1: <?php
2:
3: /**
4: * Experimental HTML5-based parser using Jeroen van der Meer's PH5P library.
5: * Occupies space in the HTML5 pseudo-namespace, which may cause conflicts.
6: *
7: * @note
8: * Recent changes to PHP's DOM extension have resulted in some fatal
9: * error conditions with the original version of PH5P. Pending changes,
10: * this lexer will punt to DirectLex if DOM throughs an exception.
11: */
12:
13: class HTMLPurifier_Lexer_PH5P extends HTMLPurifier_Lexer_DOMLex {
14:
15: public function tokenizeHTML($html, $config, $context) {
16: $new_html = $this->normalize($html, $config, $context);
17: $new_html = $this->wrapHTML($new_html, $config, $context);
18: try {
19: $parser = new HTML5($new_html);
20: $doc = $parser->save();
21: } catch (DOMException $e) {
22: // Uh oh, it failed. Punt to DirectLex.
23: $lexer = new HTMLPurifier_Lexer_DirectLex();
24: $context->register('PH5PError', $e); // save the error, so we can detect it
25: return $lexer->tokenizeHTML($html, $config, $context); // use original HTML
26: }
27: $tokens = array();
28: $this->tokenizeDOM(
29: $doc->getElementsByTagName('html')->item(0)-> // <html>
30: getElementsByTagName('body')->item(0)-> // <body>
31: getElementsByTagName('div')->item(0) // <div>
32: , $tokens);
33: return $tokens;
34: }
35:
36: }
37:
38: /*
39:
40: Copyright 2007 Jeroen van der Meer <http://jero.net/>
41:
42: Permission is hereby granted, free of charge, to any person obtaining a
43: copy of this software and associated documentation files (the
44: "Software"), to deal in the Software without restriction, including
45: without limitation the rights to use, copy, modify, merge, publish,
46: distribute, sublicense, and/or sell copies of the Software, and to
47: permit persons to whom the Software is furnished to do so, subject to
48: the following conditions:
49:
50: The above copyright notice and this permission notice shall be included
51: in all copies or substantial portions of the Software.
52:
53: THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
54: OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF
55: MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.
56: IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY
57: CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT,
58: TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE
59: SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.
60:
61: */
62:
63: class HTML5 {
64: private $data;
65: private $char;
66: private $EOF;
67: private $state;
68: private $tree;
69: private $token;
70: private $content_model;
71: private $escape = false;
72: private $entities = array('AElig;','AElig','AMP;','AMP','Aacute;','Aacute',
73: 'Acirc;','Acirc','Agrave;','Agrave','Alpha;','Aring;','Aring','Atilde;',
74: 'Atilde','Auml;','Auml','Beta;','COPY;','COPY','Ccedil;','Ccedil','Chi;',
75: 'Dagger;','Delta;','ETH;','ETH','Eacute;','Eacute','Ecirc;','Ecirc','Egrave;',
76: 'Egrave','Epsilon;','Eta;','Euml;','Euml','GT;','GT','Gamma;','Iacute;',
77: 'Iacute','Icirc;','Icirc','Igrave;','Igrave','Iota;','Iuml;','Iuml','Kappa;',
78: 'LT;','LT','Lambda;','Mu;','Ntilde;','Ntilde','Nu;','OElig;','Oacute;',
79: 'Oacute','Ocirc;','Ocirc','Ograve;','Ograve','Omega;','Omicron;','Oslash;',
80: 'Oslash','Otilde;','Otilde','Ouml;','Ouml','Phi;','Pi;','Prime;','Psi;',
81: 'QUOT;','QUOT','REG;','REG','Rho;','Scaron;','Sigma;','THORN;','THORN',
82: 'TRADE;','Tau;','Theta;','Uacute;','Uacute','Ucirc;','Ucirc','Ugrave;',
83: 'Ugrave','Upsilon;','Uuml;','Uuml','Xi;','Yacute;','Yacute','Yuml;','Zeta;',
84: 'aacute;','aacute','acirc;','acirc','acute;','acute','aelig;','aelig',
85: 'agrave;','agrave','alefsym;','alpha;','amp;','amp','and;','ang;','apos;',
86: 'aring;','aring','asymp;','atilde;','atilde','auml;','auml','bdquo;','beta;',
87: 'brvbar;','brvbar','bull;','cap;','ccedil;','ccedil','cedil;','cedil',
88: 'cent;','cent','chi;','circ;','clubs;','cong;','copy;','copy','crarr;',
89: 'cup;','curren;','curren','dArr;','dagger;','darr;','deg;','deg','delta;',
90: 'diams;','divide;','divide','eacute;','eacute','ecirc;','ecirc','egrave;',
91: 'egrave','empty;','emsp;','ensp;','epsilon;','equiv;','eta;','eth;','eth',
92: 'euml;','euml','euro;','exist;','fnof;','forall;','frac12;','frac12',
93: 'frac14;','frac14','frac34;','frac34','frasl;','gamma;','ge;','gt;','gt',
94: 'hArr;','harr;','hearts;','hellip;','iacute;','iacute','icirc;','icirc',
95: 'iexcl;','iexcl','igrave;','igrave','image;','infin;','int;','iota;',
96: 'iquest;','iquest','isin;','iuml;','iuml','kappa;','lArr;','lambda;','lang;',
97: 'laquo;','laquo','larr;','lceil;','ldquo;','le;','lfloor;','lowast;','loz;',
98: 'lrm;','lsaquo;','lsquo;','lt;','lt','macr;','macr','mdash;','micro;','micro',
99: 'middot;','middot','minus;','mu;','nabla;','nbsp;','nbsp','ndash;','ne;',
100: 'ni;','not;','not','notin;','nsub;','ntilde;','ntilde','nu;','oacute;',
101: 'oacute','ocirc;','ocirc','oelig;','ograve;','ograve','oline;','omega;',
102: 'omicron;','oplus;','or;','ordf;','ordf','ordm;','ordm','oslash;','oslash',
103: 'otilde;','otilde','otimes;','ouml;','ouml','para;','para','part;','permil;',
104: 'perp;','phi;','pi;','piv;','plusmn;','plusmn','pound;','pound','prime;',
105: 'prod;','prop;','psi;','quot;','quot','rArr;','radic;','rang;','raquo;',
106: 'raquo','rarr;','rceil;','rdquo;','real;','reg;','reg','rfloor;','rho;',
107: 'rlm;','rsaquo;','rsquo;','sbquo;','scaron;','sdot;','sect;','sect','shy;',
108: 'shy','sigma;','sigmaf;','sim;','spades;','sub;','sube;','sum;','sup1;',
109: 'sup1','sup2;','sup2','sup3;','sup3','sup;','supe;','szlig;','szlig','tau;',
110: 'there4;','theta;','thetasym;','thinsp;','thorn;','thorn','tilde;','times;',
111: 'times','trade;','uArr;','uacute;','uacute','uarr;','ucirc;','ucirc',
112: 'ugrave;','ugrave','uml;','uml','upsih;','upsilon;','uuml;','uuml','weierp;',
113: 'xi;','yacute;','yacute','yen;','yen','yuml;','yuml','zeta;','zwj;','zwnj;');
114:
115: const PCDATA = 0;
116: const RCDATA = 1;
117: const CDATA = 2;
118: const PLAINTEXT = 3;
119:
120: const DOCTYPE = 0;
121: const STARTTAG = 1;
122: const ENDTAG = 2;
123: const COMMENT = 3;
124: const CHARACTR = 4;
125: const EOF = 5;
126:
127: public function __construct($data) {
128:
129: $this->data = $data;
130: $this->char = -1;
131: $this->EOF = strlen($data);
132: $this->tree = new HTML5TreeConstructer;
133: $this->content_model = self::PCDATA;
134:
135: $this->state = 'data';
136:
137: while($this->state !== null) {
138: $this->{$this->state.'State'}();
139: }
140: }
141:
142: public function save() {
143: return $this->tree->save();
144: }
145:
146: private function char() {
147: return ($this->char < $this->EOF)
148: ? $this->data[$this->char]
149: : false;
150: }
151:
152: private function character($s, $l = 0) {
153: if($s + $l < $this->EOF) {
154: if($l === 0) {
155: return $this->data[$s];
156: } else {
157: return substr($this->data, $s, $l);
158: }
159: }
160: }
161:
162: private function characters($char_class, $start) {
163: return preg_replace('#^(['.$char_class.']+).*#s', '\\1', substr($this->data, $start));
164: }
165:
166: private function dataState() {
167: // Consume the next input character
168: $this->char++;
169: $char = $this->char();
170:
171: if($char === '&' && ($this->content_model === self::PCDATA || $this->content_model === self::RCDATA)) {
172: /* U+0026 AMPERSAND (&)
173: When the content model flag is set to one of the PCDATA or RCDATA
174: states: switch to the entity data state. Otherwise: treat it as per
175: the "anything else" entry below. */
176: $this->state = 'entityData';
177:
178: } elseif($char === '-') {
179: /* If the content model flag is set to either the RCDATA state or
180: the CDATA state, and the escape flag is false, and there are at
181: least three characters before this one in the input stream, and the
182: last four characters in the input stream, including this one, are
183: U+003C LESS-THAN SIGN, U+0021 EXCLAMATION MARK, U+002D HYPHEN-MINUS,
184: and U+002D HYPHEN-MINUS ("<!--"), then set the escape flag to true. */
185: if(($this->content_model === self::RCDATA || $this->content_model ===
186: self::CDATA) && $this->escape === false &&
187: $this->char >= 3 && $this->character($this->char - 4, 4) === '<!--') {
188: $this->escape = true;
189: }
190:
191: /* In any case, emit the input character as a character token. Stay
192: in the data state. */
193: $this->emitToken(array(
194: 'type' => self::CHARACTR,
195: 'data' => $char
196: ));
197:
198: /* U+003C LESS-THAN SIGN (<) */
199: } elseif($char === '<' && ($this->content_model === self::PCDATA ||
200: (($this->content_model === self::RCDATA ||
201: $this->content_model === self::CDATA) && $this->escape === false))) {
202: /* When the content model flag is set to the PCDATA state: switch
203: to the tag open state.
204:
205: When the content model flag is set to either the RCDATA state or
206: the CDATA state and the escape flag is false: switch to the tag
207: open state.
208:
209: Otherwise: treat it as per the "anything else" entry below. */
210: $this->state = 'tagOpen';
211:
212: /* U+003E GREATER-THAN SIGN (>) */
213: } elseif($char === '>') {
214: /* If the content model flag is set to either the RCDATA state or
215: the CDATA state, and the escape flag is true, and the last three
216: characters in the input stream including this one are U+002D
217: HYPHEN-MINUS, U+002D HYPHEN-MINUS, U+003E GREATER-THAN SIGN ("-->"),
218: set the escape flag to false. */
219: if(($this->content_model === self::RCDATA ||
220: $this->content_model === self::CDATA) && $this->escape === true &&
221: $this->character($this->char, 3) === '-->') {
222: $this->escape = false;
223: }
224:
225: /* In any case, emit the input character as a character token.
226: Stay in the data state. */
227: $this->emitToken(array(
228: 'type' => self::CHARACTR,
229: 'data' => $char
230: ));
231:
232: } elseif($this->char === $this->EOF) {
233: /* EOF
234: Emit an end-of-file token. */
235: $this->EOF();
236:
237: } elseif($this->content_model === self::PLAINTEXT) {
238: /* When the content model flag is set to the PLAINTEXT state
239: THIS DIFFERS GREATLY FROM THE SPEC: Get the remaining characters of
240: the text and emit it as a character token. */
241: $this->emitToken(array(
242: 'type' => self::CHARACTR,
243: 'data' => substr($this->data, $this->char)
244: ));
245:
246: $this->EOF();
247:
248: } else {
249: /* Anything else
250: THIS DIFFERS GREATLY FROM THE SPEC: Get as many character that
251: otherwise would also be treated as a character token and emit it
252: as a single character token. Stay in the data state. */
253: $len = strcspn($this->data, '<&', $this->char);
254: $char = substr($this->data, $this->char, $len);
255: $this->char += $len - 1;
256:
257: $this->emitToken(array(
258: 'type' => self::CHARACTR,
259: 'data' => $char
260: ));
261:
262: $this->state = 'data';
263: }
264: }
265:
266: private function entityDataState() {
267: // Attempt to consume an entity.
268: $entity = $this->entity();
269:
270: // If nothing is returned, emit a U+0026 AMPERSAND character token.
271: // Otherwise, emit the character token that was returned.
272: $char = (!$entity) ? '&' : $entity;
273: $this->emitToken(array(
274: 'type' => self::CHARACTR,
275: 'data' => $char
276: ));
277:
278: // Finally, switch to the data state.
279: $this->state = 'data';
280: }
281:
282: private function tagOpenState() {
283: switch($this->content_model) {
284: case self::RCDATA:
285: case self::CDATA:
286: /* If the next input character is a U+002F SOLIDUS (/) character,
287: consume it and switch to the close tag open state. If the next
288: input character is not a U+002F SOLIDUS (/) character, emit a
289: U+003C LESS-THAN SIGN character token and switch to the data
290: state to process the next input character. */
291: if($this->character($this->char + 1) === '/') {
292: $this->char++;
293: $this->state = 'closeTagOpen';
294:
295: } else {
296: $this->emitToken(array(
297: 'type' => self::CHARACTR,
298: 'data' => '<'
299: ));
300:
301: $this->state = 'data';
302: }
303: break;
304:
305: case self::PCDATA:
306: // If the content model flag is set to the PCDATA state
307: // Consume the next input character:
308: $this->char++;
309: $char = $this->char();
310:
311: if($char === '!') {
312: /* U+0021 EXCLAMATION MARK (!)
313: Switch to the markup declaration open state. */
314: $this->state = 'markupDeclarationOpen';
315:
316: } elseif($char === '/') {
317: /* U+002F SOLIDUS (/)
318: Switch to the close tag open state. */
319: $this->state = 'closeTagOpen';
320:
321: } elseif(preg_match('/^[A-Za-z]$/', $char)) {
322: /* U+0041 LATIN LETTER A through to U+005A LATIN LETTER Z
323: Create a new start tag token, set its tag name to the lowercase
324: version of the input character (add 0x0020 to the character's code
325: point), then switch to the tag name state. (Don't emit the token
326: yet; further details will be filled in before it is emitted.) */
327: $this->token = array(
328: 'name' => strtolower($char),
329: 'type' => self::STARTTAG,
330: 'attr' => array()
331: );
332:
333: $this->state = 'tagName';
334:
335: } elseif($char === '>') {
336: /* U+003E GREATER-THAN SIGN (>)
337: Parse error. Emit a U+003C LESS-THAN SIGN character token and a
338: U+003E GREATER-THAN SIGN character token. Switch to the data state. */
339: $this->emitToken(array(
340: 'type' => self::CHARACTR,
341: 'data' => '<>'
342: ));
343:
344: $this->state = 'data';
345:
346: } elseif($char === '?') {
347: /* U+003F QUESTION MARK (?)
348: Parse error. Switch to the bogus comment state. */
349: $this->state = 'bogusComment';
350:
351: } else {
352: /* Anything else
353: Parse error. Emit a U+003C LESS-THAN SIGN character token and
354: reconsume the current input character in the data state. */
355: $this->emitToken(array(
356: 'type' => self::CHARACTR,
357: 'data' => '<'
358: ));
359:
360: $this->char--;
361: $this->state = 'data';
362: }
363: break;
364: }
365: }
366:
367: private function closeTagOpenState() {
368: $next_node = strtolower($this->characters('A-Za-z', $this->char + 1));
369: $the_same = count($this->tree->stack) > 0 && $next_node === end($this->tree->stack)->nodeName;
370:
371: if(($this->content_model === self::RCDATA || $this->content_model === self::CDATA) &&
372: (!$the_same || ($the_same && (!preg_match('/[\t\n\x0b\x0c >\/]/',
373: $this->character($this->char + 1 + strlen($next_node))) || $this->EOF === $this->char)))) {
374: /* If the content model flag is set to the RCDATA or CDATA states then
375: examine the next few characters. If they do not match the tag name of
376: the last start tag token emitted (case insensitively), or if they do but
377: they are not immediately followed by one of the following characters:
378: * U+0009 CHARACTER TABULATION
379: * U+000A LINE FEED (LF)
380: * U+000B LINE TABULATION
381: * U+000C FORM FEED (FF)
382: * U+0020 SPACE
383: * U+003E GREATER-THAN SIGN (>)
384: * U+002F SOLIDUS (/)
385: * EOF
386: ...then there is a parse error. Emit a U+003C LESS-THAN SIGN character
387: token, a U+002F SOLIDUS character token, and switch to the data state
388: to process the next input character. */
389: $this->emitToken(array(
390: 'type' => self::CHARACTR,
391: 'data' => '</'
392: ));
393:
394: $this->state = 'data';
395:
396: } else {
397: /* Otherwise, if the content model flag is set to the PCDATA state,
398: or if the next few characters do match that tag name, consume the
399: next input character: */
400: $this->char++;
401: $char = $this->char();
402:
403: if(preg_match('/^[A-Za-z]$/', $char)) {
404: /* U+0041 LATIN LETTER A through to U+005A LATIN LETTER Z
405: Create a new end tag token, set its tag name to the lowercase version
406: of the input character (add 0x0020 to the character's code point), then
407: switch to the tag name state. (Don't emit the token yet; further details
408: will be filled in before it is emitted.) */
409: $this->token = array(
410: 'name' => strtolower($char),
411: 'type' => self::ENDTAG
412: );
413:
414: $this->state = 'tagName';
415:
416: } elseif($char === '>') {
417: /* U+003E GREATER-THAN SIGN (>)
418: Parse error. Switch to the data state. */
419: $this->state = 'data';
420:
421: } elseif($this->char === $this->EOF) {
422: /* EOF
423: Parse error. Emit a U+003C LESS-THAN SIGN character token and a U+002F
424: SOLIDUS character token. Reconsume the EOF character in the data state. */
425: $this->emitToken(array(
426: 'type' => self::CHARACTR,
427: 'data' => '</'
428: ));
429:
430: $this->char--;
431: $this->state = 'data';
432:
433: } else {
434: /* Parse error. Switch to the bogus comment state. */
435: $this->state = 'bogusComment';
436: }
437: }
438: }
439:
440: private function tagNameState() {
441: // Consume the next input character:
442: $this->char++;
443: $char = $this->character($this->char);
444:
445: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
446: /* U+0009 CHARACTER TABULATION
447: U+000A LINE FEED (LF)
448: U+000B LINE TABULATION
449: U+000C FORM FEED (FF)
450: U+0020 SPACE
451: Switch to the before attribute name state. */
452: $this->state = 'beforeAttributeName';
453:
454: } elseif($char === '>') {
455: /* U+003E GREATER-THAN SIGN (>)
456: Emit the current tag token. Switch to the data state. */
457: $this->emitToken($this->token);
458: $this->state = 'data';
459:
460: } elseif($this->char === $this->EOF) {
461: /* EOF
462: Parse error. Emit the current tag token. Reconsume the EOF
463: character in the data state. */
464: $this->emitToken($this->token);
465:
466: $this->char--;
467: $this->state = 'data';
468:
469: } elseif($char === '/') {
470: /* U+002F SOLIDUS (/)
471: Parse error unless this is a permitted slash. Switch to the before
472: attribute name state. */
473: $this->state = 'beforeAttributeName';
474:
475: } else {
476: /* Anything else
477: Append the current input character to the current tag token's tag name.
478: Stay in the tag name state. */
479: $this->token['name'] .= strtolower($char);
480: $this->state = 'tagName';
481: }
482: }
483:
484: private function beforeAttributeNameState() {
485: // Consume the next input character:
486: $this->char++;
487: $char = $this->character($this->char);
488:
489: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
490: /* U+0009 CHARACTER TABULATION
491: U+000A LINE FEED (LF)
492: U+000B LINE TABULATION
493: U+000C FORM FEED (FF)
494: U+0020 SPACE
495: Stay in the before attribute name state. */
496: $this->state = 'beforeAttributeName';
497:
498: } elseif($char === '>') {
499: /* U+003E GREATER-THAN SIGN (>)
500: Emit the current tag token. Switch to the data state. */
501: $this->emitToken($this->token);
502: $this->state = 'data';
503:
504: } elseif($char === '/') {
505: /* U+002F SOLIDUS (/)
506: Parse error unless this is a permitted slash. Stay in the before
507: attribute name state. */
508: $this->state = 'beforeAttributeName';
509:
510: } elseif($this->char === $this->EOF) {
511: /* EOF
512: Parse error. Emit the current tag token. Reconsume the EOF
513: character in the data state. */
514: $this->emitToken($this->token);
515:
516: $this->char--;
517: $this->state = 'data';
518:
519: } else {
520: /* Anything else
521: Start a new attribute in the current tag token. Set that attribute's
522: name to the current input character, and its value to the empty string.
523: Switch to the attribute name state. */
524: $this->token['attr'][] = array(
525: 'name' => strtolower($char),
526: 'value' => null
527: );
528:
529: $this->state = 'attributeName';
530: }
531: }
532:
533: private function attributeNameState() {
534: // Consume the next input character:
535: $this->char++;
536: $char = $this->character($this->char);
537:
538: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
539: /* U+0009 CHARACTER TABULATION
540: U+000A LINE FEED (LF)
541: U+000B LINE TABULATION
542: U+000C FORM FEED (FF)
543: U+0020 SPACE
544: Stay in the before attribute name state. */
545: $this->state = 'afterAttributeName';
546:
547: } elseif($char === '=') {
548: /* U+003D EQUALS SIGN (=)
549: Switch to the before attribute value state. */
550: $this->state = 'beforeAttributeValue';
551:
552: } elseif($char === '>') {
553: /* U+003E GREATER-THAN SIGN (>)
554: Emit the current tag token. Switch to the data state. */
555: $this->emitToken($this->token);
556: $this->state = 'data';
557:
558: } elseif($char === '/' && $this->character($this->char + 1) !== '>') {
559: /* U+002F SOLIDUS (/)
560: Parse error unless this is a permitted slash. Switch to the before
561: attribute name state. */
562: $this->state = 'beforeAttributeName';
563:
564: } elseif($this->char === $this->EOF) {
565: /* EOF
566: Parse error. Emit the current tag token. Reconsume the EOF
567: character in the data state. */
568: $this->emitToken($this->token);
569:
570: $this->char--;
571: $this->state = 'data';
572:
573: } else {
574: /* Anything else
575: Append the current input character to the current attribute's name.
576: Stay in the attribute name state. */
577: $last = count($this->token['attr']) - 1;
578: $this->token['attr'][$last]['name'] .= strtolower($char);
579:
580: $this->state = 'attributeName';
581: }
582: }
583:
584: private function afterAttributeNameState() {
585: // Consume the next input character:
586: $this->char++;
587: $char = $this->character($this->char);
588:
589: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
590: /* U+0009 CHARACTER TABULATION
591: U+000A LINE FEED (LF)
592: U+000B LINE TABULATION
593: U+000C FORM FEED (FF)
594: U+0020 SPACE
595: Stay in the after attribute name state. */
596: $this->state = 'afterAttributeName';
597:
598: } elseif($char === '=') {
599: /* U+003D EQUALS SIGN (=)
600: Switch to the before attribute value state. */
601: $this->state = 'beforeAttributeValue';
602:
603: } elseif($char === '>') {
604: /* U+003E GREATER-THAN SIGN (>)
605: Emit the current tag token. Switch to the data state. */
606: $this->emitToken($this->token);
607: $this->state = 'data';
608:
609: } elseif($char === '/' && $this->character($this->char + 1) !== '>') {
610: /* U+002F SOLIDUS (/)
611: Parse error unless this is a permitted slash. Switch to the
612: before attribute name state. */
613: $this->state = 'beforeAttributeName';
614:
615: } elseif($this->char === $this->EOF) {
616: /* EOF
617: Parse error. Emit the current tag token. Reconsume the EOF
618: character in the data state. */
619: $this->emitToken($this->token);
620:
621: $this->char--;
622: $this->state = 'data';
623:
624: } else {
625: /* Anything else
626: Start a new attribute in the current tag token. Set that attribute's
627: name to the current input character, and its value to the empty string.
628: Switch to the attribute name state. */
629: $this->token['attr'][] = array(
630: 'name' => strtolower($char),
631: 'value' => null
632: );
633:
634: $this->state = 'attributeName';
635: }
636: }
637:
638: private function beforeAttributeValueState() {
639: // Consume the next input character:
640: $this->char++;
641: $char = $this->character($this->char);
642:
643: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
644: /* U+0009 CHARACTER TABULATION
645: U+000A LINE FEED (LF)
646: U+000B LINE TABULATION
647: U+000C FORM FEED (FF)
648: U+0020 SPACE
649: Stay in the before attribute value state. */
650: $this->state = 'beforeAttributeValue';
651:
652: } elseif($char === '"') {
653: /* U+0022 QUOTATION MARK (")
654: Switch to the attribute value (double-quoted) state. */
655: $this->state = 'attributeValueDoubleQuoted';
656:
657: } elseif($char === '&') {
658: /* U+0026 AMPERSAND (&)
659: Switch to the attribute value (unquoted) state and reconsume
660: this input character. */
661: $this->char--;
662: $this->state = 'attributeValueUnquoted';
663:
664: } elseif($char === '\'') {
665: /* U+0027 APOSTROPHE (')
666: Switch to the attribute value (single-quoted) state. */
667: $this->state = 'attributeValueSingleQuoted';
668:
669: } elseif($char === '>') {
670: /* U+003E GREATER-THAN SIGN (>)
671: Emit the current tag token. Switch to the data state. */
672: $this->emitToken($this->token);
673: $this->state = 'data';
674:
675: } else {
676: /* Anything else
677: Append the current input character to the current attribute's value.
678: Switch to the attribute value (unquoted) state. */
679: $last = count($this->token['attr']) - 1;
680: $this->token['attr'][$last]['value'] .= $char;
681:
682: $this->state = 'attributeValueUnquoted';
683: }
684: }
685:
686: private function attributeValueDoubleQuotedState() {
687: // Consume the next input character:
688: $this->char++;
689: $char = $this->character($this->char);
690:
691: if($char === '"') {
692: /* U+0022 QUOTATION MARK (")
693: Switch to the before attribute name state. */
694: $this->state = 'beforeAttributeName';
695:
696: } elseif($char === '&') {
697: /* U+0026 AMPERSAND (&)
698: Switch to the entity in attribute value state. */
699: $this->entityInAttributeValueState('double');
700:
701: } elseif($this->char === $this->EOF) {
702: /* EOF
703: Parse error. Emit the current tag token. Reconsume the character
704: in the data state. */
705: $this->emitToken($this->token);
706:
707: $this->char--;
708: $this->state = 'data';
709:
710: } else {
711: /* Anything else
712: Append the current input character to the current attribute's value.
713: Stay in the attribute value (double-quoted) state. */
714: $last = count($this->token['attr']) - 1;
715: $this->token['attr'][$last]['value'] .= $char;
716:
717: $this->state = 'attributeValueDoubleQuoted';
718: }
719: }
720:
721: private function attributeValueSingleQuotedState() {
722: // Consume the next input character:
723: $this->char++;
724: $char = $this->character($this->char);
725:
726: if($char === '\'') {
727: /* U+0022 QUOTATION MARK (')
728: Switch to the before attribute name state. */
729: $this->state = 'beforeAttributeName';
730:
731: } elseif($char === '&') {
732: /* U+0026 AMPERSAND (&)
733: Switch to the entity in attribute value state. */
734: $this->entityInAttributeValueState('single');
735:
736: } elseif($this->char === $this->EOF) {
737: /* EOF
738: Parse error. Emit the current tag token. Reconsume the character
739: in the data state. */
740: $this->emitToken($this->token);
741:
742: $this->char--;
743: $this->state = 'data';
744:
745: } else {
746: /* Anything else
747: Append the current input character to the current attribute's value.
748: Stay in the attribute value (single-quoted) state. */
749: $last = count($this->token['attr']) - 1;
750: $this->token['attr'][$last]['value'] .= $char;
751:
752: $this->state = 'attributeValueSingleQuoted';
753: }
754: }
755:
756: private function attributeValueUnquotedState() {
757: // Consume the next input character:
758: $this->char++;
759: $char = $this->character($this->char);
760:
761: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
762: /* U+0009 CHARACTER TABULATION
763: U+000A LINE FEED (LF)
764: U+000B LINE TABULATION
765: U+000C FORM FEED (FF)
766: U+0020 SPACE
767: Switch to the before attribute name state. */
768: $this->state = 'beforeAttributeName';
769:
770: } elseif($char === '&') {
771: /* U+0026 AMPERSAND (&)
772: Switch to the entity in attribute value state. */
773: $this->entityInAttributeValueState();
774:
775: } elseif($char === '>') {
776: /* U+003E GREATER-THAN SIGN (>)
777: Emit the current tag token. Switch to the data state. */
778: $this->emitToken($this->token);
779: $this->state = 'data';
780:
781: } else {
782: /* Anything else
783: Append the current input character to the current attribute's value.
784: Stay in the attribute value (unquoted) state. */
785: $last = count($this->token['attr']) - 1;
786: $this->token['attr'][$last]['value'] .= $char;
787:
788: $this->state = 'attributeValueUnquoted';
789: }
790: }
791:
792: private function entityInAttributeValueState() {
793: // Attempt to consume an entity.
794: $entity = $this->entity();
795:
796: // If nothing is returned, append a U+0026 AMPERSAND character to the
797: // current attribute's value. Otherwise, emit the character token that
798: // was returned.
799: $char = (!$entity)
800: ? '&'
801: : $entity;
802:
803: $last = count($this->token['attr']) - 1;
804: $this->token['attr'][$last]['value'] .= $char;
805: }
806:
807: private function bogusCommentState() {
808: /* Consume every character up to the first U+003E GREATER-THAN SIGN
809: character (>) or the end of the file (EOF), whichever comes first. Emit
810: a comment token whose data is the concatenation of all the characters
811: starting from and including the character that caused the state machine
812: to switch into the bogus comment state, up to and including the last
813: consumed character before the U+003E character, if any, or up to the
814: end of the file otherwise. (If the comment was started by the end of
815: the file (EOF), the token is empty.) */
816: $data = $this->characters('^>', $this->char);
817: $this->emitToken(array(
818: 'data' => $data,
819: 'type' => self::COMMENT
820: ));
821:
822: $this->char += strlen($data);
823:
824: /* Switch to the data state. */
825: $this->state = 'data';
826:
827: /* If the end of the file was reached, reconsume the EOF character. */
828: if($this->char === $this->EOF) {
829: $this->char = $this->EOF - 1;
830: }
831: }
832:
833: private function markupDeclarationOpenState() {
834: /* If the next two characters are both U+002D HYPHEN-MINUS (-)
835: characters, consume those two characters, create a comment token whose
836: data is the empty string, and switch to the comment state. */
837: if($this->character($this->char + 1, 2) === '--') {
838: $this->char += 2;
839: $this->state = 'comment';
840: $this->token = array(
841: 'data' => null,
842: 'type' => self::COMMENT
843: );
844:
845: /* Otherwise if the next seven chacacters are a case-insensitive match
846: for the word "DOCTYPE", then consume those characters and switch to the
847: DOCTYPE state. */
848: } elseif(strtolower($this->character($this->char + 1, 7)) === 'doctype') {
849: $this->char += 7;
850: $this->state = 'doctype';
851:
852: /* Otherwise, is is a parse error. Switch to the bogus comment state.
853: The next character that is consumed, if any, is the first character
854: that will be in the comment. */
855: } else {
856: $this->char++;
857: $this->state = 'bogusComment';
858: }
859: }
860:
861: private function commentState() {
862: /* Consume the next input character: */
863: $this->char++;
864: $char = $this->char();
865:
866: /* U+002D HYPHEN-MINUS (-) */
867: if($char === '-') {
868: /* Switch to the comment dash state */
869: $this->state = 'commentDash';
870:
871: /* EOF */
872: } elseif($this->char === $this->EOF) {
873: /* Parse error. Emit the comment token. Reconsume the EOF character
874: in the data state. */
875: $this->emitToken($this->token);
876: $this->char--;
877: $this->state = 'data';
878:
879: /* Anything else */
880: } else {
881: /* Append the input character to the comment token's data. Stay in
882: the comment state. */
883: $this->token['data'] .= $char;
884: }
885: }
886:
887: private function commentDashState() {
888: /* Consume the next input character: */
889: $this->char++;
890: $char = $this->char();
891:
892: /* U+002D HYPHEN-MINUS (-) */
893: if($char === '-') {
894: /* Switch to the comment end state */
895: $this->state = 'commentEnd';
896:
897: /* EOF */
898: } elseif($this->char === $this->EOF) {
899: /* Parse error. Emit the comment token. Reconsume the EOF character
900: in the data state. */
901: $this->emitToken($this->token);
902: $this->char--;
903: $this->state = 'data';
904:
905: /* Anything else */
906: } else {
907: /* Append a U+002D HYPHEN-MINUS (-) character and the input
908: character to the comment token's data. Switch to the comment state. */
909: $this->token['data'] .= '-'.$char;
910: $this->state = 'comment';
911: }
912: }
913:
914: private function commentEndState() {
915: /* Consume the next input character: */
916: $this->char++;
917: $char = $this->char();
918:
919: if($char === '>') {
920: $this->emitToken($this->token);
921: $this->state = 'data';
922:
923: } elseif($char === '-') {
924: $this->token['data'] .= '-';
925:
926: } elseif($this->char === $this->EOF) {
927: $this->emitToken($this->token);
928: $this->char--;
929: $this->state = 'data';
930:
931: } else {
932: $this->token['data'] .= '--'.$char;
933: $this->state = 'comment';
934: }
935: }
936:
937: private function doctypeState() {
938: /* Consume the next input character: */
939: $this->char++;
940: $char = $this->char();
941:
942: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
943: $this->state = 'beforeDoctypeName';
944:
945: } else {
946: $this->char--;
947: $this->state = 'beforeDoctypeName';
948: }
949: }
950:
951: private function beforeDoctypeNameState() {
952: /* Consume the next input character: */
953: $this->char++;
954: $char = $this->char();
955:
956: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
957: // Stay in the before DOCTYPE name state.
958:
959: } elseif(preg_match('/^[a-z]$/', $char)) {
960: $this->token = array(
961: 'name' => strtoupper($char),
962: 'type' => self::DOCTYPE,
963: 'error' => true
964: );
965:
966: $this->state = 'doctypeName';
967:
968: } elseif($char === '>') {
969: $this->emitToken(array(
970: 'name' => null,
971: 'type' => self::DOCTYPE,
972: 'error' => true
973: ));
974:
975: $this->state = 'data';
976:
977: } elseif($this->char === $this->EOF) {
978: $this->emitToken(array(
979: 'name' => null,
980: 'type' => self::DOCTYPE,
981: 'error' => true
982: ));
983:
984: $this->char--;
985: $this->state = 'data';
986:
987: } else {
988: $this->token = array(
989: 'name' => $char,
990: 'type' => self::DOCTYPE,
991: 'error' => true
992: );
993:
994: $this->state = 'doctypeName';
995: }
996: }
997:
998: private function doctypeNameState() {
999: /* Consume the next input character: */
1000: $this->char++;
1001: $char = $this->char();
1002:
1003: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
1004: $this->state = 'AfterDoctypeName';
1005:
1006: } elseif($char === '>') {
1007: $this->emitToken($this->token);
1008: $this->state = 'data';
1009:
1010: } elseif(preg_match('/^[a-z]$/', $char)) {
1011: $this->token['name'] .= strtoupper($char);
1012:
1013: } elseif($this->char === $this->EOF) {
1014: $this->emitToken($this->token);
1015: $this->char--;
1016: $this->state = 'data';
1017:
1018: } else {
1019: $this->token['name'] .= $char;
1020: }
1021:
1022: $this->token['error'] = ($this->token['name'] === 'HTML')
1023: ? false
1024: : true;
1025: }
1026:
1027: private function afterDoctypeNameState() {
1028: /* Consume the next input character: */
1029: $this->char++;
1030: $char = $this->char();
1031:
1032: if(preg_match('/^[\t\n\x0b\x0c ]$/', $char)) {
1033: // Stay in the DOCTYPE name state.
1034:
1035: } elseif($char === '>') {
1036: $this->emitToken($this->token);
1037: $this->state = 'data';
1038:
1039: } elseif($this->char === $this->EOF) {
1040: $this->emitToken($this->token);
1041: $this->char--;
1042: $this->state = 'data';
1043:
1044: } else {
1045: $this->token['error'] = true;
1046: $this->state = 'bogusDoctype';
1047: }
1048: }
1049:
1050: private function bogusDoctypeState() {
1051: /* Consume the next input character: */
1052: $this->char++;
1053: $char = $this->char();
1054:
1055: if($char === '>') {
1056: $this->emitToken($this->token);
1057: $this->state = 'data';
1058:
1059: } elseif($this->char === $this->EOF) {
1060: $this->emitToken($this->token);
1061: $this->char--;
1062: $this->state = 'data';
1063:
1064: } else {
1065: // Stay in the bogus DOCTYPE state.
1066: }
1067: }
1068:
1069: private function entity() {
1070: $start = $this->char;
1071:
1072: // This section defines how to consume an entity. This definition is
1073: // used when parsing entities in text and in attributes.
1074:
1075: // The behaviour depends on the identity of the next character (the
1076: // one immediately after the U+0026 AMPERSAND character):
1077:
1078: switch($this->character($this->char + 1)) {
1079: // U+0023 NUMBER SIGN (#)
1080: case '#':
1081:
1082: // The behaviour further depends on the character after the
1083: // U+0023 NUMBER SIGN:
1084: switch($this->character($this->char + 1)) {
1085: // U+0078 LATIN SMALL LETTER X
1086: // U+0058 LATIN CAPITAL LETTER X
1087: case 'x':
1088: case 'X':
1089: // Follow the steps below, but using the range of
1090: // characters U+0030 DIGIT ZERO through to U+0039 DIGIT
1091: // NINE, U+0061 LATIN SMALL LETTER A through to U+0066
1092: // LATIN SMALL LETTER F, and U+0041 LATIN CAPITAL LETTER
1093: // A, through to U+0046 LATIN CAPITAL LETTER F (in other
1094: // words, 0-9, A-F, a-f).
1095: $char = 1;
1096: $char_class = '0-9A-Fa-f';
1097: break;
1098:
1099: // Anything else
1100: default:
1101: // Follow the steps below, but using the range of
1102: // characters U+0030 DIGIT ZERO through to U+0039 DIGIT
1103: // NINE (i.e. just 0-9).
1104: $char = 0;
1105: $char_class = '0-9';
1106: break;
1107: }
1108:
1109: // Consume as many characters as match the range of characters
1110: // given above.
1111: $this->char++;
1112: $e_name = $this->characters($char_class, $this->char + $char + 1);
1113: $entity = $this->character($start, $this->char);
1114: $cond = strlen($e_name) > 0;
1115:
1116: // The rest of the parsing happens bellow.
1117: break;
1118:
1119: // Anything else
1120: default:
1121: // Consume the maximum number of characters possible, with the
1122: // consumed characters case-sensitively matching one of the
1123: // identifiers in the first column of the entities table.
1124: $e_name = $this->characters('0-9A-Za-z;', $this->char + 1);
1125: $len = strlen($e_name);
1126:
1127: for($c = 1; $c <= $len; $c++) {
1128: $id = substr($e_name, 0, $c);
1129: $this->char++;
1130:
1131: if(in_array($id, $this->entities)) {
1132: if ($e_name[$c-1] !== ';') {
1133: if ($c < $len && $e_name[$c] == ';') {
1134: $this->char++; // consume extra semicolon
1135: }
1136: }
1137: $entity = $id;
1138: break;
1139: }
1140: }
1141:
1142: $cond = isset($entity);
1143: // The rest of the parsing happens bellow.
1144: break;
1145: }
1146:
1147: if(!$cond) {
1148: // If no match can be made, then this is a parse error. No
1149: // characters are consumed, and nothing is returned.
1150: $this->char = $start;
1151: return false;
1152: }
1153:
1154: // Return a character token for the character corresponding to the
1155: // entity name (as given by the second column of the entities table).
1156: return html_entity_decode('&'.$entity.';', ENT_QUOTES, 'UTF-8');
1157: }
1158:
1159: private function emitToken($token) {
1160: $emit = $this->tree->emitToken($token);
1161:
1162: if(is_int($emit)) {
1163: $this->content_model = $emit;
1164:
1165: } elseif($token['type'] === self::ENDTAG) {
1166: $this->content_model = self::PCDATA;
1167: }
1168: }
1169:
1170: private function EOF() {
1171: $this->state = null;
1172: $this->tree->emitToken(array(
1173: 'type' => self::EOF
1174: ));
1175: }
1176: }
1177:
1178: class HTML5TreeConstructer {
1179: public $stack = array();
1180:
1181: private $phase;
1182: private $mode;
1183: private $dom;
1184: private $foster_parent = null;
1185: private $a_formatting = array();
1186:
1187: private $head_pointer = null;
1188: private $form_pointer = null;
1189:
1190: private $scoping = array('button','caption','html','marquee','object','table','td','th');
1191: private $formatting = array('a','b','big','em','font','i','nobr','s','small','strike','strong','tt','u');
1192: private $special = array('address','area','base','basefont','bgsound',
1193: 'blockquote','body','br','center','col','colgroup','dd','dir','div','dl',
1194: 'dt','embed','fieldset','form','frame','frameset','h1','h2','h3','h4','h5',
1195: 'h6','head','hr','iframe','image','img','input','isindex','li','link',
1196: 'listing','menu','meta','noembed','noframes','noscript','ol','optgroup',
1197: 'option','p','param','plaintext','pre','script','select','spacer','style',
1198: 'tbody','textarea','tfoot','thead','title','tr','ul','wbr');
1199:
1200: // The different phases.
1201: const INIT_PHASE = 0;
1202: const ROOT_PHASE = 1;
1203: const MAIN_PHASE = 2;
1204: const END_PHASE = 3;
1205:
1206: // The different insertion modes for the main phase.
1207: const BEFOR_HEAD = 0;
1208: const IN_HEAD = 1;
1209: const AFTER_HEAD = 2;
1210: const IN_BODY = 3;
1211: const IN_TABLE = 4;
1212: const IN_CAPTION = 5;
1213: const IN_CGROUP = 6;
1214: const IN_TBODY = 7;
1215: const IN_ROW = 8;
1216: const IN_CELL = 9;
1217: const IN_SELECT = 10;
1218: const AFTER_BODY = 11;
1219: const IN_FRAME = 12;
1220: const AFTR_FRAME = 13;
1221:
1222: // The different types of elements.
1223: const SPECIAL = 0;
1224: const SCOPING = 1;
1225: const FORMATTING = 2;
1226: const PHRASING = 3;
1227:
1228: const MARKER = 0;
1229:
1230: public function __construct() {
1231: $this->phase = self::INIT_PHASE;
1232: $this->mode = self::BEFOR_HEAD;
1233: $this->dom = new DOMDocument;
1234:
1235: $this->dom->encoding = 'UTF-8';
1236: $this->dom->preserveWhiteSpace = true;
1237: $this->dom->substituteEntities = true;
1238: $this->dom->strictErrorChecking = false;
1239: }
1240:
1241: // Process tag tokens
1242: public function emitToken($token) {
1243: switch($this->phase) {
1244: case self::INIT_PHASE: return $this->initPhase($token); break;
1245: case self::ROOT_PHASE: return $this->rootElementPhase($token); break;
1246: case self::MAIN_PHASE: return $this->mainPhase($token); break;
1247: case self::END_PHASE : return $this->trailingEndPhase($token); break;
1248: }
1249: }
1250:
1251: private function initPhase($token) {
1252: /* Initially, the tree construction stage must handle each token
1253: emitted from the tokenisation stage as follows: */
1254:
1255: /* A DOCTYPE token that is marked as being in error
1256: A comment token
1257: A start tag token
1258: An end tag token
1259: A character token that is not one of one of U+0009 CHARACTER TABULATION,
1260: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
1261: or U+0020 SPACE
1262: An end-of-file token */
1263: if((isset($token['error']) && $token['error']) ||
1264: $token['type'] === HTML5::COMMENT ||
1265: $token['type'] === HTML5::STARTTAG ||
1266: $token['type'] === HTML5::ENDTAG ||
1267: $token['type'] === HTML5::EOF ||
1268: ($token['type'] === HTML5::CHARACTR && isset($token['data']) &&
1269: !preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data']))) {
1270: /* This specification does not define how to handle this case. In
1271: particular, user agents may ignore the entirety of this specification
1272: altogether for such documents, and instead invoke special parse modes
1273: with a greater emphasis on backwards compatibility. */
1274:
1275: $this->phase = self::ROOT_PHASE;
1276: return $this->rootElementPhase($token);
1277:
1278: /* A DOCTYPE token marked as being correct */
1279: } elseif(isset($token['error']) && !$token['error']) {
1280: /* Append a DocumentType node to the Document node, with the name
1281: attribute set to the name given in the DOCTYPE token (which will be
1282: "HTML"), and the other attributes specific to DocumentType objects
1283: set to null, empty lists, or the empty string as appropriate. */
1284: $doctype = new DOMDocumentType(null, null, 'HTML');
1285:
1286: /* Then, switch to the root element phase of the tree construction
1287: stage. */
1288: $this->phase = self::ROOT_PHASE;
1289:
1290: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
1291: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
1292: or U+0020 SPACE */
1293: } elseif(isset($token['data']) && preg_match('/^[\t\n\x0b\x0c ]+$/',
1294: $token['data'])) {
1295: /* Append that character to the Document node. */
1296: $text = $this->dom->createTextNode($token['data']);
1297: $this->dom->appendChild($text);
1298: }
1299: }
1300:
1301: private function rootElementPhase($token) {
1302: /* After the initial phase, as each token is emitted from the tokenisation
1303: stage, it must be processed as described in this section. */
1304:
1305: /* A DOCTYPE token */
1306: if($token['type'] === HTML5::DOCTYPE) {
1307: // Parse error. Ignore the token.
1308:
1309: /* A comment token */
1310: } elseif($token['type'] === HTML5::COMMENT) {
1311: /* Append a Comment node to the Document object with the data
1312: attribute set to the data given in the comment token. */
1313: $comment = $this->dom->createComment($token['data']);
1314: $this->dom->appendChild($comment);
1315:
1316: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
1317: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
1318: or U+0020 SPACE */
1319: } elseif($token['type'] === HTML5::CHARACTR &&
1320: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
1321: /* Append that character to the Document node. */
1322: $text = $this->dom->createTextNode($token['data']);
1323: $this->dom->appendChild($text);
1324:
1325: /* A character token that is not one of U+0009 CHARACTER TABULATION,
1326: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED
1327: (FF), or U+0020 SPACE
1328: A start tag token
1329: An end tag token
1330: An end-of-file token */
1331: } elseif(($token['type'] === HTML5::CHARACTR &&
1332: !preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) ||
1333: $token['type'] === HTML5::STARTTAG ||
1334: $token['type'] === HTML5::ENDTAG ||
1335: $token['type'] === HTML5::EOF) {
1336: /* Create an HTMLElement node with the tag name html, in the HTML
1337: namespace. Append it to the Document object. Switch to the main
1338: phase and reprocess the current token. */
1339: $html = $this->dom->createElement('html');
1340: $this->dom->appendChild($html);
1341: $this->stack[] = $html;
1342:
1343: $this->phase = self::MAIN_PHASE;
1344: return $this->mainPhase($token);
1345: }
1346: }
1347:
1348: private function mainPhase($token) {
1349: /* Tokens in the main phase must be handled as follows: */
1350:
1351: /* A DOCTYPE token */
1352: if($token['type'] === HTML5::DOCTYPE) {
1353: // Parse error. Ignore the token.
1354:
1355: /* A start tag token with the tag name "html" */
1356: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'html') {
1357: /* If this start tag token was not the first start tag token, then
1358: it is a parse error. */
1359:
1360: /* For each attribute on the token, check to see if the attribute
1361: is already present on the top element of the stack of open elements.
1362: If it is not, add the attribute and its corresponding value to that
1363: element. */
1364: foreach($token['attr'] as $attr) {
1365: if(!$this->stack[0]->hasAttribute($attr['name'])) {
1366: $this->stack[0]->setAttribute($attr['name'], $attr['value']);
1367: }
1368: }
1369:
1370: /* An end-of-file token */
1371: } elseif($token['type'] === HTML5::EOF) {
1372: /* Generate implied end tags. */
1373: $this->generateImpliedEndTags();
1374:
1375: /* Anything else. */
1376: } else {
1377: /* Depends on the insertion mode: */
1378: switch($this->mode) {
1379: case self::BEFOR_HEAD: return $this->beforeHead($token); break;
1380: case self::IN_HEAD: return $this->inHead($token); break;
1381: case self::AFTER_HEAD: return $this->afterHead($token); break;
1382: case self::IN_BODY: return $this->inBody($token); break;
1383: case self::IN_TABLE: return $this->inTable($token); break;
1384: case self::IN_CAPTION: return $this->inCaption($token); break;
1385: case self::IN_CGROUP: return $this->inColumnGroup($token); break;
1386: case self::IN_TBODY: return $this->inTableBody($token); break;
1387: case self::IN_ROW: return $this->inRow($token); break;
1388: case self::IN_CELL: return $this->inCell($token); break;
1389: case self::IN_SELECT: return $this->inSelect($token); break;
1390: case self::AFTER_BODY: return $this->afterBody($token); break;
1391: case self::IN_FRAME: return $this->inFrameset($token); break;
1392: case self::AFTR_FRAME: return $this->afterFrameset($token); break;
1393: case self::END_PHASE: return $this->trailingEndPhase($token); break;
1394: }
1395: }
1396: }
1397:
1398: private function beforeHead($token) {
1399: /* Handle the token as follows: */
1400:
1401: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
1402: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
1403: or U+0020 SPACE */
1404: if($token['type'] === HTML5::CHARACTR &&
1405: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
1406: /* Append the character to the current node. */
1407: $this->insertText($token['data']);
1408:
1409: /* A comment token */
1410: } elseif($token['type'] === HTML5::COMMENT) {
1411: /* Append a Comment node to the current node with the data attribute
1412: set to the data given in the comment token. */
1413: $this->insertComment($token['data']);
1414:
1415: /* A start tag token with the tag name "head" */
1416: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'head') {
1417: /* Create an element for the token, append the new element to the
1418: current node and push it onto the stack of open elements. */
1419: $element = $this->insertElement($token);
1420:
1421: /* Set the head element pointer to this new element node. */
1422: $this->head_pointer = $element;
1423:
1424: /* Change the insertion mode to "in head". */
1425: $this->mode = self::IN_HEAD;
1426:
1427: /* A start tag token whose tag name is one of: "base", "link", "meta",
1428: "script", "style", "title". Or an end tag with the tag name "html".
1429: Or a character token that is not one of U+0009 CHARACTER TABULATION,
1430: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
1431: or U+0020 SPACE. Or any other start tag token */
1432: } elseif($token['type'] === HTML5::STARTTAG ||
1433: ($token['type'] === HTML5::ENDTAG && $token['name'] === 'html') ||
1434: ($token['type'] === HTML5::CHARACTR && !preg_match('/^[\t\n\x0b\x0c ]$/',
1435: $token['data']))) {
1436: /* Act as if a start tag token with the tag name "head" and no
1437: attributes had been seen, then reprocess the current token. */
1438: $this->beforeHead(array(
1439: 'name' => 'head',
1440: 'type' => HTML5::STARTTAG,
1441: 'attr' => array()
1442: ));
1443:
1444: return $this->inHead($token);
1445:
1446: /* Any other end tag */
1447: } elseif($token['type'] === HTML5::ENDTAG) {
1448: /* Parse error. Ignore the token. */
1449: }
1450: }
1451:
1452: private function inHead($token) {
1453: /* Handle the token as follows: */
1454:
1455: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
1456: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
1457: or U+0020 SPACE.
1458:
1459: THIS DIFFERS FROM THE SPEC: If the current node is either a title, style
1460: or script element, append the character to the current node regardless
1461: of its content. */
1462: if(($token['type'] === HTML5::CHARACTR &&
1463: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) || (
1464: $token['type'] === HTML5::CHARACTR && in_array(end($this->stack)->nodeName,
1465: array('title', 'style', 'script')))) {
1466: /* Append the character to the current node. */
1467: $this->insertText($token['data']);
1468:
1469: /* A comment token */
1470: } elseif($token['type'] === HTML5::COMMENT) {
1471: /* Append a Comment node to the current node with the data attribute
1472: set to the data given in the comment token. */
1473: $this->insertComment($token['data']);
1474:
1475: } elseif($token['type'] === HTML5::ENDTAG &&
1476: in_array($token['name'], array('title', 'style', 'script'))) {
1477: array_pop($this->stack);
1478: return HTML5::PCDATA;
1479:
1480: /* A start tag with the tag name "title" */
1481: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'title') {
1482: /* Create an element for the token and append the new element to the
1483: node pointed to by the head element pointer, or, if that is null
1484: (innerHTML case), to the current node. */
1485: if($this->head_pointer !== null) {
1486: $element = $this->insertElement($token, false);
1487: $this->head_pointer->appendChild($element);
1488:
1489: } else {
1490: $element = $this->insertElement($token);
1491: }
1492:
1493: /* Switch the tokeniser's content model flag to the RCDATA state. */
1494: return HTML5::RCDATA;
1495:
1496: /* A start tag with the tag name "style" */
1497: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'style') {
1498: /* Create an element for the token and append the new element to the
1499: node pointed to by the head element pointer, or, if that is null
1500: (innerHTML case), to the current node. */
1501: if($this->head_pointer !== null) {
1502: $element = $this->insertElement($token, false);
1503: $this->head_pointer->appendChild($element);
1504:
1505: } else {
1506: $this->insertElement($token);
1507: }
1508:
1509: /* Switch the tokeniser's content model flag to the CDATA state. */
1510: return HTML5::CDATA;
1511:
1512: /* A start tag with the tag name "script" */
1513: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'script') {
1514: /* Create an element for the token. */
1515: $element = $this->insertElement($token, false);
1516: $this->head_pointer->appendChild($element);
1517:
1518: /* Switch the tokeniser's content model flag to the CDATA state. */
1519: return HTML5::CDATA;
1520:
1521: /* A start tag with the tag name "base", "link", or "meta" */
1522: } elseif($token['type'] === HTML5::STARTTAG && in_array($token['name'],
1523: array('base', 'link', 'meta'))) {
1524: /* Create an element for the token and append the new element to the
1525: node pointed to by the head element pointer, or, if that is null
1526: (innerHTML case), to the current node. */
1527: if($this->head_pointer !== null) {
1528: $element = $this->insertElement($token, false);
1529: $this->head_pointer->appendChild($element);
1530: array_pop($this->stack);
1531:
1532: } else {
1533: $this->insertElement($token);
1534: }
1535:
1536: /* An end tag with the tag name "head" */
1537: } elseif($token['type'] === HTML5::ENDTAG && $token['name'] === 'head') {
1538: /* If the current node is a head element, pop the current node off
1539: the stack of open elements. */
1540: if($this->head_pointer->isSameNode(end($this->stack))) {
1541: array_pop($this->stack);
1542:
1543: /* Otherwise, this is a parse error. */
1544: } else {
1545: // k
1546: }
1547:
1548: /* Change the insertion mode to "after head". */
1549: $this->mode = self::AFTER_HEAD;
1550:
1551: /* A start tag with the tag name "head" or an end tag except "html". */
1552: } elseif(($token['type'] === HTML5::STARTTAG && $token['name'] === 'head') ||
1553: ($token['type'] === HTML5::ENDTAG && $token['name'] !== 'html')) {
1554: // Parse error. Ignore the token.
1555:
1556: /* Anything else */
1557: } else {
1558: /* If the current node is a head element, act as if an end tag
1559: token with the tag name "head" had been seen. */
1560: if($this->head_pointer->isSameNode(end($this->stack))) {
1561: $this->inHead(array(
1562: 'name' => 'head',
1563: 'type' => HTML5::ENDTAG
1564: ));
1565:
1566: /* Otherwise, change the insertion mode to "after head". */
1567: } else {
1568: $this->mode = self::AFTER_HEAD;
1569: }
1570:
1571: /* Then, reprocess the current token. */
1572: return $this->afterHead($token);
1573: }
1574: }
1575:
1576: private function afterHead($token) {
1577: /* Handle the token as follows: */
1578:
1579: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
1580: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
1581: or U+0020 SPACE */
1582: if($token['type'] === HTML5::CHARACTR &&
1583: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
1584: /* Append the character to the current node. */
1585: $this->insertText($token['data']);
1586:
1587: /* A comment token */
1588: } elseif($token['type'] === HTML5::COMMENT) {
1589: /* Append a Comment node to the current node with the data attribute
1590: set to the data given in the comment token. */
1591: $this->insertComment($token['data']);
1592:
1593: /* A start tag token with the tag name "body" */
1594: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'body') {
1595: /* Insert a body element for the token. */
1596: $this->insertElement($token);
1597:
1598: /* Change the insertion mode to "in body". */
1599: $this->mode = self::IN_BODY;
1600:
1601: /* A start tag token with the tag name "frameset" */
1602: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'frameset') {
1603: /* Insert a frameset element for the token. */
1604: $this->insertElement($token);
1605:
1606: /* Change the insertion mode to "in frameset". */
1607: $this->mode = self::IN_FRAME;
1608:
1609: /* A start tag token whose tag name is one of: "base", "link", "meta",
1610: "script", "style", "title" */
1611: } elseif($token['type'] === HTML5::STARTTAG && in_array($token['name'],
1612: array('base', 'link', 'meta', 'script', 'style', 'title'))) {
1613: /* Parse error. Switch the insertion mode back to "in head" and
1614: reprocess the token. */
1615: $this->mode = self::IN_HEAD;
1616: return $this->inHead($token);
1617:
1618: /* Anything else */
1619: } else {
1620: /* Act as if a start tag token with the tag name "body" and no
1621: attributes had been seen, and then reprocess the current token. */
1622: $this->afterHead(array(
1623: 'name' => 'body',
1624: 'type' => HTML5::STARTTAG,
1625: 'attr' => array()
1626: ));
1627:
1628: return $this->inBody($token);
1629: }
1630: }
1631:
1632: private function inBody($token) {
1633: /* Handle the token as follows: */
1634:
1635: switch($token['type']) {
1636: /* A character token */
1637: case HTML5::CHARACTR:
1638: /* Reconstruct the active formatting elements, if any. */
1639: $this->reconstructActiveFormattingElements();
1640:
1641: /* Append the token's character to the current node. */
1642: $this->insertText($token['data']);
1643: break;
1644:
1645: /* A comment token */
1646: case HTML5::COMMENT:
1647: /* Append a Comment node to the current node with the data
1648: attribute set to the data given in the comment token. */
1649: $this->insertComment($token['data']);
1650: break;
1651:
1652: case HTML5::STARTTAG:
1653: switch($token['name']) {
1654: /* A start tag token whose tag name is one of: "script",
1655: "style" */
1656: case 'script': case 'style':
1657: /* Process the token as if the insertion mode had been "in
1658: head". */
1659: return $this->inHead($token);
1660: break;
1661:
1662: /* A start tag token whose tag name is one of: "base", "link",
1663: "meta", "title" */
1664: case 'base': case 'link': case 'meta': case 'title':
1665: /* Parse error. Process the token as if the insertion mode
1666: had been "in head". */
1667: return $this->inHead($token);
1668: break;
1669:
1670: /* A start tag token with the tag name "body" */
1671: case 'body':
1672: /* Parse error. If the second element on the stack of open
1673: elements is not a body element, or, if the stack of open
1674: elements has only one node on it, then ignore the token.
1675: (innerHTML case) */
1676: if(count($this->stack) === 1 || $this->stack[1]->nodeName !== 'body') {
1677: // Ignore
1678:
1679: /* Otherwise, for each attribute on the token, check to see
1680: if the attribute is already present on the body element (the
1681: second element) on the stack of open elements. If it is not,
1682: add the attribute and its corresponding value to that
1683: element. */
1684: } else {
1685: foreach($token['attr'] as $attr) {
1686: if(!$this->stack[1]->hasAttribute($attr['name'])) {
1687: $this->stack[1]->setAttribute($attr['name'], $attr['value']);
1688: }
1689: }
1690: }
1691: break;
1692:
1693: /* A start tag whose tag name is one of: "address",
1694: "blockquote", "center", "dir", "div", "dl", "fieldset",
1695: "listing", "menu", "ol", "p", "ul" */
1696: case 'address': case 'blockquote': case 'center': case 'dir':
1697: case 'div': case 'dl': case 'fieldset': case 'listing':
1698: case 'menu': case 'ol': case 'p': case 'ul':
1699: /* If the stack of open elements has a p element in scope,
1700: then act as if an end tag with the tag name p had been
1701: seen. */
1702: if($this->elementInScope('p')) {
1703: $this->emitToken(array(
1704: 'name' => 'p',
1705: 'type' => HTML5::ENDTAG
1706: ));
1707: }
1708:
1709: /* Insert an HTML element for the token. */
1710: $this->insertElement($token);
1711: break;
1712:
1713: /* A start tag whose tag name is "form" */
1714: case 'form':
1715: /* If the form element pointer is not null, ignore the
1716: token with a parse error. */
1717: if($this->form_pointer !== null) {
1718: // Ignore.
1719:
1720: /* Otherwise: */
1721: } else {
1722: /* If the stack of open elements has a p element in
1723: scope, then act as if an end tag with the tag name p
1724: had been seen. */
1725: if($this->elementInScope('p')) {
1726: $this->emitToken(array(
1727: 'name' => 'p',
1728: 'type' => HTML5::ENDTAG
1729: ));
1730: }
1731:
1732: /* Insert an HTML element for the token, and set the
1733: form element pointer to point to the element created. */
1734: $element = $this->insertElement($token);
1735: $this->form_pointer = $element;
1736: }
1737: break;
1738:
1739: /* A start tag whose tag name is "li", "dd" or "dt" */
1740: case 'li': case 'dd': case 'dt':
1741: /* If the stack of open elements has a p element in scope,
1742: then act as if an end tag with the tag name p had been
1743: seen. */
1744: if($this->elementInScope('p')) {
1745: $this->emitToken(array(
1746: 'name' => 'p',
1747: 'type' => HTML5::ENDTAG
1748: ));
1749: }
1750:
1751: $stack_length = count($this->stack) - 1;
1752:
1753: for($n = $stack_length; 0 <= $n; $n--) {
1754: /* 1. Initialise node to be the current node (the
1755: bottommost node of the stack). */
1756: $stop = false;
1757: $node = $this->stack[$n];
1758: $cat = $this->getElementCategory($node->tagName);
1759:
1760: /* 2. If node is an li, dd or dt element, then pop all
1761: the nodes from the current node up to node, including
1762: node, then stop this algorithm. */
1763: if($token['name'] === $node->tagName || ($token['name'] !== 'li'
1764: && ($node->tagName === 'dd' || $node->tagName === 'dt'))) {
1765: for($x = $stack_length; $x >= $n ; $x--) {
1766: array_pop($this->stack);
1767: }
1768:
1769: break;
1770: }
1771:
1772: /* 3. If node is not in the formatting category, and is
1773: not in the phrasing category, and is not an address or
1774: div element, then stop this algorithm. */
1775: if($cat !== self::FORMATTING && $cat !== self::PHRASING &&
1776: $node->tagName !== 'address' && $node->tagName !== 'div') {
1777: break;
1778: }
1779: }
1780:
1781: /* Finally, insert an HTML element with the same tag
1782: name as the token's. */
1783: $this->insertElement($token);
1784: break;
1785:
1786: /* A start tag token whose tag name is "plaintext" */
1787: case 'plaintext':
1788: /* If the stack of open elements has a p element in scope,
1789: then act as if an end tag with the tag name p had been
1790: seen. */
1791: if($this->elementInScope('p')) {
1792: $this->emitToken(array(
1793: 'name' => 'p',
1794: 'type' => HTML5::ENDTAG
1795: ));
1796: }
1797:
1798: /* Insert an HTML element for the token. */
1799: $this->insertElement($token);
1800:
1801: return HTML5::PLAINTEXT;
1802: break;
1803:
1804: /* A start tag whose tag name is one of: "h1", "h2", "h3", "h4",
1805: "h5", "h6" */
1806: case 'h1': case 'h2': case 'h3': case 'h4': case 'h5': case 'h6':
1807: /* If the stack of open elements has a p element in scope,
1808: then act as if an end tag with the tag name p had been seen. */
1809: if($this->elementInScope('p')) {
1810: $this->emitToken(array(
1811: 'name' => 'p',
1812: 'type' => HTML5::ENDTAG
1813: ));
1814: }
1815:
1816: /* If the stack of open elements has in scope an element whose
1817: tag name is one of "h1", "h2", "h3", "h4", "h5", or "h6", then
1818: this is a parse error; pop elements from the stack until an
1819: element with one of those tag names has been popped from the
1820: stack. */
1821: while($this->elementInScope(array('h1', 'h2', 'h3', 'h4', 'h5', 'h6'))) {
1822: array_pop($this->stack);
1823: }
1824:
1825: /* Insert an HTML element for the token. */
1826: $this->insertElement($token);
1827: break;
1828:
1829: /* A start tag whose tag name is "a" */
1830: case 'a':
1831: /* If the list of active formatting elements contains
1832: an element whose tag name is "a" between the end of the
1833: list and the last marker on the list (or the start of
1834: the list if there is no marker on the list), then this
1835: is a parse error; act as if an end tag with the tag name
1836: "a" had been seen, then remove that element from the list
1837: of active formatting elements and the stack of open
1838: elements if the end tag didn't already remove it (it
1839: might not have if the element is not in table scope). */
1840: $leng = count($this->a_formatting);
1841:
1842: for($n = $leng - 1; $n >= 0; $n--) {
1843: if($this->a_formatting[$n] === self::MARKER) {
1844: break;
1845:
1846: } elseif($this->a_formatting[$n]->nodeName === 'a') {
1847: $this->emitToken(array(
1848: 'name' => 'a',
1849: 'type' => HTML5::ENDTAG
1850: ));
1851: break;
1852: }
1853: }
1854:
1855: /* Reconstruct the active formatting elements, if any. */
1856: $this->reconstructActiveFormattingElements();
1857:
1858: /* Insert an HTML element for the token. */
1859: $el = $this->insertElement($token);
1860:
1861: /* Add that element to the list of active formatting
1862: elements. */
1863: $this->a_formatting[] = $el;
1864: break;
1865:
1866: /* A start tag whose tag name is one of: "b", "big", "em", "font",
1867: "i", "nobr", "s", "small", "strike", "strong", "tt", "u" */
1868: case 'b': case 'big': case 'em': case 'font': case 'i':
1869: case 'nobr': case 's': case 'small': case 'strike':
1870: case 'strong': case 'tt': case 'u':
1871: /* Reconstruct the active formatting elements, if any. */
1872: $this->reconstructActiveFormattingElements();
1873:
1874: /* Insert an HTML element for the token. */
1875: $el = $this->insertElement($token);
1876:
1877: /* Add that element to the list of active formatting
1878: elements. */
1879: $this->a_formatting[] = $el;
1880: break;
1881:
1882: /* A start tag token whose tag name is "button" */
1883: case 'button':
1884: /* If the stack of open elements has a button element in scope,
1885: then this is a parse error; act as if an end tag with the tag
1886: name "button" had been seen, then reprocess the token. (We don't
1887: do that. Unnecessary.) */
1888: if($this->elementInScope('button')) {
1889: $this->inBody(array(
1890: 'name' => 'button',
1891: 'type' => HTML5::ENDTAG
1892: ));
1893: }
1894:
1895: /* Reconstruct the active formatting elements, if any. */
1896: $this->reconstructActiveFormattingElements();
1897:
1898: /* Insert an HTML element for the token. */
1899: $this->insertElement($token);
1900:
1901: /* Insert a marker at the end of the list of active
1902: formatting elements. */
1903: $this->a_formatting[] = self::MARKER;
1904: break;
1905:
1906: /* A start tag token whose tag name is one of: "marquee", "object" */
1907: case 'marquee': case 'object':
1908: /* Reconstruct the active formatting elements, if any. */
1909: $this->reconstructActiveFormattingElements();
1910:
1911: /* Insert an HTML element for the token. */
1912: $this->insertElement($token);
1913:
1914: /* Insert a marker at the end of the list of active
1915: formatting elements. */
1916: $this->a_formatting[] = self::MARKER;
1917: break;
1918:
1919: /* A start tag token whose tag name is "xmp" */
1920: case 'xmp':
1921: /* Reconstruct the active formatting elements, if any. */
1922: $this->reconstructActiveFormattingElements();
1923:
1924: /* Insert an HTML element for the token. */
1925: $this->insertElement($token);
1926:
1927: /* Switch the content model flag to the CDATA state. */
1928: return HTML5::CDATA;
1929: break;
1930:
1931: /* A start tag whose tag name is "table" */
1932: case 'table':
1933: /* If the stack of open elements has a p element in scope,
1934: then act as if an end tag with the tag name p had been seen. */
1935: if($this->elementInScope('p')) {
1936: $this->emitToken(array(
1937: 'name' => 'p',
1938: 'type' => HTML5::ENDTAG
1939: ));
1940: }
1941:
1942: /* Insert an HTML element for the token. */
1943: $this->insertElement($token);
1944:
1945: /* Change the insertion mode to "in table". */
1946: $this->mode = self::IN_TABLE;
1947: break;
1948:
1949: /* A start tag whose tag name is one of: "area", "basefont",
1950: "bgsound", "br", "embed", "img", "param", "spacer", "wbr" */
1951: case 'area': case 'basefont': case 'bgsound': case 'br':
1952: case 'embed': case 'img': case 'param': case 'spacer':
1953: case 'wbr':
1954: /* Reconstruct the active formatting elements, if any. */
1955: $this->reconstructActiveFormattingElements();
1956:
1957: /* Insert an HTML element for the token. */
1958: $this->insertElement($token);
1959:
1960: /* Immediately pop the current node off the stack of open elements. */
1961: array_pop($this->stack);
1962: break;
1963:
1964: /* A start tag whose tag name is "hr" */
1965: case 'hr':
1966: /* If the stack of open elements has a p element in scope,
1967: then act as if an end tag with the tag name p had been seen. */
1968: if($this->elementInScope('p')) {
1969: $this->emitToken(array(
1970: 'name' => 'p',
1971: 'type' => HTML5::ENDTAG
1972: ));
1973: }
1974:
1975: /* Insert an HTML element for the token. */
1976: $this->insertElement($token);
1977:
1978: /* Immediately pop the current node off the stack of open elements. */
1979: array_pop($this->stack);
1980: break;
1981:
1982: /* A start tag whose tag name is "image" */
1983: case 'image':
1984: /* Parse error. Change the token's tag name to "img" and
1985: reprocess it. (Don't ask.) */
1986: $token['name'] = 'img';
1987: return $this->inBody($token);
1988: break;
1989:
1990: /* A start tag whose tag name is "input" */
1991: case 'input':
1992: /* Reconstruct the active formatting elements, if any. */
1993: $this->reconstructActiveFormattingElements();
1994:
1995: /* Insert an input element for the token. */
1996: $element = $this->insertElement($token, false);
1997:
1998: /* If the form element pointer is not null, then associate the
1999: input element with the form element pointed to by the form
2000: element pointer. */
2001: $this->form_pointer !== null
2002: ? $this->form_pointer->appendChild($element)
2003: : end($this->stack)->appendChild($element);
2004:
2005: /* Pop that input element off the stack of open elements. */
2006: array_pop($this->stack);
2007: break;
2008:
2009: /* A start tag whose tag name is "isindex" */
2010: case 'isindex':
2011: /* Parse error. */
2012: // w/e
2013:
2014: /* If the form element pointer is not null,
2015: then ignore the token. */
2016: if($this->form_pointer === null) {
2017: /* Act as if a start tag token with the tag name "form" had
2018: been seen. */
2019: $this->inBody(array(
2020: 'name' => 'body',
2021: 'type' => HTML5::STARTTAG,
2022: 'attr' => array()
2023: ));
2024:
2025: /* Act as if a start tag token with the tag name "hr" had
2026: been seen. */
2027: $this->inBody(array(
2028: 'name' => 'hr',
2029: 'type' => HTML5::STARTTAG,
2030: 'attr' => array()
2031: ));
2032:
2033: /* Act as if a start tag token with the tag name "p" had
2034: been seen. */
2035: $this->inBody(array(
2036: 'name' => 'p',
2037: 'type' => HTML5::STARTTAG,
2038: 'attr' => array()
2039: ));
2040:
2041: /* Act as if a start tag token with the tag name "label"
2042: had been seen. */
2043: $this->inBody(array(
2044: 'name' => 'label',
2045: 'type' => HTML5::STARTTAG,
2046: 'attr' => array()
2047: ));
2048:
2049: /* Act as if a stream of character tokens had been seen. */
2050: $this->insertText('This is a searchable index. '.
2051: 'Insert your search keywords here: ');
2052:
2053: /* Act as if a start tag token with the tag name "input"
2054: had been seen, with all the attributes from the "isindex"
2055: token, except with the "name" attribute set to the value
2056: "isindex" (ignoring any explicit "name" attribute). */
2057: $attr = $token['attr'];
2058: $attr[] = array('name' => 'name', 'value' => 'isindex');
2059:
2060: $this->inBody(array(
2061: 'name' => 'input',
2062: 'type' => HTML5::STARTTAG,
2063: 'attr' => $attr
2064: ));
2065:
2066: /* Act as if a stream of character tokens had been seen
2067: (see below for what they should say). */
2068: $this->insertText('This is a searchable index. '.
2069: 'Insert your search keywords here: ');
2070:
2071: /* Act as if an end tag token with the tag name "label"
2072: had been seen. */
2073: $this->inBody(array(
2074: 'name' => 'label',
2075: 'type' => HTML5::ENDTAG
2076: ));
2077:
2078: /* Act as if an end tag token with the tag name "p" had
2079: been seen. */
2080: $this->inBody(array(
2081: 'name' => 'p',
2082: 'type' => HTML5::ENDTAG
2083: ));
2084:
2085: /* Act as if a start tag token with the tag name "hr" had
2086: been seen. */
2087: $this->inBody(array(
2088: 'name' => 'hr',
2089: 'type' => HTML5::ENDTAG
2090: ));
2091:
2092: /* Act as if an end tag token with the tag name "form" had
2093: been seen. */
2094: $this->inBody(array(
2095: 'name' => 'form',
2096: 'type' => HTML5::ENDTAG
2097: ));
2098: }
2099: break;
2100:
2101: /* A start tag whose tag name is "textarea" */
2102: case 'textarea':
2103: $this->insertElement($token);
2104:
2105: /* Switch the tokeniser's content model flag to the
2106: RCDATA state. */
2107: return HTML5::RCDATA;
2108: break;
2109:
2110: /* A start tag whose tag name is one of: "iframe", "noembed",
2111: "noframes" */
2112: case 'iframe': case 'noembed': case 'noframes':
2113: $this->insertElement($token);
2114:
2115: /* Switch the tokeniser's content model flag to the CDATA state. */
2116: return HTML5::CDATA;
2117: break;
2118:
2119: /* A start tag whose tag name is "select" */
2120: case 'select':
2121: /* Reconstruct the active formatting elements, if any. */
2122: $this->reconstructActiveFormattingElements();
2123:
2124: /* Insert an HTML element for the token. */
2125: $this->insertElement($token);
2126:
2127: /* Change the insertion mode to "in select". */
2128: $this->mode = self::IN_SELECT;
2129: break;
2130:
2131: /* A start or end tag whose tag name is one of: "caption", "col",
2132: "colgroup", "frame", "frameset", "head", "option", "optgroup",
2133: "tbody", "td", "tfoot", "th", "thead", "tr". */
2134: case 'caption': case 'col': case 'colgroup': case 'frame':
2135: case 'frameset': case 'head': case 'option': case 'optgroup':
2136: case 'tbody': case 'td': case 'tfoot': case 'th': case 'thead':
2137: case 'tr':
2138: // Parse error. Ignore the token.
2139: break;
2140:
2141: /* A start or end tag whose tag name is one of: "event-source",
2142: "section", "nav", "article", "aside", "header", "footer",
2143: "datagrid", "command" */
2144: case 'event-source': case 'section': case 'nav': case 'article':
2145: case 'aside': case 'header': case 'footer': case 'datagrid':
2146: case 'command':
2147: // Work in progress!
2148: break;
2149:
2150: /* A start tag token not covered by the previous entries */
2151: default:
2152: /* Reconstruct the active formatting elements, if any. */
2153: $this->reconstructActiveFormattingElements();
2154:
2155: $this->insertElement($token, true, true);
2156: break;
2157: }
2158: break;
2159:
2160: case HTML5::ENDTAG:
2161: switch($token['name']) {
2162: /* An end tag with the tag name "body" */
2163: case 'body':
2164: /* If the second element in the stack of open elements is
2165: not a body element, this is a parse error. Ignore the token.
2166: (innerHTML case) */
2167: if(count($this->stack) < 2 || $this->stack[1]->nodeName !== 'body') {
2168: // Ignore.
2169:
2170: /* If the current node is not the body element, then this
2171: is a parse error. */
2172: } elseif(end($this->stack)->nodeName !== 'body') {
2173: // Parse error.
2174: }
2175:
2176: /* Change the insertion mode to "after body". */
2177: $this->mode = self::AFTER_BODY;
2178: break;
2179:
2180: /* An end tag with the tag name "html" */
2181: case 'html':
2182: /* Act as if an end tag with tag name "body" had been seen,
2183: then, if that token wasn't ignored, reprocess the current
2184: token. */
2185: $this->inBody(array(
2186: 'name' => 'body',
2187: 'type' => HTML5::ENDTAG
2188: ));
2189:
2190: return $this->afterBody($token);
2191: break;
2192:
2193: /* An end tag whose tag name is one of: "address", "blockquote",
2194: "center", "dir", "div", "dl", "fieldset", "listing", "menu",
2195: "ol", "pre", "ul" */
2196: case 'address': case 'blockquote': case 'center': case 'dir':
2197: case 'div': case 'dl': case 'fieldset': case 'listing':
2198: case 'menu': case 'ol': case 'pre': case 'ul':
2199: /* If the stack of open elements has an element in scope
2200: with the same tag name as that of the token, then generate
2201: implied end tags. */
2202: if($this->elementInScope($token['name'])) {
2203: $this->generateImpliedEndTags();
2204:
2205: /* Now, if the current node is not an element with
2206: the same tag name as that of the token, then this
2207: is a parse error. */
2208: // w/e
2209:
2210: /* If the stack of open elements has an element in
2211: scope with the same tag name as that of the token,
2212: then pop elements from this stack until an element
2213: with that tag name has been popped from the stack. */
2214: for($n = count($this->stack) - 1; $n >= 0; $n--) {
2215: if($this->stack[$n]->nodeName === $token['name']) {
2216: $n = -1;
2217: }
2218:
2219: array_pop($this->stack);
2220: }
2221: }
2222: break;
2223:
2224: /* An end tag whose tag name is "form" */
2225: case 'form':
2226: /* If the stack of open elements has an element in scope
2227: with the same tag name as that of the token, then generate
2228: implied end tags. */
2229: if($this->elementInScope($token['name'])) {
2230: $this->generateImpliedEndTags();
2231:
2232: }
2233:
2234: if(end($this->stack)->nodeName !== $token['name']) {
2235: /* Now, if the current node is not an element with the
2236: same tag name as that of the token, then this is a parse
2237: error. */
2238: // w/e
2239:
2240: } else {
2241: /* Otherwise, if the current node is an element with
2242: the same tag name as that of the token pop that element
2243: from the stack. */
2244: array_pop($this->stack);
2245: }
2246:
2247: /* In any case, set the form element pointer to null. */
2248: $this->form_pointer = null;
2249: break;
2250:
2251: /* An end tag whose tag name is "p" */
2252: case 'p':
2253: /* If the stack of open elements has a p element in scope,
2254: then generate implied end tags, except for p elements. */
2255: if($this->elementInScope('p')) {
2256: $this->generateImpliedEndTags(array('p'));
2257:
2258: /* If the current node is not a p element, then this is
2259: a parse error. */
2260: // k
2261:
2262: /* If the stack of open elements has a p element in
2263: scope, then pop elements from this stack until the stack
2264: no longer has a p element in scope. */
2265: for($n = count($this->stack) - 1; $n >= 0; $n--) {
2266: if($this->elementInScope('p')) {
2267: array_pop($this->stack);
2268:
2269: } else {
2270: break;
2271: }
2272: }
2273: }
2274: break;
2275:
2276: /* An end tag whose tag name is "dd", "dt", or "li" */
2277: case 'dd': case 'dt': case 'li':
2278: /* If the stack of open elements has an element in scope
2279: whose tag name matches the tag name of the token, then
2280: generate implied end tags, except for elements with the
2281: same tag name as the token. */
2282: if($this->elementInScope($token['name'])) {
2283: $this->generateImpliedEndTags(array($token['name']));
2284:
2285: /* If the current node is not an element with the same
2286: tag name as the token, then this is a parse error. */
2287: // w/e
2288:
2289: /* If the stack of open elements has an element in scope
2290: whose tag name matches the tag name of the token, then
2291: pop elements from this stack until an element with that
2292: tag name has been popped from the stack. */
2293: for($n = count($this->stack) - 1; $n >= 0; $n--) {
2294: if($this->stack[$n]->nodeName === $token['name']) {
2295: $n = -1;
2296: }
2297:
2298: array_pop($this->stack);
2299: }
2300: }
2301: break;
2302:
2303: /* An end tag whose tag name is one of: "h1", "h2", "h3", "h4",
2304: "h5", "h6" */
2305: case 'h1': case 'h2': case 'h3': case 'h4': case 'h5': case 'h6':
2306: $elements = array('h1', 'h2', 'h3', 'h4', 'h5', 'h6');
2307:
2308: /* If the stack of open elements has in scope an element whose
2309: tag name is one of "h1", "h2", "h3", "h4", "h5", or "h6", then
2310: generate implied end tags. */
2311: if($this->elementInScope($elements)) {
2312: $this->generateImpliedEndTags();
2313:
2314: /* Now, if the current node is not an element with the same
2315: tag name as that of the token, then this is a parse error. */
2316: // w/e
2317:
2318: /* If the stack of open elements has in scope an element
2319: whose tag name is one of "h1", "h2", "h3", "h4", "h5", or
2320: "h6", then pop elements from the stack until an element
2321: with one of those tag names has been popped from the stack. */
2322: while($this->elementInScope($elements)) {
2323: array_pop($this->stack);
2324: }
2325: }
2326: break;
2327:
2328: /* An end tag whose tag name is one of: "a", "b", "big", "em",
2329: "font", "i", "nobr", "s", "small", "strike", "strong", "tt", "u" */
2330: case 'a': case 'b': case 'big': case 'em': case 'font':
2331: case 'i': case 'nobr': case 's': case 'small': case 'strike':
2332: case 'strong': case 'tt': case 'u':
2333: /* 1. Let the formatting element be the last element in
2334: the list of active formatting elements that:
2335: * is between the end of the list and the last scope
2336: marker in the list, if any, or the start of the list
2337: otherwise, and
2338: * has the same tag name as the token.
2339: */
2340: while(true) {
2341: for($a = count($this->a_formatting) - 1; $a >= 0; $a--) {
2342: if($this->a_formatting[$a] === self::MARKER) {
2343: break;
2344:
2345: } elseif($this->a_formatting[$a]->tagName === $token['name']) {
2346: $formatting_element = $this->a_formatting[$a];
2347: $in_stack = in_array($formatting_element, $this->stack, true);
2348: $fe_af_pos = $a;
2349: break;
2350: }
2351: }
2352:
2353: /* If there is no such node, or, if that node is
2354: also in the stack of open elements but the element
2355: is not in scope, then this is a parse error. Abort
2356: these steps. The token is ignored. */
2357: if(!isset($formatting_element) || ($in_stack &&
2358: !$this->elementInScope($token['name']))) {
2359: break;
2360:
2361: /* Otherwise, if there is such a node, but that node
2362: is not in the stack of open elements, then this is a
2363: parse error; remove the element from the list, and
2364: abort these steps. */
2365: } elseif(isset($formatting_element) && !$in_stack) {
2366: unset($this->a_formatting[$fe_af_pos]);
2367: $this->a_formatting = array_merge($this->a_formatting);
2368: break;
2369: }
2370:
2371: /* 2. Let the furthest block be the topmost node in the
2372: stack of open elements that is lower in the stack
2373: than the formatting element, and is not an element in
2374: the phrasing or formatting categories. There might
2375: not be one. */
2376: $fe_s_pos = array_search($formatting_element, $this->stack, true);
2377: $length = count($this->stack);
2378:
2379: for($s = $fe_s_pos + 1; $s < $length; $s++) {
2380: $category = $this->getElementCategory($this->stack[$s]->nodeName);
2381:
2382: if($category !== self::PHRASING && $category !== self::FORMATTING) {
2383: $furthest_block = $this->stack[$s];
2384: }
2385: }
2386:
2387: /* 3. If there is no furthest block, then the UA must
2388: skip the subsequent steps and instead just pop all
2389: the nodes from the bottom of the stack of open
2390: elements, from the current node up to the formatting
2391: element, and remove the formatting element from the
2392: list of active formatting elements. */
2393: if(!isset($furthest_block)) {
2394: for($n = $length - 1; $n >= $fe_s_pos; $n--) {
2395: array_pop($this->stack);
2396: }
2397:
2398: unset($this->a_formatting[$fe_af_pos]);
2399: $this->a_formatting = array_merge($this->a_formatting);
2400: break;
2401: }
2402:
2403: /* 4. Let the common ancestor be the element
2404: immediately above the formatting element in the stack
2405: of open elements. */
2406: $common_ancestor = $this->stack[$fe_s_pos - 1];
2407:
2408: /* 5. If the furthest block has a parent node, then
2409: remove the furthest block from its parent node. */
2410: if($furthest_block->parentNode !== null) {
2411: $furthest_block->parentNode->removeChild($furthest_block);
2412: }
2413:
2414: /* 6. Let a bookmark note the position of the
2415: formatting element in the list of active formatting
2416: elements relative to the elements on either side
2417: of it in the list. */
2418: $bookmark = $fe_af_pos;
2419:
2420: /* 7. Let node and last node be the furthest block.
2421: Follow these steps: */
2422: $node = $furthest_block;
2423: $last_node = $furthest_block;
2424:
2425: while(true) {
2426: for($n = array_search($node, $this->stack, true) - 1; $n >= 0; $n--) {
2427: /* 7.1 Let node be the element immediately
2428: prior to node in the stack of open elements. */
2429: $node = $this->stack[$n];
2430:
2431: /* 7.2 If node is not in the list of active
2432: formatting elements, then remove node from
2433: the stack of open elements and then go back
2434: to step 1. */
2435: if(!in_array($node, $this->a_formatting, true)) {
2436: unset($this->stack[$n]);
2437: $this->stack = array_merge($this->stack);
2438:
2439: } else {
2440: break;
2441: }
2442: }
2443:
2444: /* 7.3 Otherwise, if node is the formatting
2445: element, then go to the next step in the overall
2446: algorithm. */
2447: if($node === $formatting_element) {
2448: break;
2449:
2450: /* 7.4 Otherwise, if last node is the furthest
2451: block, then move the aforementioned bookmark to
2452: be immediately after the node in the list of
2453: active formatting elements. */
2454: } elseif($last_node === $furthest_block) {
2455: $bookmark = array_search($node, $this->a_formatting, true) + 1;
2456: }
2457:
2458: /* 7.5 If node has any children, perform a
2459: shallow clone of node, replace the entry for
2460: node in the list of active formatting elements
2461: with an entry for the clone, replace the entry
2462: for node in the stack of open elements with an
2463: entry for the clone, and let node be the clone. */
2464: if($node->hasChildNodes()) {
2465: $clone = $node->cloneNode();
2466: $s_pos = array_search($node, $this->stack, true);
2467: $a_pos = array_search($node, $this->a_formatting, true);
2468:
2469: $this->stack[$s_pos] = $clone;
2470: $this->a_formatting[$a_pos] = $clone;
2471: $node = $clone;
2472: }
2473:
2474: /* 7.6 Insert last node into node, first removing
2475: it from its previous parent node if any. */
2476: if($last_node->parentNode !== null) {
2477: $last_node->parentNode->removeChild($last_node);
2478: }
2479:
2480: $node->appendChild($last_node);
2481:
2482: /* 7.7 Let last node be node. */
2483: $last_node = $node;
2484: }
2485:
2486: /* 8. Insert whatever last node ended up being in
2487: the previous step into the common ancestor node,
2488: first removing it from its previous parent node if
2489: any. */
2490: if($last_node->parentNode !== null) {
2491: $last_node->parentNode->removeChild($last_node);
2492: }
2493:
2494: $common_ancestor->appendChild($last_node);
2495:
2496: /* 9. Perform a shallow clone of the formatting
2497: element. */
2498: $clone = $formatting_element->cloneNode();
2499:
2500: /* 10. Take all of the child nodes of the furthest
2501: block and append them to the clone created in the
2502: last step. */
2503: while($furthest_block->hasChildNodes()) {
2504: $child = $furthest_block->firstChild;
2505: $furthest_block->removeChild($child);
2506: $clone->appendChild($child);
2507: }
2508:
2509: /* 11. Append that clone to the furthest block. */
2510: $furthest_block->appendChild($clone);
2511:
2512: /* 12. Remove the formatting element from the list
2513: of active formatting elements, and insert the clone
2514: into the list of active formatting elements at the
2515: position of the aforementioned bookmark. */
2516: $fe_af_pos = array_search($formatting_element, $this->a_formatting, true);
2517: unset($this->a_formatting[$fe_af_pos]);
2518: $this->a_formatting = array_merge($this->a_formatting);
2519:
2520: $af_part1 = array_slice($this->a_formatting, 0, $bookmark - 1);
2521: $af_part2 = array_slice($this->a_formatting, $bookmark, count($this->a_formatting));
2522: $this->a_formatting = array_merge($af_part1, array($clone), $af_part2);
2523:
2524: /* 13. Remove the formatting element from the stack
2525: of open elements, and insert the clone into the stack
2526: of open elements immediately after (i.e. in a more
2527: deeply nested position than) the position of the
2528: furthest block in that stack. */
2529: $fe_s_pos = array_search($formatting_element, $this->stack, true);
2530: $fb_s_pos = array_search($furthest_block, $this->stack, true);
2531: unset($this->stack[$fe_s_pos]);
2532:
2533: $s_part1 = array_slice($this->stack, 0, $fb_s_pos);
2534: $s_part2 = array_slice($this->stack, $fb_s_pos + 1, count($this->stack));
2535: $this->stack = array_merge($s_part1, array($clone), $s_part2);
2536:
2537: /* 14. Jump back to step 1 in this series of steps. */
2538: unset($formatting_element, $fe_af_pos, $fe_s_pos, $furthest_block);
2539: }
2540: break;
2541:
2542: /* An end tag token whose tag name is one of: "button",
2543: "marquee", "object" */
2544: case 'button': case 'marquee': case 'object':
2545: /* If the stack of open elements has an element in scope whose
2546: tag name matches the tag name of the token, then generate implied
2547: tags. */
2548: if($this->elementInScope($token['name'])) {
2549: $this->generateImpliedEndTags();
2550:
2551: /* Now, if the current node is not an element with the same
2552: tag name as the token, then this is a parse error. */
2553: // k
2554:
2555: /* Now, if the stack of open elements has an element in scope
2556: whose tag name matches the tag name of the token, then pop
2557: elements from the stack until that element has been popped from
2558: the stack, and clear the list of active formatting elements up
2559: to the last marker. */
2560: for($n = count($this->stack) - 1; $n >= 0; $n--) {
2561: if($this->stack[$n]->nodeName === $token['name']) {
2562: $n = -1;
2563: }
2564:
2565: array_pop($this->stack);
2566: }
2567:
2568: $marker = end(array_keys($this->a_formatting, self::MARKER, true));
2569:
2570: for($n = count($this->a_formatting) - 1; $n > $marker; $n--) {
2571: array_pop($this->a_formatting);
2572: }
2573: }
2574: break;
2575:
2576: /* Or an end tag whose tag name is one of: "area", "basefont",
2577: "bgsound", "br", "embed", "hr", "iframe", "image", "img",
2578: "input", "isindex", "noembed", "noframes", "param", "select",
2579: "spacer", "table", "textarea", "wbr" */
2580: case 'area': case 'basefont': case 'bgsound': case 'br':
2581: case 'embed': case 'hr': case 'iframe': case 'image':
2582: case 'img': case 'input': case 'isindex': case 'noembed':
2583: case 'noframes': case 'param': case 'select': case 'spacer':
2584: case 'table': case 'textarea': case 'wbr':
2585: // Parse error. Ignore the token.
2586: break;
2587:
2588: /* An end tag token not covered by the previous entries */
2589: default:
2590: for($n = count($this->stack) - 1; $n >= 0; $n--) {
2591: /* Initialise node to be the current node (the bottommost
2592: node of the stack). */
2593: $node = end($this->stack);
2594:
2595: /* If node has the same tag name as the end tag token,
2596: then: */
2597: if($token['name'] === $node->nodeName) {
2598: /* Generate implied end tags. */
2599: $this->generateImpliedEndTags();
2600:
2601: /* If the tag name of the end tag token does not
2602: match the tag name of the current node, this is a
2603: parse error. */
2604: // k
2605:
2606: /* Pop all the nodes from the current node up to
2607: node, including node, then stop this algorithm. */
2608: for($x = count($this->stack) - $n; $x >= $n; $x--) {
2609: array_pop($this->stack);
2610: }
2611:
2612: } else {
2613: $category = $this->getElementCategory($node);
2614:
2615: if($category !== self::SPECIAL && $category !== self::SCOPING) {
2616: /* Otherwise, if node is in neither the formatting
2617: category nor the phrasing category, then this is a
2618: parse error. Stop this algorithm. The end tag token
2619: is ignored. */
2620: return false;
2621: }
2622: }
2623: }
2624: break;
2625: }
2626: break;
2627: }
2628: }
2629:
2630: private function inTable($token) {
2631: $clear = array('html', 'table');
2632:
2633: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
2634: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
2635: or U+0020 SPACE */
2636: if($token['type'] === HTML5::CHARACTR &&
2637: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
2638: /* Append the character to the current node. */
2639: $text = $this->dom->createTextNode($token['data']);
2640: end($this->stack)->appendChild($text);
2641:
2642: /* A comment token */
2643: } elseif($token['type'] === HTML5::COMMENT) {
2644: /* Append a Comment node to the current node with the data
2645: attribute set to the data given in the comment token. */
2646: $comment = $this->dom->createComment($token['data']);
2647: end($this->stack)->appendChild($comment);
2648:
2649: /* A start tag whose tag name is "caption" */
2650: } elseif($token['type'] === HTML5::STARTTAG &&
2651: $token['name'] === 'caption') {
2652: /* Clear the stack back to a table context. */
2653: $this->clearStackToTableContext($clear);
2654:
2655: /* Insert a marker at the end of the list of active
2656: formatting elements. */
2657: $this->a_formatting[] = self::MARKER;
2658:
2659: /* Insert an HTML element for the token, then switch the
2660: insertion mode to "in caption". */
2661: $this->insertElement($token);
2662: $this->mode = self::IN_CAPTION;
2663:
2664: /* A start tag whose tag name is "colgroup" */
2665: } elseif($token['type'] === HTML5::STARTTAG &&
2666: $token['name'] === 'colgroup') {
2667: /* Clear the stack back to a table context. */
2668: $this->clearStackToTableContext($clear);
2669:
2670: /* Insert an HTML element for the token, then switch the
2671: insertion mode to "in column group". */
2672: $this->insertElement($token);
2673: $this->mode = self::IN_CGROUP;
2674:
2675: /* A start tag whose tag name is "col" */
2676: } elseif($token['type'] === HTML5::STARTTAG &&
2677: $token['name'] === 'col') {
2678: $this->inTable(array(
2679: 'name' => 'colgroup',
2680: 'type' => HTML5::STARTTAG,
2681: 'attr' => array()
2682: ));
2683:
2684: $this->inColumnGroup($token);
2685:
2686: /* A start tag whose tag name is one of: "tbody", "tfoot", "thead" */
2687: } elseif($token['type'] === HTML5::STARTTAG && in_array($token['name'],
2688: array('tbody', 'tfoot', 'thead'))) {
2689: /* Clear the stack back to a table context. */
2690: $this->clearStackToTableContext($clear);
2691:
2692: /* Insert an HTML element for the token, then switch the insertion
2693: mode to "in table body". */
2694: $this->insertElement($token);
2695: $this->mode = self::IN_TBODY;
2696:
2697: /* A start tag whose tag name is one of: "td", "th", "tr" */
2698: } elseif($token['type'] === HTML5::STARTTAG &&
2699: in_array($token['name'], array('td', 'th', 'tr'))) {
2700: /* Act as if a start tag token with the tag name "tbody" had been
2701: seen, then reprocess the current token. */
2702: $this->inTable(array(
2703: 'name' => 'tbody',
2704: 'type' => HTML5::STARTTAG,
2705: 'attr' => array()
2706: ));
2707:
2708: return $this->inTableBody($token);
2709:
2710: /* A start tag whose tag name is "table" */
2711: } elseif($token['type'] === HTML5::STARTTAG &&
2712: $token['name'] === 'table') {
2713: /* Parse error. Act as if an end tag token with the tag name "table"
2714: had been seen, then, if that token wasn't ignored, reprocess the
2715: current token. */
2716: $this->inTable(array(
2717: 'name' => 'table',
2718: 'type' => HTML5::ENDTAG
2719: ));
2720:
2721: return $this->mainPhase($token);
2722:
2723: /* An end tag whose tag name is "table" */
2724: } elseif($token['type'] === HTML5::ENDTAG &&
2725: $token['name'] === 'table') {
2726: /* If the stack of open elements does not have an element in table
2727: scope with the same tag name as the token, this is a parse error.
2728: Ignore the token. (innerHTML case) */
2729: if(!$this->elementInScope($token['name'], true)) {
2730: return false;
2731:
2732: /* Otherwise: */
2733: } else {
2734: /* Generate implied end tags. */
2735: $this->generateImpliedEndTags();
2736:
2737: /* Now, if the current node is not a table element, then this
2738: is a parse error. */
2739: // w/e
2740:
2741: /* Pop elements from this stack until a table element has been
2742: popped from the stack. */
2743: while(true) {
2744: $current = end($this->stack)->nodeName;
2745: array_pop($this->stack);
2746:
2747: if($current === 'table') {
2748: break;
2749: }
2750: }
2751:
2752: /* Reset the insertion mode appropriately. */
2753: $this->resetInsertionMode();
2754: }
2755:
2756: /* An end tag whose tag name is one of: "body", "caption", "col",
2757: "colgroup", "html", "tbody", "td", "tfoot", "th", "thead", "tr" */
2758: } elseif($token['type'] === HTML5::ENDTAG && in_array($token['name'],
2759: array('body', 'caption', 'col', 'colgroup', 'html', 'tbody', 'td',
2760: 'tfoot', 'th', 'thead', 'tr'))) {
2761: // Parse error. Ignore the token.
2762:
2763: /* Anything else */
2764: } else {
2765: /* Parse error. Process the token as if the insertion mode was "in
2766: body", with the following exception: */
2767:
2768: /* If the current node is a table, tbody, tfoot, thead, or tr
2769: element, then, whenever a node would be inserted into the current
2770: node, it must instead be inserted into the foster parent element. */
2771: if(in_array(end($this->stack)->nodeName,
2772: array('table', 'tbody', 'tfoot', 'thead', 'tr'))) {
2773: /* The foster parent element is the parent element of the last
2774: table element in the stack of open elements, if there is a
2775: table element and it has such a parent element. If there is no
2776: table element in the stack of open elements (innerHTML case),
2777: then the foster parent element is the first element in the
2778: stack of open elements (the html element). Otherwise, if there
2779: is a table element in the stack of open elements, but the last
2780: table element in the stack of open elements has no parent, or
2781: its parent node is not an element, then the foster parent
2782: element is the element before the last table element in the
2783: stack of open elements. */
2784: for($n = count($this->stack) - 1; $n >= 0; $n--) {
2785: if($this->stack[$n]->nodeName === 'table') {
2786: $table = $this->stack[$n];
2787: break;
2788: }
2789: }
2790:
2791: if(isset($table) && $table->parentNode !== null) {
2792: $this->foster_parent = $table->parentNode;
2793:
2794: } elseif(!isset($table)) {
2795: $this->foster_parent = $this->stack[0];
2796:
2797: } elseif(isset($table) && ($table->parentNode === null ||
2798: $table->parentNode->nodeType !== XML_ELEMENT_NODE)) {
2799: $this->foster_parent = $this->stack[$n - 1];
2800: }
2801: }
2802:
2803: $this->inBody($token);
2804: }
2805: }
2806:
2807: private function inCaption($token) {
2808: /* An end tag whose tag name is "caption" */
2809: if($token['type'] === HTML5::ENDTAG && $token['name'] === 'caption') {
2810: /* If the stack of open elements does not have an element in table
2811: scope with the same tag name as the token, this is a parse error.
2812: Ignore the token. (innerHTML case) */
2813: if(!$this->elementInScope($token['name'], true)) {
2814: // Ignore
2815:
2816: /* Otherwise: */
2817: } else {
2818: /* Generate implied end tags. */
2819: $this->generateImpliedEndTags();
2820:
2821: /* Now, if the current node is not a caption element, then this
2822: is a parse error. */
2823: // w/e
2824:
2825: /* Pop elements from this stack until a caption element has
2826: been popped from the stack. */
2827: while(true) {
2828: $node = end($this->stack)->nodeName;
2829: array_pop($this->stack);
2830:
2831: if($node === 'caption') {
2832: break;
2833: }
2834: }
2835:
2836: /* Clear the list of active formatting elements up to the last
2837: marker. */
2838: $this->clearTheActiveFormattingElementsUpToTheLastMarker();
2839:
2840: /* Switch the insertion mode to "in table". */
2841: $this->mode = self::IN_TABLE;
2842: }
2843:
2844: /* A start tag whose tag name is one of: "caption", "col", "colgroup",
2845: "tbody", "td", "tfoot", "th", "thead", "tr", or an end tag whose tag
2846: name is "table" */
2847: } elseif(($token['type'] === HTML5::STARTTAG && in_array($token['name'],
2848: array('caption', 'col', 'colgroup', 'tbody', 'td', 'tfoot', 'th',
2849: 'thead', 'tr'))) || ($token['type'] === HTML5::ENDTAG &&
2850: $token['name'] === 'table')) {
2851: /* Parse error. Act as if an end tag with the tag name "caption"
2852: had been seen, then, if that token wasn't ignored, reprocess the
2853: current token. */
2854: $this->inCaption(array(
2855: 'name' => 'caption',
2856: 'type' => HTML5::ENDTAG
2857: ));
2858:
2859: return $this->inTable($token);
2860:
2861: /* An end tag whose tag name is one of: "body", "col", "colgroup",
2862: "html", "tbody", "td", "tfoot", "th", "thead", "tr" */
2863: } elseif($token['type'] === HTML5::ENDTAG && in_array($token['name'],
2864: array('body', 'col', 'colgroup', 'html', 'tbody', 'tfoot', 'th',
2865: 'thead', 'tr'))) {
2866: // Parse error. Ignore the token.
2867:
2868: /* Anything else */
2869: } else {
2870: /* Process the token as if the insertion mode was "in body". */
2871: $this->inBody($token);
2872: }
2873: }
2874:
2875: private function inColumnGroup($token) {
2876: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
2877: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
2878: or U+0020 SPACE */
2879: if($token['type'] === HTML5::CHARACTR &&
2880: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
2881: /* Append the character to the current node. */
2882: $text = $this->dom->createTextNode($token['data']);
2883: end($this->stack)->appendChild($text);
2884:
2885: /* A comment token */
2886: } elseif($token['type'] === HTML5::COMMENT) {
2887: /* Append a Comment node to the current node with the data
2888: attribute set to the data given in the comment token. */
2889: $comment = $this->dom->createComment($token['data']);
2890: end($this->stack)->appendChild($comment);
2891:
2892: /* A start tag whose tag name is "col" */
2893: } elseif($token['type'] === HTML5::STARTTAG && $token['name'] === 'col') {
2894: /* Insert a col element for the token. Immediately pop the current
2895: node off the stack of open elements. */
2896: $this->insertElement($token);
2897: array_pop($this->stack);
2898:
2899: /* An end tag whose tag name is "colgroup" */
2900: } elseif($token['type'] === HTML5::ENDTAG &&
2901: $token['name'] === 'colgroup') {
2902: /* If the current node is the root html element, then this is a
2903: parse error, ignore the token. (innerHTML case) */
2904: if(end($this->stack)->nodeName === 'html') {
2905: // Ignore
2906:
2907: /* Otherwise, pop the current node (which will be a colgroup
2908: element) from the stack of open elements. Switch the insertion
2909: mode to "in table". */
2910: } else {
2911: array_pop($this->stack);
2912: $this->mode = self::IN_TABLE;
2913: }
2914:
2915: /* An end tag whose tag name is "col" */
2916: } elseif($token['type'] === HTML5::ENDTAG && $token['name'] === 'col') {
2917: /* Parse error. Ignore the token. */
2918:
2919: /* Anything else */
2920: } else {
2921: /* Act as if an end tag with the tag name "colgroup" had been seen,
2922: and then, if that token wasn't ignored, reprocess the current token. */
2923: $this->inColumnGroup(array(
2924: 'name' => 'colgroup',
2925: 'type' => HTML5::ENDTAG
2926: ));
2927:
2928: return $this->inTable($token);
2929: }
2930: }
2931:
2932: private function inTableBody($token) {
2933: $clear = array('tbody', 'tfoot', 'thead', 'html');
2934:
2935: /* A start tag whose tag name is "tr" */
2936: if($token['type'] === HTML5::STARTTAG && $token['name'] === 'tr') {
2937: /* Clear the stack back to a table body context. */
2938: $this->clearStackToTableContext($clear);
2939:
2940: /* Insert a tr element for the token, then switch the insertion
2941: mode to "in row". */
2942: $this->insertElement($token);
2943: $this->mode = self::IN_ROW;
2944:
2945: /* A start tag whose tag name is one of: "th", "td" */
2946: } elseif($token['type'] === HTML5::STARTTAG &&
2947: ($token['name'] === 'th' || $token['name'] === 'td')) {
2948: /* Parse error. Act as if a start tag with the tag name "tr" had
2949: been seen, then reprocess the current token. */
2950: $this->inTableBody(array(
2951: 'name' => 'tr',
2952: 'type' => HTML5::STARTTAG,
2953: 'attr' => array()
2954: ));
2955:
2956: return $this->inRow($token);
2957:
2958: /* An end tag whose tag name is one of: "tbody", "tfoot", "thead" */
2959: } elseif($token['type'] === HTML5::ENDTAG &&
2960: in_array($token['name'], array('tbody', 'tfoot', 'thead'))) {
2961: /* If the stack of open elements does not have an element in table
2962: scope with the same tag name as the token, this is a parse error.
2963: Ignore the token. */
2964: if(!$this->elementInScope($token['name'], true)) {
2965: // Ignore
2966:
2967: /* Otherwise: */
2968: } else {
2969: /* Clear the stack back to a table body context. */
2970: $this->clearStackToTableContext($clear);
2971:
2972: /* Pop the current node from the stack of open elements. Switch
2973: the insertion mode to "in table". */
2974: array_pop($this->stack);
2975: $this->mode = self::IN_TABLE;
2976: }
2977:
2978: /* A start tag whose tag name is one of: "caption", "col", "colgroup",
2979: "tbody", "tfoot", "thead", or an end tag whose tag name is "table" */
2980: } elseif(($token['type'] === HTML5::STARTTAG && in_array($token['name'],
2981: array('caption', 'col', 'colgroup', 'tbody', 'tfoor', 'thead'))) ||
2982: ($token['type'] === HTML5::STARTTAG && $token['name'] === 'table')) {
2983: /* If the stack of open elements does not have a tbody, thead, or
2984: tfoot element in table scope, this is a parse error. Ignore the
2985: token. (innerHTML case) */
2986: if(!$this->elementInScope(array('tbody', 'thead', 'tfoot'), true)) {
2987: // Ignore.
2988:
2989: /* Otherwise: */
2990: } else {
2991: /* Clear the stack back to a table body context. */
2992: $this->clearStackToTableContext($clear);
2993:
2994: /* Act as if an end tag with the same tag name as the current
2995: node ("tbody", "tfoot", or "thead") had been seen, then
2996: reprocess the current token. */
2997: $this->inTableBody(array(
2998: 'name' => end($this->stack)->nodeName,
2999: 'type' => HTML5::ENDTAG
3000: ));
3001:
3002: return $this->mainPhase($token);
3003: }
3004:
3005: /* An end tag whose tag name is one of: "body", "caption", "col",
3006: "colgroup", "html", "td", "th", "tr" */
3007: } elseif($token['type'] === HTML5::ENDTAG && in_array($token['name'],
3008: array('body', 'caption', 'col', 'colgroup', 'html', 'td', 'th', 'tr'))) {
3009: /* Parse error. Ignore the token. */
3010:
3011: /* Anything else */
3012: } else {
3013: /* Process the token as if the insertion mode was "in table". */
3014: $this->inTable($token);
3015: }
3016: }
3017:
3018: private function inRow($token) {
3019: $clear = array('tr', 'html');
3020:
3021: /* A start tag whose tag name is one of: "th", "td" */
3022: if($token['type'] === HTML5::STARTTAG &&
3023: ($token['name'] === 'th' || $token['name'] === 'td')) {
3024: /* Clear the stack back to a table row context. */
3025: $this->clearStackToTableContext($clear);
3026:
3027: /* Insert an HTML element for the token, then switch the insertion
3028: mode to "in cell". */
3029: $this->insertElement($token);
3030: $this->mode = self::IN_CELL;
3031:
3032: /* Insert a marker at the end of the list of active formatting
3033: elements. */
3034: $this->a_formatting[] = self::MARKER;
3035:
3036: /* An end tag whose tag name is "tr" */
3037: } elseif($token['type'] === HTML5::ENDTAG && $token['name'] === 'tr') {
3038: /* If the stack of open elements does not have an element in table
3039: scope with the same tag name as the token, this is a parse error.
3040: Ignore the token. (innerHTML case) */
3041: if(!$this->elementInScope($token['name'], true)) {
3042: // Ignore.
3043:
3044: /* Otherwise: */
3045: } else {
3046: /* Clear the stack back to a table row context. */
3047: $this->clearStackToTableContext($clear);
3048:
3049: /* Pop the current node (which will be a tr element) from the
3050: stack of open elements. Switch the insertion mode to "in table
3051: body". */
3052: array_pop($this->stack);
3053: $this->mode = self::IN_TBODY;
3054: }
3055:
3056: /* A start tag whose tag name is one of: "caption", "col", "colgroup",
3057: "tbody", "tfoot", "thead", "tr" or an end tag whose tag name is "table" */
3058: } elseif($token['type'] === HTML5::STARTTAG && in_array($token['name'],
3059: array('caption', 'col', 'colgroup', 'tbody', 'tfoot', 'thead', 'tr'))) {
3060: /* Act as if an end tag with the tag name "tr" had been seen, then,
3061: if that token wasn't ignored, reprocess the current token. */
3062: $this->inRow(array(
3063: 'name' => 'tr',
3064: 'type' => HTML5::ENDTAG
3065: ));
3066:
3067: return $this->inCell($token);
3068:
3069: /* An end tag whose tag name is one of: "tbody", "tfoot", "thead" */
3070: } elseif($token['type'] === HTML5::ENDTAG &&
3071: in_array($token['name'], array('tbody', 'tfoot', 'thead'))) {
3072: /* If the stack of open elements does not have an element in table
3073: scope with the same tag name as the token, this is a parse error.
3074: Ignore the token. */
3075: if(!$this->elementInScope($token['name'], true)) {
3076: // Ignore.
3077:
3078: /* Otherwise: */
3079: } else {
3080: /* Otherwise, act as if an end tag with the tag name "tr" had
3081: been seen, then reprocess the current token. */
3082: $this->inRow(array(
3083: 'name' => 'tr',
3084: 'type' => HTML5::ENDTAG
3085: ));
3086:
3087: return $this->inCell($token);
3088: }
3089:
3090: /* An end tag whose tag name is one of: "body", "caption", "col",
3091: "colgroup", "html", "td", "th" */
3092: } elseif($token['type'] === HTML5::ENDTAG && in_array($token['name'],
3093: array('body', 'caption', 'col', 'colgroup', 'html', 'td', 'th', 'tr'))) {
3094: /* Parse error. Ignore the token. */
3095:
3096: /* Anything else */
3097: } else {
3098: /* Process the token as if the insertion mode was "in table". */
3099: $this->inTable($token);
3100: }
3101: }
3102:
3103: private function inCell($token) {
3104: /* An end tag whose tag name is one of: "td", "th" */
3105: if($token['type'] === HTML5::ENDTAG &&
3106: ($token['name'] === 'td' || $token['name'] === 'th')) {
3107: /* If the stack of open elements does not have an element in table
3108: scope with the same tag name as that of the token, then this is a
3109: parse error and the token must be ignored. */
3110: if(!$this->elementInScope($token['name'], true)) {
3111: // Ignore.
3112:
3113: /* Otherwise: */
3114: } else {
3115: /* Generate implied end tags, except for elements with the same
3116: tag name as the token. */
3117: $this->generateImpliedEndTags(array($token['name']));
3118:
3119: /* Now, if the current node is not an element with the same tag
3120: name as the token, then this is a parse error. */
3121: // k
3122:
3123: /* Pop elements from this stack until an element with the same
3124: tag name as the token has been popped from the stack. */
3125: while(true) {
3126: $node = end($this->stack)->nodeName;
3127: array_pop($this->stack);
3128:
3129: if($node === $token['name']) {
3130: break;
3131: }
3132: }
3133:
3134: /* Clear the list of active formatting elements up to the last
3135: marker. */
3136: $this->clearTheActiveFormattingElementsUpToTheLastMarker();
3137:
3138: /* Switch the insertion mode to "in row". (The current node
3139: will be a tr element at this point.) */
3140: $this->mode = self::IN_ROW;
3141: }
3142:
3143: /* A start tag whose tag name is one of: "caption", "col", "colgroup",
3144: "tbody", "td", "tfoot", "th", "thead", "tr" */
3145: } elseif($token['type'] === HTML5::STARTTAG && in_array($token['name'],
3146: array('caption', 'col', 'colgroup', 'tbody', 'td', 'tfoot', 'th',
3147: 'thead', 'tr'))) {
3148: /* If the stack of open elements does not have a td or th element
3149: in table scope, then this is a parse error; ignore the token.
3150: (innerHTML case) */
3151: if(!$this->elementInScope(array('td', 'th'), true)) {
3152: // Ignore.
3153:
3154: /* Otherwise, close the cell (see below) and reprocess the current
3155: token. */
3156: } else {
3157: $this->closeCell();
3158: return $this->inRow($token);
3159: }
3160:
3161: /* A start tag whose tag name is one of: "caption", "col", "colgroup",
3162: "tbody", "td", "tfoot", "th", "thead", "tr" */
3163: } elseif($token['type'] === HTML5::STARTTAG && in_array($token['name'],
3164: array('caption', 'col', 'colgroup', 'tbody', 'td', 'tfoot', 'th',
3165: 'thead', 'tr'))) {
3166: /* If the stack of open elements does not have a td or th element
3167: in table scope, then this is a parse error; ignore the token.
3168: (innerHTML case) */
3169: if(!$this->elementInScope(array('td', 'th'), true)) {
3170: // Ignore.
3171:
3172: /* Otherwise, close the cell (see below) and reprocess the current
3173: token. */
3174: } else {
3175: $this->closeCell();
3176: return $this->inRow($token);
3177: }
3178:
3179: /* An end tag whose tag name is one of: "body", "caption", "col",
3180: "colgroup", "html" */
3181: } elseif($token['type'] === HTML5::ENDTAG && in_array($token['name'],
3182: array('body', 'caption', 'col', 'colgroup', 'html'))) {
3183: /* Parse error. Ignore the token. */
3184:
3185: /* An end tag whose tag name is one of: "table", "tbody", "tfoot",
3186: "thead", "tr" */
3187: } elseif($token['type'] === HTML5::ENDTAG && in_array($token['name'],
3188: array('table', 'tbody', 'tfoot', 'thead', 'tr'))) {
3189: /* If the stack of open elements does not have an element in table
3190: scope with the same tag name as that of the token (which can only
3191: happen for "tbody", "tfoot" and "thead", or, in the innerHTML case),
3192: then this is a parse error and the token must be ignored. */
3193: if(!$this->elementInScope($token['name'], true)) {
3194: // Ignore.
3195:
3196: /* Otherwise, close the cell (see below) and reprocess the current
3197: token. */
3198: } else {
3199: $this->closeCell();
3200: return $this->inRow($token);
3201: }
3202:
3203: /* Anything else */
3204: } else {
3205: /* Process the token as if the insertion mode was "in body". */
3206: $this->inBody($token);
3207: }
3208: }
3209:
3210: private function inSelect($token) {
3211: /* Handle the token as follows: */
3212:
3213: /* A character token */
3214: if($token['type'] === HTML5::CHARACTR) {
3215: /* Append the token's character to the current node. */
3216: $this->insertText($token['data']);
3217:
3218: /* A comment token */
3219: } elseif($token['type'] === HTML5::COMMENT) {
3220: /* Append a Comment node to the current node with the data
3221: attribute set to the data given in the comment token. */
3222: $this->insertComment($token['data']);
3223:
3224: /* A start tag token whose tag name is "option" */
3225: } elseif($token['type'] === HTML5::STARTTAG &&
3226: $token['name'] === 'option') {
3227: /* If the current node is an option element, act as if an end tag
3228: with the tag name "option" had been seen. */
3229: if(end($this->stack)->nodeName === 'option') {
3230: $this->inSelect(array(
3231: 'name' => 'option',
3232: 'type' => HTML5::ENDTAG
3233: ));
3234: }
3235:
3236: /* Insert an HTML element for the token. */
3237: $this->insertElement($token);
3238:
3239: /* A start tag token whose tag name is "optgroup" */
3240: } elseif($token['type'] === HTML5::STARTTAG &&
3241: $token['name'] === 'optgroup') {
3242: /* If the current node is an option element, act as if an end tag
3243: with the tag name "option" had been seen. */
3244: if(end($this->stack)->nodeName === 'option') {
3245: $this->inSelect(array(
3246: 'name' => 'option',
3247: 'type' => HTML5::ENDTAG
3248: ));
3249: }
3250:
3251: /* If the current node is an optgroup element, act as if an end tag
3252: with the tag name "optgroup" had been seen. */
3253: if(end($this->stack)->nodeName === 'optgroup') {
3254: $this->inSelect(array(
3255: 'name' => 'optgroup',
3256: 'type' => HTML5::ENDTAG
3257: ));
3258: }
3259:
3260: /* Insert an HTML element for the token. */
3261: $this->insertElement($token);
3262:
3263: /* An end tag token whose tag name is "optgroup" */
3264: } elseif($token['type'] === HTML5::ENDTAG &&
3265: $token['name'] === 'optgroup') {
3266: /* First, if the current node is an option element, and the node
3267: immediately before it in the stack of open elements is an optgroup
3268: element, then act as if an end tag with the tag name "option" had
3269: been seen. */
3270: $elements_in_stack = count($this->stack);
3271:
3272: if($this->stack[$elements_in_stack - 1]->nodeName === 'option' &&
3273: $this->stack[$elements_in_stack - 2]->nodeName === 'optgroup') {
3274: $this->inSelect(array(
3275: 'name' => 'option',
3276: 'type' => HTML5::ENDTAG
3277: ));
3278: }
3279:
3280: /* If the current node is an optgroup element, then pop that node
3281: from the stack of open elements. Otherwise, this is a parse error,
3282: ignore the token. */
3283: if($this->stack[$elements_in_stack - 1] === 'optgroup') {
3284: array_pop($this->stack);
3285: }
3286:
3287: /* An end tag token whose tag name is "option" */
3288: } elseif($token['type'] === HTML5::ENDTAG &&
3289: $token['name'] === 'option') {
3290: /* If the current node is an option element, then pop that node
3291: from the stack of open elements. Otherwise, this is a parse error,
3292: ignore the token. */
3293: if(end($this->stack)->nodeName === 'option') {
3294: array_pop($this->stack);
3295: }
3296:
3297: /* An end tag whose tag name is "select" */
3298: } elseif($token['type'] === HTML5::ENDTAG &&
3299: $token['name'] === 'select') {
3300: /* If the stack of open elements does not have an element in table
3301: scope with the same tag name as the token, this is a parse error.
3302: Ignore the token. (innerHTML case) */
3303: if(!$this->elementInScope($token['name'], true)) {
3304: // w/e
3305:
3306: /* Otherwise: */
3307: } else {
3308: /* Pop elements from the stack of open elements until a select
3309: element has been popped from the stack. */
3310: while(true) {
3311: $current = end($this->stack)->nodeName;
3312: array_pop($this->stack);
3313:
3314: if($current === 'select') {
3315: break;
3316: }
3317: }
3318:
3319: /* Reset the insertion mode appropriately. */
3320: $this->resetInsertionMode();
3321: }
3322:
3323: /* A start tag whose tag name is "select" */
3324: } elseif($token['name'] === 'select' &&
3325: $token['type'] === HTML5::STARTTAG) {
3326: /* Parse error. Act as if the token had been an end tag with the
3327: tag name "select" instead. */
3328: $this->inSelect(array(
3329: 'name' => 'select',
3330: 'type' => HTML5::ENDTAG
3331: ));
3332:
3333: /* An end tag whose tag name is one of: "caption", "table", "tbody",
3334: "tfoot", "thead", "tr", "td", "th" */
3335: } elseif(in_array($token['name'], array('caption', 'table', 'tbody',
3336: 'tfoot', 'thead', 'tr', 'td', 'th')) && $token['type'] === HTML5::ENDTAG) {
3337: /* Parse error. */
3338: // w/e
3339:
3340: /* If the stack of open elements has an element in table scope with
3341: the same tag name as that of the token, then act as if an end tag
3342: with the tag name "select" had been seen, and reprocess the token.
3343: Otherwise, ignore the token. */
3344: if($this->elementInScope($token['name'], true)) {
3345: $this->inSelect(array(
3346: 'name' => 'select',
3347: 'type' => HTML5::ENDTAG
3348: ));
3349:
3350: $this->mainPhase($token);
3351: }
3352:
3353: /* Anything else */
3354: } else {
3355: /* Parse error. Ignore the token. */
3356: }
3357: }
3358:
3359: private function afterBody($token) {
3360: /* Handle the token as follows: */
3361:
3362: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
3363: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
3364: or U+0020 SPACE */
3365: if($token['type'] === HTML5::CHARACTR &&
3366: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
3367: /* Process the token as it would be processed if the insertion mode
3368: was "in body". */
3369: $this->inBody($token);
3370:
3371: /* A comment token */
3372: } elseif($token['type'] === HTML5::COMMENT) {
3373: /* Append a Comment node to the first element in the stack of open
3374: elements (the html element), with the data attribute set to the
3375: data given in the comment token. */
3376: $comment = $this->dom->createComment($token['data']);
3377: $this->stack[0]->appendChild($comment);
3378:
3379: /* An end tag with the tag name "html" */
3380: } elseif($token['type'] === HTML5::ENDTAG && $token['name'] === 'html') {
3381: /* If the parser was originally created in order to handle the
3382: setting of an element's innerHTML attribute, this is a parse error;
3383: ignore the token. (The element will be an html element in this
3384: case.) (innerHTML case) */
3385:
3386: /* Otherwise, switch to the trailing end phase. */
3387: $this->phase = self::END_PHASE;
3388:
3389: /* Anything else */
3390: } else {
3391: /* Parse error. Set the insertion mode to "in body" and reprocess
3392: the token. */
3393: $this->mode = self::IN_BODY;
3394: return $this->inBody($token);
3395: }
3396: }
3397:
3398: private function inFrameset($token) {
3399: /* Handle the token as follows: */
3400:
3401: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
3402: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
3403: U+000D CARRIAGE RETURN (CR), or U+0020 SPACE */
3404: if($token['type'] === HTML5::CHARACTR &&
3405: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
3406: /* Append the character to the current node. */
3407: $this->insertText($token['data']);
3408:
3409: /* A comment token */
3410: } elseif($token['type'] === HTML5::COMMENT) {
3411: /* Append a Comment node to the current node with the data
3412: attribute set to the data given in the comment token. */
3413: $this->insertComment($token['data']);
3414:
3415: /* A start tag with the tag name "frameset" */
3416: } elseif($token['name'] === 'frameset' &&
3417: $token['type'] === HTML5::STARTTAG) {
3418: $this->insertElement($token);
3419:
3420: /* An end tag with the tag name "frameset" */
3421: } elseif($token['name'] === 'frameset' &&
3422: $token['type'] === HTML5::ENDTAG) {
3423: /* If the current node is the root html element, then this is a
3424: parse error; ignore the token. (innerHTML case) */
3425: if(end($this->stack)->nodeName === 'html') {
3426: // Ignore
3427:
3428: } else {
3429: /* Otherwise, pop the current node from the stack of open
3430: elements. */
3431: array_pop($this->stack);
3432:
3433: /* If the parser was not originally created in order to handle
3434: the setting of an element's innerHTML attribute (innerHTML case),
3435: and the current node is no longer a frameset element, then change
3436: the insertion mode to "after frameset". */
3437: $this->mode = self::AFTR_FRAME;
3438: }
3439:
3440: /* A start tag with the tag name "frame" */
3441: } elseif($token['name'] === 'frame' &&
3442: $token['type'] === HTML5::STARTTAG) {
3443: /* Insert an HTML element for the token. */
3444: $this->insertElement($token);
3445:
3446: /* Immediately pop the current node off the stack of open elements. */
3447: array_pop($this->stack);
3448:
3449: /* A start tag with the tag name "noframes" */
3450: } elseif($token['name'] === 'noframes' &&
3451: $token['type'] === HTML5::STARTTAG) {
3452: /* Process the token as if the insertion mode had been "in body". */
3453: $this->inBody($token);
3454:
3455: /* Anything else */
3456: } else {
3457: /* Parse error. Ignore the token. */
3458: }
3459: }
3460:
3461: private function afterFrameset($token) {
3462: /* Handle the token as follows: */
3463:
3464: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
3465: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
3466: U+000D CARRIAGE RETURN (CR), or U+0020 SPACE */
3467: if($token['type'] === HTML5::CHARACTR &&
3468: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
3469: /* Append the character to the current node. */
3470: $this->insertText($token['data']);
3471:
3472: /* A comment token */
3473: } elseif($token['type'] === HTML5::COMMENT) {
3474: /* Append a Comment node to the current node with the data
3475: attribute set to the data given in the comment token. */
3476: $this->insertComment($token['data']);
3477:
3478: /* An end tag with the tag name "html" */
3479: } elseif($token['name'] === 'html' &&
3480: $token['type'] === HTML5::ENDTAG) {
3481: /* Switch to the trailing end phase. */
3482: $this->phase = self::END_PHASE;
3483:
3484: /* A start tag with the tag name "noframes" */
3485: } elseif($token['name'] === 'noframes' &&
3486: $token['type'] === HTML5::STARTTAG) {
3487: /* Process the token as if the insertion mode had been "in body". */
3488: $this->inBody($token);
3489:
3490: /* Anything else */
3491: } else {
3492: /* Parse error. Ignore the token. */
3493: }
3494: }
3495:
3496: private function trailingEndPhase($token) {
3497: /* After the main phase, as each token is emitted from the tokenisation
3498: stage, it must be processed as described in this section. */
3499:
3500: /* A DOCTYPE token */
3501: if($token['type'] === HTML5::DOCTYPE) {
3502: // Parse error. Ignore the token.
3503:
3504: /* A comment token */
3505: } elseif($token['type'] === HTML5::COMMENT) {
3506: /* Append a Comment node to the Document object with the data
3507: attribute set to the data given in the comment token. */
3508: $comment = $this->dom->createComment($token['data']);
3509: $this->dom->appendChild($comment);
3510:
3511: /* A character token that is one of one of U+0009 CHARACTER TABULATION,
3512: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
3513: or U+0020 SPACE */
3514: } elseif($token['type'] === HTML5::CHARACTR &&
3515: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) {
3516: /* Process the token as it would be processed in the main phase. */
3517: $this->mainPhase($token);
3518:
3519: /* A character token that is not one of U+0009 CHARACTER TABULATION,
3520: U+000A LINE FEED (LF), U+000B LINE TABULATION, U+000C FORM FEED (FF),
3521: or U+0020 SPACE. Or a start tag token. Or an end tag token. */
3522: } elseif(($token['type'] === HTML5::CHARACTR &&
3523: preg_match('/^[\t\n\x0b\x0c ]+$/', $token['data'])) ||
3524: $token['type'] === HTML5::STARTTAG || $token['type'] === HTML5::ENDTAG) {
3525: /* Parse error. Switch back to the main phase and reprocess the
3526: token. */
3527: $this->phase = self::MAIN_PHASE;
3528: return $this->mainPhase($token);
3529:
3530: /* An end-of-file token */
3531: } elseif($token['type'] === HTML5::EOF) {
3532: /* OMG DONE!! */
3533: }
3534: }
3535:
3536: private function insertElement($token, $append = true, $check = false) {
3537: // Proprietary workaround for libxml2's limitations with tag names
3538: if ($check) {
3539: // Slightly modified HTML5 tag-name modification,
3540: // removing anything that's not an ASCII letter, digit, or hyphen
3541: $token['name'] = preg_replace('/[^a-z0-9-]/i', '', $token['name']);
3542: // Remove leading hyphens and numbers
3543: $token['name'] = ltrim($token['name'], '-0..9');
3544: // In theory, this should ever be needed, but just in case
3545: if ($token['name'] === '') $token['name'] = 'span'; // arbitrary generic choice
3546: }
3547:
3548: $el = $this->dom->createElement($token['name']);
3549:
3550: foreach($token['attr'] as $attr) {
3551: if(!$el->hasAttribute($attr['name'])) {
3552: $el->setAttribute($attr['name'], $attr['value']);
3553: }
3554: }
3555:
3556: $this->appendToRealParent($el);
3557: $this->stack[] = $el;
3558:
3559: return $el;
3560: }
3561:
3562: private function insertText($data) {
3563: $text = $this->dom->createTextNode($data);
3564: $this->appendToRealParent($text);
3565: }
3566:
3567: private function insertComment($data) {
3568: $comment = $this->dom->createComment($data);
3569: $this->appendToRealParent($comment);
3570: }
3571:
3572: private function appendToRealParent($node) {
3573: if($this->foster_parent === null) {
3574: end($this->stack)->appendChild($node);
3575:
3576: } elseif($this->foster_parent !== null) {
3577: /* If the foster parent element is the parent element of the
3578: last table element in the stack of open elements, then the new
3579: node must be inserted immediately before the last table element
3580: in the stack of open elements in the foster parent element;
3581: otherwise, the new node must be appended to the foster parent
3582: element. */
3583: for($n = count($this->stack) - 1; $n >= 0; $n--) {
3584: if($this->stack[$n]->nodeName === 'table' &&
3585: $this->stack[$n]->parentNode !== null) {
3586: $table = $this->stack[$n];
3587: break;
3588: }
3589: }
3590:
3591: if(isset($table) && $this->foster_parent->isSameNode($table->parentNode))
3592: $this->foster_parent->insertBefore($node, $table);
3593: else
3594: $this->foster_parent->appendChild($node);
3595:
3596: $this->foster_parent = null;
3597: }
3598: }
3599:
3600: private function elementInScope($el, $table = false) {
3601: if(is_array($el)) {
3602: foreach($el as $element) {
3603: if($this->elementInScope($element, $table)) {
3604: return true;
3605: }
3606: }
3607:
3608: return false;
3609: }
3610:
3611: $leng = count($this->stack);
3612:
3613: for($n = 0; $n < $leng; $n++) {
3614: /* 1. Initialise node to be the current node (the bottommost node of
3615: the stack). */
3616: $node = $this->stack[$leng - 1 - $n];
3617:
3618: if($node->tagName === $el) {
3619: /* 2. If node is the target node, terminate in a match state. */
3620: return true;
3621:
3622: } elseif($node->tagName === 'table') {
3623: /* 3. Otherwise, if node is a table element, terminate in a failure
3624: state. */
3625: return false;
3626:
3627: } elseif($table === true && in_array($node->tagName, array('caption', 'td',
3628: 'th', 'button', 'marquee', 'object'))) {
3629: /* 4. Otherwise, if the algorithm is the "has an element in scope"
3630: variant (rather than the "has an element in table scope" variant),
3631: and node is one of the following, terminate in a failure state. */
3632: return false;
3633:
3634: } elseif($node === $node->ownerDocument->documentElement) {
3635: /* 5. Otherwise, if node is an html element (root element), terminate
3636: in a failure state. (This can only happen if the node is the topmost
3637: node of the stack of open elements, and prevents the next step from
3638: being invoked if there are no more elements in the stack.) */
3639: return false;
3640: }
3641:
3642: /* Otherwise, set node to the previous entry in the stack of open
3643: elements and return to step 2. (This will never fail, since the loop
3644: will always terminate in the previous step if the top of the stack
3645: is reached.) */
3646: }
3647: }
3648:
3649: private function reconstructActiveFormattingElements() {
3650: /* 1. If there are no entries in the list of active formatting elements,
3651: then there is nothing to reconstruct; stop this algorithm. */
3652: $formatting_elements = count($this->a_formatting);
3653:
3654: if($formatting_elements === 0) {
3655: return false;
3656: }
3657:
3658: /* 3. Let entry be the last (most recently added) element in the list
3659: of active formatting elements. */
3660: $entry = end($this->a_formatting);
3661:
3662: /* 2. If the last (most recently added) entry in the list of active
3663: formatting elements is a marker, or if it is an element that is in the
3664: stack of open elements, then there is nothing to reconstruct; stop this
3665: algorithm. */
3666: if($entry === self::MARKER || in_array($entry, $this->stack, true)) {
3667: return false;
3668: }
3669:
3670: for($a = $formatting_elements - 1; $a >= 0; true) {
3671: /* 4. If there are no entries before entry in the list of active
3672: formatting elements, then jump to step 8. */
3673: if($a === 0) {
3674: $step_seven = false;
3675: break;
3676: }
3677:
3678: /* 5. Let entry be the entry one earlier than entry in the list of
3679: active formatting elements. */
3680: $a--;
3681: $entry = $this->a_formatting[$a];
3682:
3683: /* 6. If entry is neither a marker nor an element that is also in
3684: thetack of open elements, go to step 4. */
3685: if($entry === self::MARKER || in_array($entry, $this->stack, true)) {
3686: break;
3687: }
3688: }
3689:
3690: while(true) {
3691: /* 7. Let entry be the element one later than entry in the list of
3692: active formatting elements. */
3693: if(isset($step_seven) && $step_seven === true) {
3694: $a++;
3695: $entry = $this->a_formatting[$a];
3696: }
3697:
3698: /* 8. Perform a shallow clone of the element entry to obtain clone. */
3699: $clone = $entry->cloneNode();
3700:
3701: /* 9. Append clone to the current node and push it onto the stack
3702: of open elements so that it is the new current node. */
3703: end($this->stack)->appendChild($clone);
3704: $this->stack[] = $clone;
3705:
3706: /* 10. Replace the entry for entry in the list with an entry for
3707: clone. */
3708: $this->a_formatting[$a] = $clone;
3709:
3710: /* 11. If the entry for clone in the list of active formatting
3711: elements is not the last entry in the list, return to step 7. */
3712: if(end($this->a_formatting) !== $clone) {
3713: $step_seven = true;
3714: } else {
3715: break;
3716: }
3717: }
3718: }
3719:
3720: private function clearTheActiveFormattingElementsUpToTheLastMarker() {
3721: /* When the steps below require the UA to clear the list of active
3722: formatting elements up to the last marker, the UA must perform the
3723: following steps: */
3724:
3725: while(true) {
3726: /* 1. Let entry be the last (most recently added) entry in the list
3727: of active formatting elements. */
3728: $entry = end($this->a_formatting);
3729:
3730: /* 2. Remove entry from the list of active formatting elements. */
3731: array_pop($this->a_formatting);
3732:
3733: /* 3. If entry was a marker, then stop the algorithm at this point.
3734: The list has been cleared up to the last marker. */
3735: if($entry === self::MARKER) {
3736: break;
3737: }
3738: }
3739: }
3740:
3741: private function generateImpliedEndTags($exclude = array()) {
3742: /* When the steps below require the UA to generate implied end tags,
3743: then, if the current node is a dd element, a dt element, an li element,
3744: a p element, a td element, a th element, or a tr element, the UA must
3745: act as if an end tag with the respective tag name had been seen and
3746: then generate implied end tags again. */
3747: $node = end($this->stack);
3748: $elements = array_diff(array('dd', 'dt', 'li', 'p', 'td', 'th', 'tr'), $exclude);
3749:
3750: while(in_array(end($this->stack)->nodeName, $elements)) {
3751: array_pop($this->stack);
3752: }
3753: }
3754:
3755: private function getElementCategory($node) {
3756: $name = $node->tagName;
3757: if(in_array($name, $this->special))
3758: return self::SPECIAL;
3759:
3760: elseif(in_array($name, $this->scoping))
3761: return self::SCOPING;
3762:
3763: elseif(in_array($name, $this->formatting))
3764: return self::FORMATTING;
3765:
3766: else
3767: return self::PHRASING;
3768: }
3769:
3770: private function clearStackToTableContext($elements) {
3771: /* When the steps above require the UA to clear the stack back to a
3772: table context, it means that the UA must, while the current node is not
3773: a table element or an html element, pop elements from the stack of open
3774: elements. If this causes any elements to be popped from the stack, then
3775: this is a parse error. */
3776: while(true) {
3777: $node = end($this->stack)->nodeName;
3778:
3779: if(in_array($node, $elements)) {
3780: break;
3781: } else {
3782: array_pop($this->stack);
3783: }
3784: }
3785: }
3786:
3787: private function resetInsertionMode() {
3788: /* 1. Let last be false. */
3789: $last = false;
3790: $leng = count($this->stack);
3791:
3792: for($n = $leng - 1; $n >= 0; $n--) {
3793: /* 2. Let node be the last node in the stack of open elements. */
3794: $node = $this->stack[$n];
3795:
3796: /* 3. If node is the first node in the stack of open elements, then
3797: set last to true. If the element whose innerHTML attribute is being
3798: set is neither a td element nor a th element, then set node to the
3799: element whose innerHTML attribute is being set. (innerHTML case) */
3800: if($this->stack[0]->isSameNode($node)) {
3801: $last = true;
3802: }
3803:
3804: /* 4. If node is a select element, then switch the insertion mode to
3805: "in select" and abort these steps. (innerHTML case) */
3806: if($node->nodeName === 'select') {
3807: $this->mode = self::IN_SELECT;
3808: break;
3809:
3810: /* 5. If node is a td or th element, then switch the insertion mode
3811: to "in cell" and abort these steps. */
3812: } elseif($node->nodeName === 'td' || $node->nodeName === 'th') {
3813: $this->mode = self::IN_CELL;
3814: break;
3815:
3816: /* 6. If node is a tr element, then switch the insertion mode to
3817: "in row" and abort these steps. */
3818: } elseif($node->nodeName === 'tr') {
3819: $this->mode = self::IN_ROW;
3820: break;
3821:
3822: /* 7. If node is a tbody, thead, or tfoot element, then switch the
3823: insertion mode to "in table body" and abort these steps. */
3824: } elseif(in_array($node->nodeName, array('tbody', 'thead', 'tfoot'))) {
3825: $this->mode = self::IN_TBODY;
3826: break;
3827:
3828: /* 8. If node is a caption element, then switch the insertion mode
3829: to "in caption" and abort these steps. */
3830: } elseif($node->nodeName === 'caption') {
3831: $this->mode = self::IN_CAPTION;
3832: break;
3833:
3834: /* 9. If node is a colgroup element, then switch the insertion mode
3835: to "in column group" and abort these steps. (innerHTML case) */
3836: } elseif($node->nodeName === 'colgroup') {
3837: $this->mode = self::IN_CGROUP;
3838: break;
3839:
3840: /* 10. If node is a table element, then switch the insertion mode
3841: to "in table" and abort these steps. */
3842: } elseif($node->nodeName === 'table') {
3843: $this->mode = self::IN_TABLE;
3844: break;
3845:
3846: /* 11. If node is a head element, then switch the insertion mode
3847: to "in body" ("in body"! not "in head"!) and abort these steps.
3848: (innerHTML case) */
3849: } elseif($node->nodeName === 'head') {
3850: $this->mode = self::IN_BODY;
3851: break;
3852:
3853: /* 12. If node is a body element, then switch the insertion mode to
3854: "in body" and abort these steps. */
3855: } elseif($node->nodeName === 'body') {
3856: $this->mode = self::IN_BODY;
3857: break;
3858:
3859: /* 13. If node is a frameset element, then switch the insertion
3860: mode to "in frameset" and abort these steps. (innerHTML case) */
3861: } elseif($node->nodeName === 'frameset') {
3862: $this->mode = self::IN_FRAME;
3863: break;
3864:
3865: /* 14. If node is an html element, then: if the head element
3866: pointer is null, switch the insertion mode to "before head",
3867: otherwise, switch the insertion mode to "after head". In either
3868: case, abort these steps. (innerHTML case) */
3869: } elseif($node->nodeName === 'html') {
3870: $this->mode = ($this->head_pointer === null)
3871: ? self::BEFOR_HEAD
3872: : self::AFTER_HEAD;
3873:
3874: break;
3875:
3876: /* 15. If last is true, then set the insertion mode to "in body"
3877: and abort these steps. (innerHTML case) */
3878: } elseif($last) {
3879: $this->mode = self::IN_BODY;
3880: break;
3881: }
3882: }
3883: }
3884:
3885: private function closeCell() {
3886: /* If the stack of open elements has a td or th element in table scope,
3887: then act as if an end tag token with that tag name had been seen. */
3888: foreach(array('td', 'th') as $cell) {
3889: if($this->elementInScope($cell, true)) {
3890: $this->inCell(array(
3891: 'name' => $cell,
3892: 'type' => HTML5::ENDTAG
3893: ));
3894:
3895: break;
3896: }
3897: }
3898: }
3899:
3900: public function save() {
3901: return $this->dom;
3902: }
3903: }
3904: ?>
3905: