntoBlock('link',$value,1);
684 foreach($blockSplit as $k => $v) {
685 $error = '';
686 if ($k%2) { // block:
687 $tagCode = t3lib_div::unQuoteFilenames(trim(substr($this->getFirstTag($v),0,-1)),true);
688 $link_param = $tagCode[1];
689 $href = '';
690 $siteUrl = $this->siteUrl();
691 // Parsing the typolink data. This parsing is roughly done like in tslib_content->typolink()
692 if(strstr($link_param,'@')) { // mailadr
693 $href = 'mailto:'.eregi_replace('^mailto:','',$link_param);
694 } elseif (substr($link_param,0,1)=='#') { // check if anchor
695 $href = $siteUrl.$link_param;
696 } else {
697 $fileChar=intval(strpos($link_param, '/'));
698 $urlChar=intval(strpos($link_param, '.'));
699
700 // Detects if a file is found in site-root OR is a simulateStaticDocument.
701 list($rootFileDat) = explode('?',$link_param);
702 $rFD_fI = pathinfo($rootFileDat);
703 if (trim($rootFileDat) && !strstr($link_param,'/') && (@is_file(PATH_site.$rootFileDat) || t3lib_div::inList('php,html,htm',strtolower($rFD_fI['extension'])))) {
704 $href = $siteUrl.$link_param;
705 } elseif($urlChar && (strstr($link_param,'//') || !$fileChar || $urlChar<$fileChar)) { // url (external): If doubleSlash or if a '.' comes before a '/'.
706 if (!ereg('^[a-z]*://',trim(strtolower($link_param)))) {$scheme='http://';} else {$scheme='';}
707 $href = $scheme.$link_param;
708 } elseif($fileChar) { // file (internal)
709 $href = $siteUrl.$link_param;
710 } else { // integer or alias (alias is without slashes or periods or commas, that is 'nospace,alphanum_x,lower,unique' according to tables.php!!)
711 $link_params_parts = explode('#',$link_param);
712 $idPart = trim($link_params_parts[0]); // Link-data del
713 if (!strcmp($idPart,'')) { $idPart=$this->recPid; } // If no id or alias is given, set it to class record pid
714
715 // FIXME commented because useless - what is it for?
716 // if ($link_params_parts[1] && !$sectionMark) {
717 // $sectionMark = '#'.trim($link_params_parts[1]);
718 // }
719
720 // Splitting the parameter by ',' and if the array counts more than 1 element it's a id/type/? pair
721 $pairParts = t3lib_div::trimExplode(',',$idPart);
722 if (count($pairParts)>1) {
723 $idPart = $pairParts[0];
724 // Type ? future support for?
725 }
726 // Checking if the id-parameter is an alias.
727 if (!t3lib_div::testInt($idPart)) {
728 list($idPartR) = t3lib_BEfunc::getRecordsByField('pages','alias',$idPart);
729 $idPart = intval($idPartR['uid']);
730 }
731 $page = t3lib_BEfunc::getRecord('pages', $idPart);
732 if (is_array($page)) { // Page must exist...
733 $href = $siteUrl.'?id='.$link_param;
734 } else if(strtolower(substr($link_param, 0, 7)) == 'record:') {
735 // linkHandler - allowing links to start with "record:"
736 $href = $link_param;
737 } else {
738 #$href = '';
739 $href = $siteUrl.'?id='.$link_param;
740 $error = 'No page found: '.$idPart;
741 }
742 }
743 }
744
745 // Setting the A-tag:
746 $bTag = '<a href="'.htmlspecialchars($href).'"'.
747 ($tagCode[2]&&$tagCode[2]!='-' ? ' target="'.htmlspecialchars($tagCode[2]).'"' : '').
748 ($tagCode[3]&&$tagCode[3]!='-' ? ' class="'.htmlspecialchars($tagCode[3]).'"' : '').
749 ($tagCode[4] ? ' title="'.htmlspecialchars($tagCode[4]).'"' : '').
750 ($error ? ' rteerror="'.htmlspecialchars($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...
751 '>';
752 $eTag = '</a>';
753 $blockSplit[$k] = $bTag.$this->TS_links_rte($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
754 }
755 }
756
757 // Return content:
758 return implode('',$blockSplit);
759 }
760
761 /**
762 * Preserve special tags
763 *
764 * @param string Content input
765 * @return string Content output
766 */
767 function TS_preserve_db($value) {
768 if (!$this->preserveTags) return $value;
769
770 // Splitting into blocks for processing (span-tags are used for special tags)
771 $blockSplit = $this->splitIntoBlock('span',$value);
772 foreach($blockSplit as $k => $v) {
773 if ($k%2) { // block:
774 $attribArray=$this->get_tag_attributes_classic($this->getFirstTag($v));
775 if ($attribArray['specialtag']) {
776 $theTag = rawurldecode($attribArray['specialtag']);
777 $theTagName = $this->getFirstTagName($theTag);
778 $blockSplit[$k] = $theTag.$this->removeFirstAndLastTag($blockSplit[$k]).'</'.$theTagName.'>';
779 }
780 }
781 }
782 return implode('',$blockSplit);
783 }
784
785 /**
786 * Preserve special tags
787 *
788 * @param string Content input
789 * @return string Content output
790 */
791 function TS_preserve_rte($value) {
792 if (!$this->preserveTags) return $value;
793
794 $blockSplit = $this->splitIntoBlock($this->preserveTags,$value);
795 foreach($blockSplit as $k => $v) {
796 if ($k%2) { // block:
797 $blockSplit[$k] = '<span specialtag="'.rawurlencode($this->getFirstTag($v)).'">'.$this->removeFirstAndLastTag($blockSplit[$k]).'</span>';
798 }
799 }
800 return implode('',$blockSplit);
801 }
802
803 /**
804 * Transformation handler: 'ts_transform' + 'css_transform' / direction: "db"
805 * Cleaning (->db) for standard content elements (ts)
806 *
807 * @param string Content input
808 * @param boolean If true, the transformation was "css_transform", otherwise "ts_transform"
809 * @return string Content output
810 * @see TS_transform_rte()
811 */
812 function TS_transform_db($value,$css=FALSE) {
813
814 // safety... so forever loops are avoided (they should not occur, but an error would potentially do this...)
815 $this->TS_transform_db_safecounter--;
816 if ($this->TS_transform_db_safecounter<0) return $value;
817
818 // Split the content from RTE by the occurence of these blocks:
819 $blockSplit = $this->splitIntoBlock('TABLE,BLOCKQUOTE,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$value);
820
821 $cc=0;
822 $aC = count($blockSplit);
823
824 // Avoid superfluous linebreaks by transform_db after ending headListTag
825 while($aC && !strcmp(trim($blockSplit[$aC-1]),'')) {
826 unset($blockSplit[$aC-1]);
827 $aC = count($blockSplit);
828 }
829
830 // Traverse the blocks
831 foreach($blockSplit as $k => $v) {
832 $cc++;
833 $lastBR = $cc==$aC ? '' : chr(10);
834
835 if ($k%2) { // Inside block:
836
837 // Init:
838 $tag=$this->getFirstTag($v);
839 $tagName=strtolower($this->getFirstTagName($v));
840
841 // Process based on the tag:
842 switch($tagName) {
843 case 'blockquote': // Keep blockquotes, but clean the inside recursively in the same manner as the main code
844 case 'dd' : // Do the same on dd elements
845 case 'div': // Do the same on div sections, if they were splitted
846 $blockSplit[$k]=$tag.$this->TS_transform_db($this->removeFirstAndLastTag($blockSplit[$k]),$css).'</'.$tagName.'>'.$lastBR;
847 break;
848 case 'ol':
849 case 'ul': // Transform lists into <typolist>-tags:
850 if (!$css) {
851 if (!isset($this->procOptions['typolist']) || $this->procOptions['typolist']) {
852 $parts = $this->getAllParts($this->splitIntoBlock('LI',$this->removeFirstAndLastTag($blockSplit[$k])),1,0);
853 while(list($k2)=each($parts)) {
854 $parts[$k2]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/','',$parts[$k2]); // remove all linesbreaks!
855 $parts[$k2]=$this->defaultTStagMapping($parts[$k2],'db');
856 $parts[$k2]=$this->cleanFontTags($parts[$k2],0,0,0);
857 $parts[$k2] = $this->HTMLcleaner_db($parts[$k2],strtolower($this->procOptions['allowTagsInTypolists']?$this->procOptions['allowTagsInTypolists']:'br,font,b,i,u,a,img,span,strong,em'));
858 }
859 if ($tagName=='ol') { $params=' type="1"'; } else { $params=''; }
860 $blockSplit[$k]='<typolist'.$params.'>'.chr(10).implode(chr(10),$parts).chr(10).'</typolist>'.$lastBR;
861 }
862 } else {
863 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
864 }
865 break;
866 case 'table': // Tables are NOT allowed in any form (unless preserveTables is set or CSS is the mode)
867 if (!$this->procOptions['preserveTables'] && !$css) {
868 $blockSplit[$k]=$this->TS_transform_db($this->removeTables($blockSplit[$k]));
869 } else {
870 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
871 }
872 break;
873 case 'h1':
874 case 'h2':
875 case 'h3':
876 case 'h4':
877 case 'h5':
878 case 'h6':
879 if (!$css) {
880 $attribArray=$this->get_tag_attributes_classic($tag);
881 // Processing inner content here:
882 $innerContent = $this->HTMLcleaner_db($this->removeFirstAndLastTag($blockSplit[$k]));
883
884 if (!isset($this->procOptions['typohead']) || $this->procOptions['typohead']) {
885 $type = intval(substr($tagName,1));
886 $blockSplit[$k]='<typohead'.
887 ($type!=6?' type="'.$type.'"':'').
888 ($attribArray['align']?' align="'.$attribArray['align'].'"':'').
889 ($attribArray['class']?' class="'.$attribArray['class'].'"':'').
890 '>'.
891 $innerContent.
892 '</typohead>'.
893 $lastBR;
894 } else {
895 $blockSplit[$k]='<'.$tagName.
896 ($attribArray['align']?' align="'.htmlspecialchars($attribArray['align']).'"':'').
897 ($attribArray['class']?' class="'.htmlspecialchars($attribArray['class']).'"':'').
898 '>'.
899 $innerContent.
900 '</'.$tagName.'>'.
901 $lastBR;
902 }
903 } else {
904 // Eliminate true linebreaks inside Hx tags
905 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
906 }
907 break;
908 default:
909 // Eliminate true linebreaks inside other headlist tags and after hr tag
910 $blockSplit[$k]=preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$this->transformStyledATags($blockSplit[$k])).$lastBR;
911 break;
912 }
913 } else { // NON-block:
914 if (strcmp(trim($blockSplit[$k]),'')) {
915 $blockSplit[$k]=$this->divideIntoLines(preg_replace('/['.preg_quote(chr(10).chr(13)).']+/',' ',$blockSplit[$k])).$lastBR;
916 $blockSplit[$k]=$this->transformStyledATags($blockSplit[$k]);
917 } else unset($blockSplit[$k]);
918 }
919 }
920 $this->TS_transform_db_safecounter++;
921
922 return implode('',$blockSplit);
923 }
924
925 /**
926 * Wraps a-tags that contain a style attribute with a span-tag
927 *
928 * @param string Content input
929 * @return string Content output
930 */
931 function transformStyledATags($value) {
932 $blockSplit = $this->splitIntoBlock('A',$value);
933 foreach($blockSplit as $k => $v) {
934 if ($k%2) { // If an A-tag was found:
935 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
936 if ($attribArray['style']) { // If "style" attribute is set!
937 $attribArray_copy['style'] = $attribArray['style'];
938 unset($attribArray['style']);
939 $bTag='<span '.t3lib_div::implodeAttributes($attribArray_copy,1).'><a '.t3lib_div::implodeAttributes($attribArray,1).'>';
940 $eTag='</a></span>';
941 $blockSplit[$k] = $bTag.$this->removeFirstAndLastTag($blockSplit[$k]).$eTag;
942 }
943 }
944 }
945 return implode('',$blockSplit);
946 }
947
948 /**
949 * Transformation handler: 'ts_transform' + 'css_transform' / direction: "rte"
950 * Set (->rte) for standard content elements (ts)
951 *
952 * @param string Content input
953 * @param boolean If true, the transformation was "css_transform", otherwise "ts_transform"
954 * @return string Content output
955 * @see TS_transform_db()
956 */
957 function TS_transform_rte($value,$css=0) {
958
959 // Split the content from Database by the occurence of these blocks:
960 $blockSplit = $this->splitIntoBlock('TABLE,BLOCKQUOTE,TYPOLIST,TYPOHEAD,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$value);
961
962 // Traverse the blocks
963 foreach($blockSplit as $k => $v) {
964 if ($k%2) { // Inside one of the blocks:
965
966 // Init:
967 $tag = $this->getFirstTag($v);
968 $tagName = strtolower($this->getFirstTagName($v));
969 $attribArray = $this->get_tag_attributes_classic($tag);
970
971 // Based on tagname, we do transformations:
972 switch($tagName) {
973 case 'blockquote': // Keep blockquotes
974 case 'dd': // Keep definitions
975 case 'div': // Keep div sections, if they were splitted
976 $blockSplit[$k] = $tag.
977 $this->TS_transform_rte($this->removeFirstAndLastTag($blockSplit[$k]),$css).
978 '</'.$tagName.'>';
979 break;
980 case 'typolist': // Transform typolist blocks into OL/UL lists. Type 1 is expected to be numerical block
981 if (!isset($this->procOptions['typolist']) || $this->procOptions['typolist']) {
982 $tListContent = $this->removeFirstAndLastTag($blockSplit[$k]);
983 $tListContent = ereg_replace('^[ ]*'.chr(10),'',$tListContent);
984 $tListContent = ereg_replace(chr(10).'[ ]*$','',$tListContent);
985 $lines = explode(chr(10),$tListContent);
986 $typ = $attribArray['type']==1 ? 'ol' : 'ul';
987 $blockSplit[$k] = '<'.$typ.'>'.chr(10).
988 '<li>'.implode('</li>'.chr(10).'<li>',$lines).'</li>'.
989 '</'.$typ.'>';
990 }
991 break;
992 case 'typohead': // Transform typohead into Hx tags.
993 if (!isset($this->procOptions['typohead']) || $this->procOptions['typohead']) {
994 $tC = $this->removeFirstAndLastTag($blockSplit[$k]);
995 $typ = t3lib_div::intInRange($attribArray['type'],0,6);
996 if (!$typ) $typ=6;
997 $align = $attribArray['align']?' align="'.$attribArray['align'].'"': '';
998 $class = $attribArray['class']?' class="'.$attribArray['class'].'"': '';
999 $blockSplit[$k] = '<h'.$typ.$align.$class.'>'.
1000 $tC.
1001 '</h'.$typ.'>';
1002 }
1003 break;
1004 }
1005 $blockSplit[$k+1] = ereg_replace('^[ ]*'.chr(10),'',$blockSplit[$k+1]); // Removing linebreak if typohead
1006 } else { // NON-block:
1007 $nextFTN = $this->getFirstTagName($blockSplit[$k+1]);
1008 $singleLineBreak = $blockSplit[$k]==chr(10);
1009 if (t3lib_div::inList('TABLE,BLOCKQUOTE,TYPOLIST,TYPOHEAD,'.($this->procOptions['preserveDIVSections']?'DIV,':'').$this->blockElementList,$nextFTN)) { // Removing linebreak if typolist/typohead
1010 $blockSplit[$k] = ereg_replace(chr(10).'[ ]*$','',$blockSplit[$k]);
1011 }
1012 // If $blockSplit[$k] is blank then unset the line. UNLESS the line happend to be a single line break.
1013 if (!strcmp($blockSplit[$k],'') && !$singleLineBreak) {
1014 unset($blockSplit[$k]);
1015 } else {
1016 $blockSplit[$k] = $this->setDivTags($blockSplit[$k],($this->procOptions['useDIVasParagraphTagForRTE']?'div':'p'));
1017 }
1018 }
1019 }
1020 return implode(chr(10),$blockSplit);
1021 }
1022
1023 /**
1024 * Transformation handler: 'ts_strip' / direction: "db"
1025 * Removing all non-allowed tags
1026 *
1027 * @param string Content input
1028 * @return string Content output
1029 */
1030 function TS_strip_db($value) {
1031 $value = strip_tags($value,'<'.implode('><',explode(',','b,i,u,a,img,br,div,center,pre,font,hr,sub,sup,p,strong,em,li,ul,ol,blockquote')).'>');
1032 return $value;
1033 }
1034
1035
1036
1037
1038
1039
1040
1041
1042
1043
1044
1045
1046
1047
1048 /***************************************************************
1049 *
1050 * Generic RTE transformation, analysis and helper functions
1051 *
1052 **************************************************************/
1053
1054 /**
1055 * Reads the file or url $url and returns the content
1056 *
1057 * @param string Filepath/URL to read
1058 * @return string The content from the resource given as input.
1059 * @see t3lib_div::getURL()
1060 */
1061 function getURL($url) {
1062 return t3lib_div::getURL($url);
1063 }
1064
1065 /**
1066 * Function for cleaning content going into the database.
1067 * Content is cleaned eg. by removing unallowed HTML and ds-HSC content
1068 * It is basically calling HTMLcleaner from the parent class with some preset configuration specifically set up for cleaning content going from the RTE into the db
1069 *
1070 * @param string Content to clean up
1071 * @param string Comma list of tags to specifically allow. Default comes from getKeepTags and is ""
1072 * @return string Clean content
1073 * @see getKeepTags()
1074 */
1075 function HTMLcleaner_db($content,$tagList='') {
1076 if (!$tagList) {
1077 $keepTags = $this->getKeepTags('db');
1078 } else {
1079 $keepTags = $this->getKeepTags('db',$tagList);
1080 }
1081 $kUknown = $this->procOptions['dontRemoveUnknownTags_db'] ? 1 : 0; // Default: remove unknown tags.
1082 $hSC = $this->procOptions['dontUndoHSC_db'] ? 0 : -1; // Default: re-convert literals to characters (that is &lt; to <)
1083
1084 // Create additional configuration in order to honor the setting RTE.default.proc.HTMLparser_db.xhtml_cleaning=1
1085 $addConfig=array();
1086 if ((is_array($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'])) {
1087 $addConfig['xhtml']=1;
1088 }
1089
1090 return $this->HTMLcleaner($content,$keepTags,$kUknown,$hSC,$addConfig);
1091 }
1092
1093 /**
1094 * Creates an array of configuration for the HTMLcleaner function based on whether content go TO or FROM the Rich Text Editor ($direction)
1095 * Unless "tagList" is given, the function will cache the configuration for next time processing goes on. (In this class that is the case only if we are processing a bulletlist)
1096 *
1097 * @param string The direction of the content being processed by the output configuration; "db" (content going into the database FROM the rte) or "rte" (content going into the form)
1098 * @param string Comma list of tags to keep (overriding default which is to keep all + take notice of internal configuration)
1099 * @return array Configuration array
1100 * @see HTMLcleaner_db()
1101 */
1102 function getKeepTags($direction='rte',$tagList='') {
1103 if (!is_array($this->getKeepTags_cache[$direction]) || $tagList) {
1104
1105 // Setting up allowed tags:
1106 if (strcmp($tagList,'')) { // If the $tagList input var is set, this will take precedence
1107 $keepTags = array_flip(t3lib_div::trimExplode(',',$tagList,1));
1108 } else { // Default is to get allowed/denied tags from internal array of processing options:
1109 // Construct default list of tags to keep:
1110 $typoScript_list = 'b,i,u,a,img,br,div,center,pre,font,hr,sub,sup,p,strong,em,li,ul,ol,blockquote,strike,span';
1111 $keepTags = array_flip(t3lib_div::trimExplode(',',$typoScript_list.','.strtolower($this->procOptions['allowTags']),1));
1112
1113 // For tags to deny, remove them from $keepTags array:
1114 $denyTags = t3lib_div::trimExplode(',',$this->procOptions['denyTags'],1);
1115 foreach($denyTags as $dKe) {
1116 unset($keepTags[$dKe]);
1117 }
1118 }
1119
1120 // Based on the direction of content, set further options:
1121 switch ($direction) {
1122
1123 // GOING from database to Rich Text Editor:
1124 case 'rte':
1125
1126 if (!isset($this->procOptions['transformBoldAndItalicTags']) || $this->procOptions['transformBoldAndItalicTags']) {
1127 // Transform bold/italics tags to strong/em
1128 if (isset($keepTags['b'])) {$keepTags['b']=array('remap'=>'STRONG');}
1129 if (isset($keepTags['i'])) {$keepTags['i']=array('remap'=>'EM');}
1130 }
1131
1132 // Transforming keepTags array so it can be understood by the HTMLcleaner function. This basically converts the format of the array from TypoScript (having .'s) to plain multi-dimensional array.
1133 list($keepTags) = $this->HTMLparserConfig($this->procOptions['HTMLparser_rte.'],$keepTags);
1134 break;
1135
1136 // GOING from RTE to database:
1137 case 'db':
1138
1139 if (!isset($this->procOptions['transformBoldAndItalicTags']) || $this->procOptions['transformBoldAndItalicTags']) {
1140 // Transform strong/em back to bold/italics:
1141 if (isset($keepTags['strong'])) { $keepTags['strong']=array('remap'=>'b'); }
1142 if (isset($keepTags['em'])) { $keepTags['em']=array('remap'=>'i'); }
1143 }
1144
1145 // Setting up span tags if they are allowed:
1146 if (isset($keepTags['span'])) {
1147 $classes=array_merge(array(''),$this->allowedClasses);
1148 $keepTags['span']=array(
1149 'allowedAttribs' => 'class,style,xml:lang',
1150 'fixAttrib' => Array(
1151 'class' => Array (
1152 'list' => $classes,
1153 'removeIfFalse' => 1
1154 )
1155 ),
1156 'rmTagIfNoAttrib' => 1
1157 );
1158 if (!$this->procOptions['allowedClasses']) unset($keepTags['span']['fixAttrib']['class']['list']);
1159 }
1160
1161 // Setting up font tags if they are allowed:
1162 if (isset($keepTags['font'])) {
1163 $colors=array_merge(array(''),t3lib_div::trimExplode(',',$this->procOptions['allowedFontColors'],1));
1164 $keepTags['font']=array(
1165 'allowedAttribs'=>'face,color,size',
1166 'fixAttrib' => Array(
1167 'face' => Array (
1168 'removeIfFalse' => 1
1169 ),
1170 'color' => Array (
1171 'removeIfFalse' => 1,
1172 'list'=>$colors
1173 ),
1174 'size' => Array (
1175 'removeIfFalse' => 1,
1176 )
1177 ),
1178 'rmTagIfNoAttrib' => 1
1179 );
1180 if (!$this->procOptions['allowedFontColors']) unset($keepTags['font']['fixAttrib']['color']['list']);
1181 }
1182
1183 // Setting further options, getting them from the processiong options:
1184 $TSc = $this->procOptions['HTMLparser_db.'];
1185 if (!$TSc['globalNesting']) $TSc['globalNesting']='b,i,u,a,center,font,sub,sup,strong,em,strike,span';
1186 if (!$TSc['noAttrib']) $TSc['noAttrib']='b,i,u,br,center,hr,sub,sup,strong,em,li,ul,ol,blockquote,strike';
1187
1188 // Transforming the array from TypoScript to regular array:
1189 list($keepTags) = $this->HTMLparserConfig($TSc,$keepTags);
1190 break;
1191 }
1192
1193 // Caching (internally, in object memory) the result unless tagList is set:
1194 if (!$tagList) {
1195 $this->getKeepTags_cache[$direction] = $keepTags;
1196 } else {
1197 return $keepTags;
1198 }
1199 }
1200
1201 // Return result:
1202 return $this->getKeepTags_cache[$direction];
1203 }
1204
1205 /**
1206 * This resolves the $value into parts based on <div></div>-sections and <p>-sections and <br />-tags. These are returned as lines separated by chr(10).
1207 * This point is to resolve the HTML-code returned from RTE into ordinary lines so it's 'human-readable'
1208 * The function ->setDivTags does the opposite.
1209 * This function processes content to go into the database.
1210 *
1211 * @param string Value to process.
1212 * @param integer Recursion brake. Decremented on each recursion down to zero. Default is 5 (which equals the allowed nesting levels of p/div tags).
1213 * @param boolean If true, an array with the lines is returned, otherwise a string of the processed input value.
1214 * @return string Processed input value.
1215 * @see setDivTags()
1216 */
1217 function divideIntoLines($value,$count=5,$returnArray=FALSE) {
1218
1219 // Internalize font tags (move them from OUTSIDE p/div to inside it that is the case):
1220 if ($this->procOptions['internalizeFontTags']) {$value = $this->internalizeFontTags($value);}
1221
1222 // Setting configuration for processing:
1223 $allowTagsOutside = t3lib_div::trimExplode(',',strtolower($this->procOptions['allowTagsOutside']?$this->procOptions['allowTagsOutside']:'img'),1);
1224 $remapParagraphTag = strtoupper($this->procOptions['remapParagraphTag']);
1225 $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...?
1226
1227 if ($this->procOptions['keepPDIVattribs']) {
1228 $keepAttribListArr = t3lib_div::trimExplode(',',strtolower($this->procOptions['keepPDIVattribs']),1);
1229 } else {
1230 $keepAttribListArr = array();
1231 }
1232
1233 // Returns plainly the value if there was no div/p sections in it
1234 if (count($divSplit)<=1 || $count<=0) {
1235 return $value;
1236 }
1237
1238 // Traverse the splitted sections:
1239 foreach($divSplit as $k => $v) {
1240 if ($k%2) { // Inside
1241 $v=$this->removeFirstAndLastTag($v);
1242
1243 // Fetching 'sub-lines' - which will explode any further p/div nesting...
1244 $subLines = $this->divideIntoLines($v,$count-1,1);
1245 if (is_array($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')
1246 // No noting.
1247 } else { //... but if NO subsection was found, we process it as a TRUE line without erronous content:
1248 $subLines = array($subLines);
1249 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...
1250 $subLines = spliti('<br[[:space:]]*[\/]?>',$v);
1251 }
1252
1253 // Traverse sublines (there is typically one, except if <br/> has been converted to lines as well!)
1254 reset($subLines);
1255 while(list($sk)=each($subLines)) {
1256
1257 // Clear up the subline for DB.
1258 $subLines[$sk]=$this->HTMLcleaner_db($subLines[$sk]);
1259
1260 // Get first tag, attributes etc:
1261 $fTag = $this->getFirstTag($divSplit[$k]);
1262 $tagName=strtolower($this->getFirstTagName($divSplit[$k]));
1263 $attribs=$this->get_tag_attributes($fTag);
1264
1265 // Keep attributes (lowercase)
1266 $newAttribs=array();
1267 if (count($keepAttribListArr)) {
1268 foreach($keepAttribListArr as $keepA) {
1269 if (isset($attribs[0][$keepA])) { $newAttribs[$keepA] = $attribs[0][$keepA]; }
1270 }
1271 }
1272
1273 // ALIGN attribute:
1274 if (!$this->procOptions['skipAlign'] && strcmp(trim($attribs[0]['align']),'') && strtolower($attribs[0]['align'])!='left') { // Set to value, but not 'left'
1275 $newAttribs['align']=strtolower($attribs[0]['align']);
1276 }
1277
1278 // CLASS attribute:
1279 if (!$this->procOptions['skipClass'] && strcmp(trim($attribs[0]['class']),'')) { // Set to whatever value
1280 if (!count($this->allowedClasses) || in_array(strtoupper($attribs[0]['class']),$this->allowedClasses)) {
1281 $newAttribs['class']=$attribs[0]['class'];
1282 }
1283 }
1284
1285 // Remove any line break char (10 or 13)
1286 $subLines[$sk]=ereg_replace(chr(10).'|'.chr(13),'',$subLines[$sk]);
1287
1288 // If there are any attributes or if we are supposed to remap the tag, then do so:
1289 if (count($newAttribs) && strcmp($remapParagraphTag,'1')) {
1290 if ($remapParagraphTag=='P') $tagName='p';
1291 if ($remapParagraphTag=='DIV') $tagName='div';
1292 $subLines[$sk]='<'.trim($tagName.' '.$this->compileTagAttribs($newAttribs)).'>'.$subLines[$sk].'</'.$tagName.'>';
1293 }
1294 }
1295 }
1296 // Add the processed line(s)
1297 $divSplit[$k] = implode(chr(10),$subLines);
1298
1299 // If it turns out the line is just blank (containing a &nbsp; possibly) then just make it pure blank.
1300 // But, prevent filtering of lines that are blank in sense above, but whose tags contain attributes.
1301 // Those attributes should have been filtered before; if they are still there they must be considered as possible content.
1302 if (trim(strip_tags($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]))) {
1303 $divSplit[$k]='';
1304 }
1305 } else { // outside div:
1306 // Remove positions which are outside div/p tags and without content
1307 $divSplit[$k]=trim(strip_tags($divSplit[$k],'<'.implode('><',$allowTagsOutside).'>'));
1308 if (!strcmp($divSplit[$k],'')) unset($divSplit[$k]); // Remove part if it's empty
1309 }
1310 }
1311
1312 // Return value:
1313 return $returnArray ? $divSplit : implode(chr(10),$divSplit);
1314 }
1315
1316 /**
1317 * Converts all lines into <div></div>/<p></p>-sections (unless the line is a div-section already)
1318 * For processing of content going FROM database TO RTE.
1319 *
1320 * @param string Value to convert
1321 * @param string Tag to wrap with. Either "p" or "div" should it be. Lowercase preferably.
1322 * @return string Processed value.
1323 * @see divideIntoLines()
1324 */
1325 function setDivTags($value,$dT='p') {
1326
1327 // First, setting configuration for the HTMLcleaner function. This will process each line between the <div>/<p> section on their way to the RTE
1328 $keepTags = $this->getKeepTags('rte');
1329 $kUknown = $this->procOptions['dontProtectUnknownTags_rte'] ? 0 : 'protect'; // Default: remove unknown tags.
1330 $hSC = $this->procOptions['dontHSC_rte'] ? 0 : 1; // Default: re-convert literals to characters (that is &lt; to <)
1331 $convNBSP = !$this->procOptions['dontConvAmpInNBSP_rte']?1:0;
1332
1333 // Divide the content into lines, based on chr(10):
1334 $parts = explode(chr(10),$value);
1335 foreach($parts as $k => $v) {
1336
1337 // Processing of line content:
1338 if (!strcmp(trim($parts[$k]),'')) { // If the line is blank, set it to &nbsp;
1339 $parts[$k]='&nbsp;';
1340 } else { // Clean the line content:
1341 $parts[$k]=$this->HTMLcleaner($parts[$k],$keepTags,$kUknown,$hSC);
1342 if ($convNBSP) $parts[$k]=str_replace('&amp;nbsp;','&nbsp;',$parts[$k]);
1343 }
1344
1345 // Wrapping the line in <$dT> is not already wrapped:
1346 $testStr = strtolower(trim($parts[$k]));
1347 if (substr($testStr,0,4)!='<div' || substr($testStr,-6)!='</div>') {
1348 if (substr($testStr,0,2)!='<p' || substr($testStr,-4)!='</p>') {
1349 // Only set p-tags if there is not already div or p tags:
1350 $parts[$k]='<'.$dT.'>'.$parts[$k].'</'.$dT.'>';
1351 }
1352 }
1353 }
1354
1355 // Implode result:
1356 return implode(chr(10),$parts);
1357 }
1358
1359 /**
1360 * This splits the $value in font-tag chunks.
1361 * If there are any <P>/<DIV> sections inside of them, the font-tag is wrapped AROUND the content INSIDE of the P/DIV sections and the outer font-tag is removed.
1362 * This functions seems to be a good choice for pre-processing content if it has been pasted into the RTE from eg. star-office.
1363 * In that case the font-tags are normally on the OUTSIDE of the sections.
1364 * This function is used by eg. divideIntoLines() if the procesing option 'internalizeFontTags' is set.
1365 *
1366 * @param string Input content
1367 * @return string Output content
1368 * @see divideIntoLines()
1369 */
1370 function internalizeFontTags($value) {
1371
1372 // Splitting into font tag blocks:
1373 $fontSplit = $this->splitIntoBlock('font',$value);
1374
1375 foreach($fontSplit as $k => $v) {
1376 if ($k%2) { // Inside
1377 $fTag = $this->getFirstTag($v); // Fint font-tag
1378
1379 $divSplit_sub = $this->splitIntoBlock('div,p',$this->removeFirstAndLastTag($v),1);
1380 if (count($divSplit_sub)>1) { // If there were div/p sections inside the font-tag, do something about it...
1381 // traverse those sections:
1382 foreach($divSplit_sub as $k2 => $v2) {
1383 if ($k2%2) { // Inside
1384 $div_p = $this->getFirstTag($v2); // Fint font-tag
1385 $div_p_tagname = $this->getFirstTagName($v2); // Fint font-tag
1386 $v2=$this->removeFirstAndLastTag($v2); // ... and remove it from original.
1387 $divSplit_sub[$k2]=$div_p.$fTag.$v2.'</font>'.'</'.$div_p_tagname.'>';
1388 } elseif (trim(strip_tags($v2))) {
1389 $divSplit_sub[$k2]=$fTag.$v2.'</font>';
1390 }
1391 }
1392 $fontSplit[$k]=implode('',$divSplit_sub);
1393 }
1394 }
1395 }
1396
1397 return implode('',$fontSplit);
1398 }
1399
1400 /**
1401 * Returns SiteURL based on thisScript.
1402 *
1403 * @return string Value of t3lib_div::getIndpEnv('TYPO3_SITE_URL');
1404 * @see t3lib_div::getIndpEnv()
1405 */
1406 function siteUrl() {
1407 return t3lib_div::getIndpEnv('TYPO3_SITE_URL');
1408 }
1409
1410 /**
1411 * Return the storage folder of RTE image files.
1412 * Default is $GLOBALS['TYPO3_CONF_VARS']['BE']['RTE_imageStorageDir'] unless something else is configured in the types configuration for the RTE.
1413 *
1414 * @return string
1415 */
1416 function rteImageStorageDir() {
1417 return $this->rte_p['imgpath'] ? $this->rte_p['imgpath'] : $GLOBALS['TYPO3_CONF_VARS']['BE']['RTE_imageStorageDir'];
1418 }
1419
1420 /**
1421 * Remove all tables from incoming code
1422 * The function is trying to to this is some more or less respectfull way. The approach is to resolve each table cells content and implode it all by <br /> chars. Thus at least the content is preserved in some way.
1423 *
1424 * @param string Input value
1425 * @param string Break character to use for linebreaks.
1426 * @return string Output value
1427 */
1428 function removeTables($value,$breakChar='<br />') {
1429
1430 // Splitting value into table blocks:
1431 $tableSplit = $this->splitIntoBlock('table',$value);
1432
1433 // Traverse blocks of tables:
1434 foreach($tableSplit as $k => $v) {
1435 if ($k%2) {
1436 $tableSplit[$k]='';
1437 $rowSplit = $this->splitIntoBlock('tr',$v);
1438 foreach($rowSplit as $k2 => $v2) {
1439 if ($k2%2) {
1440 $cellSplit = $this->getAllParts($this->splitIntoBlock('td',$v2),1,0);
1441 foreach($cellSplit as $k3 => $v3) {
1442 $tableSplit[$k].=$v3.$breakChar;
1443 }
1444 }
1445 }
1446 }
1447 }
1448
1449 // Implode it all again:
1450 return implode($breakChar,$tableSplit);
1451 }
1452
1453 /**
1454 * Default tag mapping for TS
1455 *
1456 * @param string Input code to process
1457 * @param string Direction To databsae (db) or from database to RTE (rte)
1458 * @return string Processed value
1459 */
1460 function defaultTStagMapping($code,$direction='rte') {
1461 if ($direction=='db') {
1462 $code=$this->mapTags($code,array( // Map tags
1463 'strong' => 'b',
1464 'em' => 'i'
1465 ));
1466 }
1467 if ($direction=='rte') {
1468 $code=$this->mapTags($code,array( // Map tags
1469 'b' => 'strong',
1470 'i' => 'em'
1471 ));
1472 }
1473 return $code;
1474 }
1475
1476 /**
1477 * Finds width and height from attrib-array
1478 * If the width and height is found in the style-attribute, use that!
1479 *
1480 * @param array Array of attributes from tag in which to search. More specifically the content of the key "style" is used to extract "width:xxx / height:xxx" information
1481 * @return array Integer w/h in key 0/1. Zero is returned if not found.
1482 */
1483 function getWHFromAttribs($attribArray) {
1484 $style =trim($attribArray['style']);
1485 if ($style) {
1486 $regex='[[:space:]]*:[[:space:]]*([0-9]*)[[:space:]]*px';
1487 // Width
1488 $reg = array();
1489 eregi('width'.$regex,$style,$reg);
1490 $w = intval($reg[1]);
1491 // Height
1492 eregi('height'.$regex,$style,$reg);
1493 $h = intval($reg[1]);
1494 }
1495 if (!$w) {
1496 $w = $attribArray['width'];
1497 }
1498 if (!$h) {
1499 $h = $attribArray['height'];
1500 }
1501 return array(intval($w),intval($h));
1502 }
1503
1504 /**
1505 * Parse <A>-tag href and return status of email,external,file or page
1506 *
1507 * @param string URL to analyse.
1508 * @return array Information in an array about the URL
1509 */
1510 function urlInfoForLinkTags($url) {
1511 $info = array();
1512 $url = trim($url);
1513 if (substr(strtolower($url),0,7)=='mailto:') {
1514 $info['url']=trim(substr($url,7));
1515 $info['type']='email';
1516 } else {
1517 $curURL = $this->siteUrl(); // 100502, removed this: 'http://'.t3lib_div::getThisUrl(); Reason: The url returned had typo3/ in the end - should be only the site's url as far as I see...
1518 for($a=0;$a<strlen($url);$a++) {
1519 if ($url{$a}!=$curURL{$a}) {
1520 break;
1521 }
1522 }
1523
1524 $info['relScriptPath']=substr($curURL,$a);
1525 $info['relUrl']=substr($url,$a);
1526 $info['url']=$url;
1527 $info['type']='ext';
1528
1529 $siteUrl_parts = parse_url($url);
1530 $curUrl_parts = parse_url($curURL);
1531
1532 if ($siteUrl_parts['host']==$curUrl_parts['host'] // Hosts should match
1533 && (!$info['relScriptPath'] || (defined('TYPO3_mainDir') && substr($info['relScriptPath'],0,strlen(TYPO3_mainDir))==TYPO3_mainDir))) { // If the script path seems to match or is empty (FE-EDIT)
1534
1535 // New processing order 100502
1536 $uP=parse_url($info['relUrl']);
1537
1538 if (!strcmp('#'.$siteUrl_parts['fragment'],$info['relUrl'])) {
1539 $info['url']=$info['relUrl'];
1540 $info['type']='anchor';
1541 } elseif (!trim($uP['path']) || !strcmp($uP['path'],'index.php')) {
1542 $pp = explode('id=',$uP['query']);
1543 $id = trim($pp[1]);
1544 if ($id) {
1545 $info['pageid']=$id;
1546 $info['cElement']=$uP['fragment'];
1547 $info['url']=$id.($info['cElement']?'#'.$info['cElement']:'');
1548 $info['type']='page';
1549 }
1550 } else {
1551 $info['url']=$info['relUrl'];
1552 $info['type']='file';
1553 }
1554 } else {
1555 unset($info['relScriptPath']);
1556 unset($info['relUrl']);
1557 }
1558 }
1559 return $info;
1560 }
1561
1562 /**
1563 * Converting <A>-tags to absolute URLs (+ setting rtekeep attribute)
1564 *
1565 * @param string Content input
1566 * @param boolean If true, then the "rtekeep" attribute will not be set.
1567 * @return string Content output
1568 */
1569 function TS_AtagToAbs($value,$dontSetRTEKEEP=FALSE) {
1570 $blockSplit = $this->splitIntoBlock('A',$value);
1571 reset($blockSplit);
1572 while(list($k,$v)=each($blockSplit)) {
1573 if ($k%2) { // block:
1574 $attribArray = $this->get_tag_attributes_classic($this->getFirstTag($v),1);
1575
1576 // Checking if there is a scheme, and if not, prepend the current url.
1577 if (strlen($attribArray['href'])) { // ONLY do this if href has content - the <a> tag COULD be an anchor and if so, it should be preserved...
1578 $uP = parse_url(strtolower($attribArray['href']));
1579 if (!$uP['scheme']) {
1580 $attribArray['href'] = $this->siteUrl().substr($attribArray['href'],strlen($this->relBackPath));
1581 }
1582 } else {
1583 $attribArray['rtekeep'] = 1;
1584 }
1585 if (!$dontSetRTEKEEP) $attribArray['rtekeep'] = 1;
1586
1587 $bTag='<a '.t3lib_div::implodeAttributes($attribArray,1).'>';
1588 $eTag='</a>';
1589 $blockSplit[$k] = $bTag.$this->TS_AtagToAbs($this->removeFirstAndLastTag($blockSplit[$k])).$eTag;
1590 }
1591 }
1592 return implode('',$blockSplit);
1593 }
1594 }
1595
1596
1597 if (defined('TYPO3_MODE') && $TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['t3lib/class.t3lib_parsehtml_proc.php']) {
1598 include_once($TYPO3_CONF_VARS[TYPO3_MODE]['XCLASS']['t3lib/class.t3lib_parsehtml_proc.php']);
1599 }
1600 ?>