]>
Commit | Line | Data |
---|---|---|
1 | <?php | |
2 | ||
3 | /* | |
4 | ||
5 | Copyright 2007 Jeroen van der Meer <http://jero.net/> | |
6 | Copyright 2008 Edward Z. Yang <http://htmlpurifier.org/> | |
7 | Copyright 2009 Geoffrey Sneddon <http://gsnedders.com/> | |
8 | ||
9 | Permission is hereby granted, free of charge, to any person obtaining a | |
10 | copy of this software and associated documentation files (the | |
11 | "Software"), to deal in the Software without restriction, including | |
12 | without limitation the rights to use, copy, modify, merge, publish, | |
13 | distribute, sublicense, and/or sell copies of the Software, and to | |
14 | permit persons to whom the Software is furnished to do so, subject to | |
15 | the following conditions: | |
16 | ||
17 | The above copyright notice and this permission notice shall be included | |
18 | in all copies or substantial portions of the Software. | |
19 | ||
20 | THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS | |
21 | OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF | |
22 | MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. | |
23 | IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY | |
24 | CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, | |
25 | TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE | |
26 | SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. | |
27 | ||
28 | */ | |
29 | ||
30 | // Some conventions: | |
31 | // /* */ indicates verbatim text from the HTML 5 specification | |
32 | // // indicates regular comments | |
33 | ||
34 | // all flags are in hyphenated form | |
35 | ||
36 | class HTML5_Tokenizer { | |
37 | /** | |
38 | * Points to an InputStream object. | |
39 | */ | |
40 | protected $stream; | |
41 | ||
42 | /** | |
43 | * Tree builder that the tokenizer emits token to. | |
44 | */ | |
45 | private $tree; | |
46 | ||
47 | /** | |
48 | * Current content model we are parsing as. | |
49 | */ | |
50 | protected $content_model; | |
51 | ||
52 | /** | |
53 | * Current token that is being built, but not yet emitted. Also | |
54 | * is the last token emitted, if applicable. | |
55 | */ | |
56 | protected $token; | |
57 | ||
58 | // These are constants describing the content model | |
59 | const PCDATA = 0; | |
60 | const RCDATA = 1; | |
61 | const CDATA = 2; | |
62 | const PLAINTEXT = 3; | |
63 | ||
64 | // These are constants describing tokens | |
65 | // XXX should probably be moved somewhere else, probably the | |
66 | // HTML5 class. | |
67 | const DOCTYPE = 0; | |
68 | const STARTTAG = 1; | |
69 | const ENDTAG = 2; | |
70 | const COMMENT = 3; | |
71 | const CHARACTER = 4; | |
72 | const SPACECHARACTER = 5; | |
73 | const EOF = 6; | |
74 | const PARSEERROR = 7; | |
75 | ||
76 | // These are constants representing bunches of characters. | |
77 | const ALPHA = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz'; | |
78 | const UPPER_ALPHA = 'ABCDEFGHIJKLMNOPQRSTUVWXYZ'; | |
79 | const LOWER_ALPHA = 'abcdefghijklmnopqrstuvwxyz'; | |
80 | const DIGIT = '0123456789'; | |
81 | const HEX = '0123456789ABCDEFabcdef'; | |
82 | const WHITESPACE = "\t\n\x0c "; | |
83 | ||
84 | /** | |
85 | * @param $data Data to parse | |
86 | */ | |
87 | public function __construct($data, $builder = null) { | |
88 | $this->stream = new HTML5_InputStream($data); | |
89 | if (!$builder) $this->tree = new HTML5_TreeBuilder; | |
90 | else $this->tree = $builder; | |
91 | $this->content_model = self::PCDATA; | |
92 | } | |
93 | ||
94 | public function parseFragment($context = null) { | |
95 | $this->tree->setupContext($context); | |
96 | if ($this->tree->content_model) { | |
97 | $this->content_model = $this->tree->content_model; | |
98 | $this->tree->content_model = null; | |
99 | } | |
100 | $this->parse(); | |
101 | } | |
102 | ||
103 | // XXX maybe convert this into an iterator? regardless, this function | |
104 | // and the save function should go into a Parser facade of some sort | |
105 | /** | |
106 | * Performs the actual parsing of the document. | |
107 | */ | |
108 | public function parse() { | |
109 | // Current state | |
110 | $state = 'data'; | |
111 | // This is used to avoid having to have look-behind in the data state. | |
112 | $lastFourChars = ''; | |
113 | /** | |
114 | * Escape flag as specified by the HTML5 specification: "used to | |
115 | * control the behavior of the tokeniser. It is either true or | |
116 | * false, and initially must be set to the false state." | |
117 | */ | |
118 | $escape = false; | |
119 | //echo "\n\n"; | |
120 | while($state !== null) { | |
121 | ||
122 | /*echo $state . ' '; | |
123 | switch ($this->content_model) { | |
124 | case self::PCDATA: echo 'PCDATA'; break; | |
125 | case self::RCDATA: echo 'RCDATA'; break; | |
126 | case self::CDATA: echo 'CDATA'; break; | |
127 | case self::PLAINTEXT: echo 'PLAINTEXT'; break; | |
128 | } | |
129 | if ($escape) echo " escape"; | |
130 | echo "\n";*/ | |
131 | ||
132 | switch($state) { | |
133 | case 'data': | |
134 | ||
135 | /* Consume the next input character */ | |
136 | $char = $this->stream->char(); | |
137 | $lastFourChars .= $char; | |
138 | if (strlen($lastFourChars) > 4) $lastFourChars = substr($lastFourChars, -4); | |
139 | ||
140 | // see below for meaning | |
141 | $hyp_cond = | |
142 | !$escape && | |
143 | ( | |
144 | $this->content_model === self::RCDATA || | |
145 | $this->content_model === self::CDATA | |
146 | ); | |
147 | $amp_cond = | |
148 | !$escape && | |
149 | ( | |
150 | $this->content_model === self::PCDATA || | |
151 | $this->content_model === self::RCDATA | |
152 | ); | |
153 | $lt_cond = | |
154 | $this->content_model === self::PCDATA || | |
155 | ( | |
156 | ( | |
157 | $this->content_model === self::RCDATA || | |
158 | $this->content_model === self::CDATA | |
159 | ) && | |
160 | !$escape | |
161 | ); | |
162 | $gt_cond = | |
163 | $escape && | |
164 | ( | |
165 | $this->content_model === self::RCDATA || | |
166 | $this->content_model === self::CDATA | |
167 | ); | |
168 | ||
169 | if($char === '&' && $amp_cond) { | |
170 | /* U+0026 AMPERSAND (&) | |
171 | When the content model flag is set to one of the PCDATA or RCDATA | |
172 | states and the escape flag is false: switch to the | |
173 | character reference data state. Otherwise: treat it as per | |
174 | the "anything else" entry below. */ | |
175 | $state = 'character reference data'; | |
176 | ||
177 | } elseif( | |
178 | $char === '-' && | |
179 | $hyp_cond && | |
180 | $lastFourChars === '<!--' | |
181 | ) { | |
182 | /* | |
183 | U+002D HYPHEN-MINUS (-) | |
184 | If the content model flag is set to either the RCDATA state or | |
185 | the CDATA state, and the escape flag is false, and there are at | |
186 | least three characters before this one in the input stream, and the | |
187 | last four characters in the input stream, including this one, are | |
188 | U+003C LESS-THAN SIGN, U+0021 EXCLAMATION MARK, U+002D HYPHEN-MINUS, | |
189 | and U+002D HYPHEN-MINUS ("<!--"), then set the escape flag to true. */ | |
190 | $escape = true; | |
191 | ||
192 | /* In any case, emit the input character as a character token. Stay | |
193 | in the data state. */ | |
194 | $this->emitToken(array( | |
195 | 'type' => self::CHARACTER, | |
196 | 'data' => '-' | |
197 | )); | |
198 | // We do the "any case" part as part of "anything else". | |
199 | ||
200 | /* U+003C LESS-THAN SIGN (<) */ | |
201 | } elseif($char === '<' && $lt_cond) { | |
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 | $state = 'tag open'; | |
211 | ||
212 | /* U+003E GREATER-THAN SIGN (>) */ | |
213 | } elseif( | |
214 | $char === '>' && | |
215 | $gt_cond && | |
216 | substr($lastFourChars, 1) === '-->' | |
217 | ) { | |
218 | /* If the content model flag is set to either the RCDATA state or | |
219 | the CDATA state, and the escape flag is true, and the last three | |
220 | characters in the input stream including this one are U+002D | |
221 | HYPHEN-MINUS, U+002D HYPHEN-MINUS, U+003E GREATER-THAN SIGN ("-->"), | |
222 | set the escape flag to false. */ | |
223 | $escape = false; | |
224 | ||
225 | /* In any case, emit the input character as a character token. | |
226 | Stay in the data state. */ | |
227 | $this->emitToken(array( | |
228 | 'type' => self::CHARACTER, | |
229 | 'data' => '>' | |
230 | )); | |
231 | // We do the "any case" part as part of "anything else". | |
232 | ||
233 | } elseif($char === false) { | |
234 | /* EOF | |
235 | Emit an end-of-file token. */ | |
236 | $state = null; | |
237 | $this->tree->emitToken(array( | |
238 | 'type' => self::EOF | |
239 | )); | |
240 | ||
241 | } elseif($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
242 | // Directly after emitting a token you switch back to the "data | |
243 | // state". At that point spaceCharacters are important so they are | |
244 | // emitted separately. | |
245 | $chars = $this->stream->charsWhile(self::WHITESPACE); | |
246 | $this->emitToken(array( | |
247 | 'type' => self::SPACECHARACTER, | |
248 | 'data' => $char . $chars | |
249 | )); | |
250 | $lastFourChars .= $chars; | |
251 | if (strlen($lastFourChars) > 4) $lastFourChars = substr($lastFourChars, -4); | |
252 | ||
253 | } else { | |
254 | /* Anything else | |
255 | THIS IS AN OPTIMIZATION: Get as many character that | |
256 | otherwise would also be treated as a character token and emit it | |
257 | as a single character token. Stay in the data state. */ | |
258 | ||
259 | $mask = ''; | |
260 | if ($hyp_cond) $mask .= '-'; | |
261 | if ($amp_cond) $mask .= '&'; | |
262 | if ($lt_cond) $mask .= '<'; | |
263 | if ($gt_cond) $mask .= '>'; | |
264 | ||
265 | if ($mask === '') { | |
266 | $chars = $this->stream->remainingChars(); | |
267 | } else { | |
268 | $chars = $this->stream->charsUntil($mask); | |
269 | } | |
270 | ||
271 | $this->emitToken(array( | |
272 | 'type' => self::CHARACTER, | |
273 | 'data' => $char . $chars | |
274 | )); | |
275 | ||
276 | $lastFourChars .= $chars; | |
277 | if (strlen($lastFourChars) > 4) $lastFourChars = substr($lastFourChars, -4); | |
278 | ||
279 | $state = 'data'; | |
280 | } | |
281 | break; | |
282 | ||
283 | case 'character reference data': | |
284 | /* (This cannot happen if the content model flag | |
285 | is set to the CDATA state.) */ | |
286 | ||
287 | /* Attempt to consume a character reference, with no | |
288 | additional allowed character. */ | |
289 | $entity = $this->consumeCharacterReference(); | |
290 | ||
291 | /* If nothing is returned, emit a U+0026 AMPERSAND | |
292 | character token. Otherwise, emit the character token that | |
293 | was returned. */ | |
294 | // This is all done when consuming the character reference. | |
295 | $this->emitToken(array( | |
296 | 'type' => self::CHARACTER, | |
297 | 'data' => $entity | |
298 | )); | |
299 | ||
300 | /* Finally, switch to the data state. */ | |
301 | $state = 'data'; | |
302 | break; | |
303 | ||
304 | case 'tag open': | |
305 | $char = $this->stream->char(); | |
306 | ||
307 | switch($this->content_model) { | |
308 | case self::RCDATA: | |
309 | case self::CDATA: | |
310 | /* Consume the next input character. If it is a | |
311 | U+002F SOLIDUS (/) character, switch to the close | |
312 | tag open state. Otherwise, emit a U+003C LESS-THAN | |
313 | SIGN character token and reconsume the current input | |
314 | character in the data state. */ | |
315 | // We consumed above. | |
316 | ||
317 | if($char === '/') { | |
318 | $state = 'close tag open'; | |
319 | ||
320 | } else { | |
321 | $this->emitToken(array( | |
322 | 'type' => self::CHARACTER, | |
323 | 'data' => '<' | |
324 | )); | |
325 | ||
326 | $this->stream->unget(); | |
327 | ||
328 | $state = 'data'; | |
329 | } | |
330 | break; | |
331 | ||
332 | case self::PCDATA: | |
333 | /* If the content model flag is set to the PCDATA state | |
334 | Consume the next input character: */ | |
335 | // We consumed above. | |
336 | ||
337 | if($char === '!') { | |
338 | /* U+0021 EXCLAMATION MARK (!) | |
339 | Switch to the markup declaration open state. */ | |
340 | $state = 'markup declaration open'; | |
341 | ||
342 | } elseif($char === '/') { | |
343 | /* U+002F SOLIDUS (/) | |
344 | Switch to the close tag open state. */ | |
345 | $state = 'close tag open'; | |
346 | ||
347 | } elseif('A' <= $char && $char <= 'Z') { | |
348 | /* U+0041 LATIN LETTER A through to U+005A LATIN LETTER Z | |
349 | Create a new start tag token, set its tag name to the lowercase | |
350 | version of the input character (add 0x0020 to the character's code | |
351 | point), then switch to the tag name state. (Don't emit the token | |
352 | yet; further details will be filled in before it is emitted.) */ | |
353 | $this->token = array( | |
354 | 'name' => strtolower($char), | |
355 | 'type' => self::STARTTAG, | |
356 | 'attr' => array() | |
357 | ); | |
358 | ||
359 | $state = 'tag name'; | |
360 | ||
361 | } elseif('a' <= $char && $char <= 'z') { | |
362 | /* U+0061 LATIN SMALL LETTER A through to U+007A LATIN SMALL LETTER Z | |
363 | Create a new start tag token, set its tag name to the input | |
364 | character, then switch to the tag name state. (Don't emit | |
365 | the token yet; further details will be filled in before it | |
366 | is emitted.) */ | |
367 | $this->token = array( | |
368 | 'name' => $char, | |
369 | 'type' => self::STARTTAG, | |
370 | 'attr' => array() | |
371 | ); | |
372 | ||
373 | $state = 'tag name'; | |
374 | ||
375 | } elseif($char === '>') { | |
376 | /* U+003E GREATER-THAN SIGN (>) | |
377 | Parse error. Emit a U+003C LESS-THAN SIGN character token and a | |
378 | U+003E GREATER-THAN SIGN character token. Switch to the data state. */ | |
379 | $this->emitToken(array( | |
380 | 'type' => self::PARSEERROR, | |
381 | 'data' => 'expected-tag-name-but-got-right-bracket' | |
382 | )); | |
383 | $this->emitToken(array( | |
384 | 'type' => self::CHARACTER, | |
385 | 'data' => '<>' | |
386 | )); | |
387 | ||
388 | $state = 'data'; | |
389 | ||
390 | } elseif($char === '?') { | |
391 | /* U+003F QUESTION MARK (?) | |
392 | Parse error. Switch to the bogus comment state. */ | |
393 | $this->emitToken(array( | |
394 | 'type' => self::PARSEERROR, | |
395 | 'data' => 'expected-tag-name-but-got-question-mark' | |
396 | )); | |
397 | $this->token = array( | |
398 | 'data' => '?', | |
399 | 'type' => self::COMMENT | |
400 | ); | |
401 | $state = 'bogus comment'; | |
402 | ||
403 | } else { | |
404 | /* Anything else | |
405 | Parse error. Emit a U+003C LESS-THAN SIGN character token and | |
406 | reconsume the current input character in the data state. */ | |
407 | $this->emitToken(array( | |
408 | 'type' => self::PARSEERROR, | |
409 | 'data' => 'expected-tag-name' | |
410 | )); | |
411 | $this->emitToken(array( | |
412 | 'type' => self::CHARACTER, | |
413 | 'data' => '<' | |
414 | )); | |
415 | ||
416 | $state = 'data'; | |
417 | $this->stream->unget(); | |
418 | } | |
419 | break; | |
420 | } | |
421 | break; | |
422 | ||
423 | case 'close tag open': | |
424 | if ( | |
425 | $this->content_model === self::RCDATA || | |
426 | $this->content_model === self::CDATA | |
427 | ) { | |
428 | /* If the content model flag is set to the RCDATA or CDATA | |
429 | states... */ | |
430 | $name = strtolower($this->stream->charsWhile(self::ALPHA)); | |
431 | $following = $this->stream->char(); | |
432 | $this->stream->unget(); | |
433 | if ( | |
434 | !$this->token || | |
435 | $this->token['name'] !== $name || | |
436 | $this->token['name'] === $name && !in_array($following, array("\x09", "\x0A", "\x0C", "\x20", "\x3E", "\x2F", false)) | |
437 | ) { | |
438 | /* if no start tag token has ever been emitted by this instance | |
439 | of the tokenizer (fragment case), or, if the next few | |
440 | characters do not match the tag name of the last start tag | |
441 | token emitted (compared in an ASCII case-insensitive manner), | |
442 | or if they do but they are not immediately followed by one of | |
443 | the following characters: | |
444 | ||
445 | * U+0009 CHARACTER TABULATION | |
446 | * U+000A LINE FEED (LF) | |
447 | * U+000C FORM FEED (FF) | |
448 | * U+0020 SPACE | |
449 | * U+003E GREATER-THAN SIGN (>) | |
450 | * U+002F SOLIDUS (/) | |
451 | * EOF | |
452 | ||
453 | ...then emit a U+003C LESS-THAN SIGN character token, a | |
454 | U+002F SOLIDUS character token, and switch to the data | |
455 | state to process the next input character. */ | |
456 | // XXX: Probably ought to replace in_array with $following === x ||... | |
457 | ||
458 | // We also need to emit $name now we've consumed that, as we | |
459 | // know it'll just be emitted as a character token. | |
460 | $this->emitToken(array( | |
461 | 'type' => self::CHARACTER, | |
462 | 'data' => '</' . $name | |
463 | )); | |
464 | ||
465 | $state = 'data'; | |
466 | } else { | |
467 | // This matches what would happen if we actually did the | |
468 | // otherwise below (but we can't because we've consumed too | |
469 | // much). | |
470 | ||
471 | // Start the end tag token with the name we already have. | |
472 | $this->token = array( | |
473 | 'name' => $name, | |
474 | 'type' => self::ENDTAG | |
475 | ); | |
476 | ||
477 | // Change to tag name state. | |
478 | $state = 'tag name'; | |
479 | } | |
480 | } elseif ($this->content_model === self::PCDATA) { | |
481 | /* Otherwise, if the content model flag is set to the PCDATA | |
482 | state [...]: */ | |
483 | $char = $this->stream->char(); | |
484 | ||
485 | if ('A' <= $char && $char <= 'Z') { | |
486 | /* U+0041 LATIN LETTER A through to U+005A LATIN LETTER Z | |
487 | Create a new end tag token, set its tag name to the lowercase version | |
488 | of the input character (add 0x0020 to the character's code point), then | |
489 | switch to the tag name state. (Don't emit the token yet; further details | |
490 | will be filled in before it is emitted.) */ | |
491 | $this->token = array( | |
492 | 'name' => strtolower($char), | |
493 | 'type' => self::ENDTAG | |
494 | ); | |
495 | ||
496 | $state = 'tag name'; | |
497 | ||
498 | } elseif ('a' <= $char && $char <= 'z') { | |
499 | /* U+0061 LATIN SMALL LETTER A through to U+007A LATIN SMALL LETTER Z | |
500 | Create a new end tag token, set its tag name to the | |
501 | input character, then switch to the tag name state. | |
502 | (Don't emit the token yet; further details will be | |
503 | filled in before it is emitted.) */ | |
504 | $this->token = array( | |
505 | 'name' => $char, | |
506 | 'type' => self::ENDTAG | |
507 | ); | |
508 | ||
509 | $state = 'tag name'; | |
510 | ||
511 | } elseif($char === '>') { | |
512 | /* U+003E GREATER-THAN SIGN (>) | |
513 | Parse error. Switch to the data state. */ | |
514 | $this->emitToken(array( | |
515 | 'type' => self::PARSEERROR, | |
516 | 'data' => 'expected-closing-tag-but-got-right-bracket' | |
517 | )); | |
518 | $state = 'data'; | |
519 | ||
520 | } elseif($char === false) { | |
521 | /* EOF | |
522 | Parse error. Emit a U+003C LESS-THAN SIGN character token and a U+002F | |
523 | SOLIDUS character token. Reconsume the EOF character in the data state. */ | |
524 | $this->emitToken(array( | |
525 | 'type' => self::PARSEERROR, | |
526 | 'data' => 'expected-closing-tag-but-got-eof' | |
527 | )); | |
528 | $this->emitToken(array( | |
529 | 'type' => self::CHARACTER, | |
530 | 'data' => '</' | |
531 | )); | |
532 | ||
533 | $this->stream->unget(); | |
534 | $state = 'data'; | |
535 | ||
536 | } else { | |
537 | /* Parse error. Switch to the bogus comment state. */ | |
538 | $this->emitToken(array( | |
539 | 'type' => self::PARSEERROR, | |
540 | 'data' => 'expected-closing-tag-but-got-char' | |
541 | )); | |
542 | $this->token = array( | |
543 | 'data' => $char, | |
544 | 'type' => self::COMMENT | |
545 | ); | |
546 | $state = 'bogus comment'; | |
547 | } | |
548 | } | |
549 | break; | |
550 | ||
551 | case 'tag name': | |
552 | /* Consume the next input character: */ | |
553 | $char = $this->stream->char(); | |
554 | ||
555 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
556 | /* U+0009 CHARACTER TABULATION | |
557 | U+000A LINE FEED (LF) | |
558 | U+000C FORM FEED (FF) | |
559 | U+0020 SPACE | |
560 | Switch to the before attribute name state. */ | |
561 | $state = 'before attribute name'; | |
562 | ||
563 | } elseif($char === '/') { | |
564 | /* U+002F SOLIDUS (/) | |
565 | Switch to the self-closing start tag state. */ | |
566 | $state = 'self-closing start tag'; | |
567 | ||
568 | } elseif($char === '>') { | |
569 | /* U+003E GREATER-THAN SIGN (>) | |
570 | Emit the current tag token. Switch to the data state. */ | |
571 | $this->emitToken($this->token); | |
572 | $state = 'data'; | |
573 | ||
574 | } elseif('A' <= $char && $char <= 'Z') { | |
575 | /* U+0041 LATIN CAPITAL LETTER A through to U+005A LATIN CAPITAL LETTER Z | |
576 | Append the lowercase version of the current input | |
577 | character (add 0x0020 to the character's code point) to | |
578 | the current tag token's tag name. Stay in the tag name state. */ | |
579 | $chars = $this->stream->charsWhile(self::UPPER_ALPHA); | |
580 | ||
581 | $this->token['name'] .= strtolower($char . $chars); | |
582 | $state = 'tag name'; | |
583 | ||
584 | } elseif($char === false) { | |
585 | /* EOF | |
586 | Parse error. Reconsume the EOF character in the data state. */ | |
587 | $this->emitToken(array( | |
588 | 'type' => self::PARSEERROR, | |
589 | 'data' => 'eof-in-tag-name' | |
590 | )); | |
591 | ||
592 | $this->stream->unget(); | |
593 | $state = 'data'; | |
594 | ||
595 | } else { | |
596 | /* Anything else | |
597 | Append the current input character to the current tag token's tag name. | |
598 | Stay in the tag name state. */ | |
599 | $chars = $this->stream->charsUntil("\t\n\x0C />" . self::UPPER_ALPHA); | |
600 | ||
601 | $this->token['name'] .= $char . $chars; | |
602 | $state = 'tag name'; | |
603 | } | |
604 | break; | |
605 | ||
606 | case 'before attribute name': | |
607 | /* Consume the next input character: */ | |
608 | $char = $this->stream->char(); | |
609 | ||
610 | // this conditional is optimized, check bottom | |
611 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
612 | /* U+0009 CHARACTER TABULATION | |
613 | U+000A LINE FEED (LF) | |
614 | U+000C FORM FEED (FF) | |
615 | U+0020 SPACE | |
616 | Stay in the before attribute name state. */ | |
617 | $state = 'before attribute name'; | |
618 | ||
619 | } elseif($char === '/') { | |
620 | /* U+002F SOLIDUS (/) | |
621 | Switch to the self-closing start tag state. */ | |
622 | $state = 'self-closing start tag'; | |
623 | ||
624 | } elseif($char === '>') { | |
625 | /* U+003E GREATER-THAN SIGN (>) | |
626 | Emit the current tag token. Switch to the data state. */ | |
627 | $this->emitToken($this->token); | |
628 | $state = 'data'; | |
629 | ||
630 | } elseif('A' <= $char && $char <= 'Z') { | |
631 | /* U+0041 LATIN CAPITAL LETTER A through to U+005A LATIN CAPITAL LETTER Z | |
632 | Start a new attribute in the current tag token. Set that | |
633 | attribute's name to the lowercase version of the current | |
634 | input character (add 0x0020 to the character's code | |
635 | point), and its value to the empty string. Switch to the | |
636 | attribute name state.*/ | |
637 | $this->token['attr'][] = array( | |
638 | 'name' => strtolower($char), | |
639 | 'value' => '' | |
640 | ); | |
641 | ||
642 | $state = 'attribute name'; | |
643 | ||
644 | } elseif($char === false) { | |
645 | /* EOF | |
646 | Parse error. Reconsume the EOF character in the data state. */ | |
647 | $this->emitToken(array( | |
648 | 'type' => self::PARSEERROR, | |
649 | 'data' => 'expected-attribute-name-but-got-eof' | |
650 | )); | |
651 | ||
652 | $this->stream->unget(); | |
653 | $state = 'data'; | |
654 | ||
655 | } else { | |
656 | /* U+0022 QUOTATION MARK (") | |
657 | U+0027 APOSTROPHE (') | |
658 | U+003C LESS-THAN SIGN (<) | |
659 | U+003D EQUALS SIGN (=) | |
660 | Parse error. Treat it as per the "anything else" entry | |
661 | below. */ | |
662 | if($char === '"' || $char === "'" || $char === '<' || $char === '=') { | |
663 | $this->emitToken(array( | |
664 | 'type' => self::PARSEERROR, | |
665 | 'data' => 'invalid-character-in-attribute-name' | |
666 | )); | |
667 | } | |
668 | ||
669 | /* Anything else | |
670 | Start a new attribute in the current tag token. Set that attribute's | |
671 | name to the current input character, and its value to the empty string. | |
672 | Switch to the attribute name state. */ | |
673 | $this->token['attr'][] = array( | |
674 | 'name' => $char, | |
675 | 'value' => '' | |
676 | ); | |
677 | ||
678 | $state = 'attribute name'; | |
679 | } | |
680 | break; | |
681 | ||
682 | case 'attribute name': | |
683 | // Consume the next input character: | |
684 | $char = $this->stream->char(); | |
685 | ||
686 | // this conditional is optimized, check bottom | |
687 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
688 | /* U+0009 CHARACTER TABULATION | |
689 | U+000A LINE FEED (LF) | |
690 | U+000C FORM FEED (FF) | |
691 | U+0020 SPACE | |
692 | Switch to the after attribute name state. */ | |
693 | $state = 'after attribute name'; | |
694 | ||
695 | } elseif($char === '/') { | |
696 | /* U+002F SOLIDUS (/) | |
697 | Switch to the self-closing start tag state. */ | |
698 | $state = 'self-closing start tag'; | |
699 | ||
700 | } elseif($char === '=') { | |
701 | /* U+003D EQUALS SIGN (=) | |
702 | Switch to the before attribute value state. */ | |
703 | $state = 'before attribute value'; | |
704 | ||
705 | } elseif($char === '>') { | |
706 | /* U+003E GREATER-THAN SIGN (>) | |
707 | Emit the current tag token. Switch to the data state. */ | |
708 | $this->emitToken($this->token); | |
709 | $state = 'data'; | |
710 | ||
711 | } elseif('A' <= $char && $char <= 'Z') { | |
712 | /* U+0041 LATIN CAPITAL LETTER A through to U+005A LATIN CAPITAL LETTER Z | |
713 | Append the lowercase version of the current input | |
714 | character (add 0x0020 to the character's code point) to | |
715 | the current attribute's name. Stay in the attribute name | |
716 | state. */ | |
717 | $chars = $this->stream->charsWhile(self::UPPER_ALPHA); | |
718 | ||
719 | $last = count($this->token['attr']) - 1; | |
720 | $this->token['attr'][$last]['name'] .= strtolower($char . $chars); | |
721 | ||
722 | $state = 'attribute name'; | |
723 | ||
724 | } elseif($char === false) { | |
725 | /* EOF | |
726 | Parse error. Reconsume the EOF character in the data state. */ | |
727 | $this->emitToken(array( | |
728 | 'type' => self::PARSEERROR, | |
729 | 'data' => 'eof-in-attribute-name' | |
730 | )); | |
731 | ||
732 | $this->stream->unget(); | |
733 | $state = 'data'; | |
734 | ||
735 | } else { | |
736 | /* U+0022 QUOTATION MARK (") | |
737 | U+0027 APOSTROPHE (') | |
738 | U+003C LESS-THAN SIGN (<) | |
739 | Parse error. Treat it as per the "anything else" | |
740 | entry below. */ | |
741 | if($char === '"' || $char === "'" || $char === '<') { | |
742 | $this->emitToken(array( | |
743 | 'type' => self::PARSEERROR, | |
744 | 'data' => 'invalid-character-in-attribute-name' | |
745 | )); | |
746 | } | |
747 | ||
748 | /* Anything else | |
749 | Append the current input character to the current attribute's name. | |
750 | Stay in the attribute name state. */ | |
751 | $chars = $this->stream->charsUntil("\t\n\x0C /=>\"'" . self::UPPER_ALPHA); | |
752 | ||
753 | $last = count($this->token['attr']) - 1; | |
754 | $this->token['attr'][$last]['name'] .= $char . $chars; | |
755 | ||
756 | $state = 'attribute name'; | |
757 | } | |
758 | ||
759 | /* When the user agent leaves the attribute name state | |
760 | (and before emitting the tag token, if appropriate), the | |
761 | complete attribute's name must be compared to the other | |
762 | attributes on the same token; if there is already an | |
763 | attribute on the token with the exact same name, then this | |
764 | is a parse error and the new attribute must be dropped, along | |
765 | with the value that gets associated with it (if any). */ | |
766 | // this might be implemented in the emitToken method | |
767 | break; | |
768 | ||
769 | case 'after attribute name': | |
770 | // Consume the next input character: | |
771 | $char = $this->stream->char(); | |
772 | ||
773 | // this is an optimized conditional, check the bottom | |
774 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
775 | /* U+0009 CHARACTER TABULATION | |
776 | U+000A LINE FEED (LF) | |
777 | U+000C FORM FEED (FF) | |
778 | U+0020 SPACE | |
779 | Stay in the after attribute name state. */ | |
780 | $state = 'after attribute name'; | |
781 | ||
782 | } elseif($char === '/') { | |
783 | /* U+002F SOLIDUS (/) | |
784 | Switch to the self-closing start tag state. */ | |
785 | $state = 'self-closing start tag'; | |
786 | ||
787 | } elseif($char === '=') { | |
788 | /* U+003D EQUALS SIGN (=) | |
789 | Switch to the before attribute value state. */ | |
790 | $state = 'before attribute value'; | |
791 | ||
792 | } elseif($char === '>') { | |
793 | /* U+003E GREATER-THAN SIGN (>) | |
794 | Emit the current tag token. Switch to the data state. */ | |
795 | $this->emitToken($this->token); | |
796 | $state = 'data'; | |
797 | ||
798 | } elseif('A' <= $char && $char <= 'Z') { | |
799 | /* U+0041 LATIN CAPITAL LETTER A through to U+005A LATIN CAPITAL LETTER Z | |
800 | Start a new attribute in the current tag token. Set that | |
801 | attribute's name to the lowercase version of the current | |
802 | input character (add 0x0020 to the character's code | |
803 | point), and its value to the empty string. Switch to the | |
804 | attribute name state. */ | |
805 | $this->token['attr'][] = array( | |
806 | 'name' => strtolower($char), | |
807 | 'value' => '' | |
808 | ); | |
809 | ||
810 | $state = 'attribute name'; | |
811 | ||
812 | } elseif($char === false) { | |
813 | /* EOF | |
814 | Parse error. Reconsume the EOF character in the data state. */ | |
815 | $this->emitToken(array( | |
816 | 'type' => self::PARSEERROR, | |
817 | 'data' => 'expected-end-of-tag-but-got-eof' | |
818 | )); | |
819 | ||
820 | $this->stream->unget(); | |
821 | $state = 'data'; | |
822 | ||
823 | } else { | |
824 | /* U+0022 QUOTATION MARK (") | |
825 | U+0027 APOSTROPHE (') | |
826 | U+003C LESS-THAN SIGN(<) | |
827 | Parse error. Treat it as per the "anything else" | |
828 | entry below. */ | |
829 | if($char === '"' || $char === "'" || $char === "<") { | |
830 | $this->emitToken(array( | |
831 | 'type' => self::PARSEERROR, | |
832 | 'data' => 'invalid-character-after-attribute-name' | |
833 | )); | |
834 | } | |
835 | ||
836 | /* Anything else | |
837 | Start a new attribute in the current tag token. Set that attribute's | |
838 | name to the current input character, and its value to the empty string. | |
839 | Switch to the attribute name state. */ | |
840 | $this->token['attr'][] = array( | |
841 | 'name' => $char, | |
842 | 'value' => '' | |
843 | ); | |
844 | ||
845 | $state = 'attribute name'; | |
846 | } | |
847 | break; | |
848 | ||
849 | case 'before attribute value': | |
850 | // Consume the next input character: | |
851 | $char = $this->stream->char(); | |
852 | ||
853 | // this is an optimized conditional | |
854 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
855 | /* U+0009 CHARACTER TABULATION | |
856 | U+000A LINE FEED (LF) | |
857 | U+000C FORM FEED (FF) | |
858 | U+0020 SPACE | |
859 | Stay in the before attribute value state. */ | |
860 | $state = 'before attribute value'; | |
861 | ||
862 | } elseif($char === '"') { | |
863 | /* U+0022 QUOTATION MARK (") | |
864 | Switch to the attribute value (double-quoted) state. */ | |
865 | $state = 'attribute value (double-quoted)'; | |
866 | ||
867 | } elseif($char === '&') { | |
868 | /* U+0026 AMPERSAND (&) | |
869 | Switch to the attribute value (unquoted) state and reconsume | |
870 | this input character. */ | |
871 | $this->stream->unget(); | |
872 | $state = 'attribute value (unquoted)'; | |
873 | ||
874 | } elseif($char === '\'') { | |
875 | /* U+0027 APOSTROPHE (') | |
876 | Switch to the attribute value (single-quoted) state. */ | |
877 | $state = 'attribute value (single-quoted)'; | |
878 | ||
879 | } elseif($char === '>') { | |
880 | /* U+003E GREATER-THAN SIGN (>) | |
881 | Parse error. Emit the current tag token. Switch to the data state. */ | |
882 | $this->emitToken(array( | |
883 | 'type' => self::PARSEERROR, | |
884 | 'data' => 'expected-attribute-value-but-got-right-bracket' | |
885 | )); | |
886 | $this->emitToken($this->token); | |
887 | $state = 'data'; | |
888 | ||
889 | } elseif($char === false) { | |
890 | /* EOF | |
891 | Parse error. Reconsume the EOF character in the data state. */ | |
892 | $this->emitToken(array( | |
893 | 'type' => self::PARSEERROR, | |
894 | 'data' => 'expected-attribute-value-but-got-eof' | |
895 | )); | |
896 | $this->stream->unget(); | |
897 | $state = 'data'; | |
898 | ||
899 | } else { | |
900 | /* U+003D EQUALS SIGN (=) | |
901 | * U+003C LESS-THAN SIGN (<) | |
902 | Parse error. Treat it as per the "anything else" entry below. */ | |
903 | if($char === '=' || $char === '<') { | |
904 | $this->emitToken(array( | |
905 | 'type' => self::PARSEERROR, | |
906 | 'data' => 'equals-in-unquoted-attribute-value' | |
907 | )); | |
908 | } | |
909 | ||
910 | /* Anything else | |
911 | Append the current input character to the current attribute's value. | |
912 | Switch to the attribute value (unquoted) state. */ | |
913 | $last = count($this->token['attr']) - 1; | |
914 | $this->token['attr'][$last]['value'] .= $char; | |
915 | ||
916 | $state = 'attribute value (unquoted)'; | |
917 | } | |
918 | break; | |
919 | ||
920 | case 'attribute value (double-quoted)': | |
921 | // Consume the next input character: | |
922 | $char = $this->stream->char(); | |
923 | ||
924 | if($char === '"') { | |
925 | /* U+0022 QUOTATION MARK (") | |
926 | Switch to the after attribute value (quoted) state. */ | |
927 | $state = 'after attribute value (quoted)'; | |
928 | ||
929 | } elseif($char === '&') { | |
930 | /* U+0026 AMPERSAND (&) | |
931 | Switch to the character reference in attribute value | |
932 | state, with the additional allowed character | |
933 | being U+0022 QUOTATION MARK ("). */ | |
934 | $this->characterReferenceInAttributeValue('"'); | |
935 | ||
936 | } elseif($char === false) { | |
937 | /* EOF | |
938 | Parse error. Reconsume the EOF character in the data state. */ | |
939 | $this->emitToken(array( | |
940 | 'type' => self::PARSEERROR, | |
941 | 'data' => 'eof-in-attribute-value-double-quote' | |
942 | )); | |
943 | ||
944 | $this->stream->unget(); | |
945 | $state = 'data'; | |
946 | ||
947 | } else { | |
948 | /* Anything else | |
949 | Append the current input character to the current attribute's value. | |
950 | Stay in the attribute value (double-quoted) state. */ | |
951 | $chars = $this->stream->charsUntil('"&'); | |
952 | ||
953 | $last = count($this->token['attr']) - 1; | |
954 | $this->token['attr'][$last]['value'] .= $char . $chars; | |
955 | ||
956 | $state = 'attribute value (double-quoted)'; | |
957 | } | |
958 | break; | |
959 | ||
960 | case 'attribute value (single-quoted)': | |
961 | // Consume the next input character: | |
962 | $char = $this->stream->char(); | |
963 | ||
964 | if($char === "'") { | |
965 | /* U+0022 QUOTATION MARK (') | |
966 | Switch to the after attribute value state. */ | |
967 | $state = 'after attribute value (quoted)'; | |
968 | ||
969 | } elseif($char === '&') { | |
970 | /* U+0026 AMPERSAND (&) | |
971 | Switch to the entity in attribute value state. */ | |
972 | $this->characterReferenceInAttributeValue("'"); | |
973 | ||
974 | } elseif($char === false) { | |
975 | /* EOF | |
976 | Parse error. Reconsume the EOF character in the data state. */ | |
977 | $this->emitToken(array( | |
978 | 'type' => self::PARSEERROR, | |
979 | 'data' => 'eof-in-attribute-value-single-quote' | |
980 | )); | |
981 | ||
982 | $this->stream->unget(); | |
983 | $state = 'data'; | |
984 | ||
985 | } else { | |
986 | /* Anything else | |
987 | Append the current input character to the current attribute's value. | |
988 | Stay in the attribute value (single-quoted) state. */ | |
989 | $chars = $this->stream->charsUntil("'&"); | |
990 | ||
991 | $last = count($this->token['attr']) - 1; | |
992 | $this->token['attr'][$last]['value'] .= $char . $chars; | |
993 | ||
994 | $state = 'attribute value (single-quoted)'; | |
995 | } | |
996 | break; | |
997 | ||
998 | case 'attribute value (unquoted)': | |
999 | // Consume the next input character: | |
1000 | $char = $this->stream->char(); | |
1001 | ||
1002 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1003 | /* U+0009 CHARACTER TABULATION | |
1004 | U+000A LINE FEED (LF) | |
1005 | U+000C FORM FEED (FF) | |
1006 | U+0020 SPACE | |
1007 | Switch to the before attribute name state. */ | |
1008 | $state = 'before attribute name'; | |
1009 | ||
1010 | } elseif($char === '&') { | |
1011 | /* U+0026 AMPERSAND (&) | |
1012 | Switch to the entity in attribute value state, with the | |
1013 | additional allowed character being U+003E | |
1014 | GREATER-THAN SIGN (>). */ | |
1015 | $this->characterReferenceInAttributeValue('>'); | |
1016 | ||
1017 | } elseif($char === '>') { | |
1018 | /* U+003E GREATER-THAN SIGN (>) | |
1019 | Emit the current tag token. Switch to the data state. */ | |
1020 | $this->emitToken($this->token); | |
1021 | $state = 'data'; | |
1022 | ||
1023 | } elseif ($char === false) { | |
1024 | /* EOF | |
1025 | Parse error. Reconsume the EOF character in the data state. */ | |
1026 | $this->emitToken(array( | |
1027 | 'type' => self::PARSEERROR, | |
1028 | 'data' => 'eof-in-attribute-value-no-quotes' | |
1029 | )); | |
1030 | $this->stream->unget(); | |
1031 | $state = 'data'; | |
1032 | ||
1033 | } else { | |
1034 | /* U+0022 QUOTATION MARK (") | |
1035 | U+0027 APOSTROPHE (') | |
1036 | U+003C LESS-THAN SIGN (<) | |
1037 | U+003D EQUALS SIGN (=) | |
1038 | Parse error. Treat it as per the "anything else" | |
1039 | entry below. */ | |
1040 | if($char === '"' || $char === "'" || $char === '=' || $char == '<') { | |
1041 | $this->emitToken(array( | |
1042 | 'type' => self::PARSEERROR, | |
1043 | 'data' => 'unexpected-character-in-unquoted-attribute-value' | |
1044 | )); | |
1045 | } | |
1046 | ||
1047 | /* Anything else | |
1048 | Append the current input character to the current attribute's value. | |
1049 | Stay in the attribute value (unquoted) state. */ | |
1050 | $chars = $this->stream->charsUntil("\t\n\x0c &>\"'="); | |
1051 | ||
1052 | $last = count($this->token['attr']) - 1; | |
1053 | $this->token['attr'][$last]['value'] .= $char . $chars; | |
1054 | ||
1055 | $state = 'attribute value (unquoted)'; | |
1056 | } | |
1057 | break; | |
1058 | ||
1059 | case 'after attribute value (quoted)': | |
1060 | /* Consume the next input character: */ | |
1061 | $char = $this->stream->char(); | |
1062 | ||
1063 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1064 | /* U+0009 CHARACTER TABULATION | |
1065 | U+000A LINE FEED (LF) | |
1066 | U+000C FORM FEED (FF) | |
1067 | U+0020 SPACE | |
1068 | Switch to the before attribute name state. */ | |
1069 | $state = 'before attribute name'; | |
1070 | ||
1071 | } elseif ($char === '/') { | |
1072 | /* U+002F SOLIDUS (/) | |
1073 | Switch to the self-closing start tag state. */ | |
1074 | $state = 'self-closing start tag'; | |
1075 | ||
1076 | } elseif ($char === '>') { | |
1077 | /* U+003E GREATER-THAN SIGN (>) | |
1078 | Emit the current tag token. Switch to the data state. */ | |
1079 | $this->emitToken($this->token); | |
1080 | $state = 'data'; | |
1081 | ||
1082 | } elseif ($char === false) { | |
1083 | /* EOF | |
1084 | Parse error. Reconsume the EOF character in the data state. */ | |
1085 | $this->emitToken(array( | |
1086 | 'type' => self::PARSEERROR, | |
1087 | 'data' => 'unexpected-EOF-after-attribute-value' | |
1088 | )); | |
1089 | $this->stream->unget(); | |
1090 | $state = 'data'; | |
1091 | ||
1092 | } else { | |
1093 | /* Anything else | |
1094 | Parse error. Reconsume the character in the before attribute | |
1095 | name state. */ | |
1096 | $this->emitToken(array( | |
1097 | 'type' => self::PARSEERROR, | |
1098 | 'data' => 'unexpected-character-after-attribute-value' | |
1099 | )); | |
1100 | $this->stream->unget(); | |
1101 | $state = 'before attribute name'; | |
1102 | } | |
1103 | break; | |
1104 | ||
1105 | case 'self-closing start tag': | |
1106 | /* Consume the next input character: */ | |
1107 | $char = $this->stream->char(); | |
1108 | ||
1109 | if ($char === '>') { | |
1110 | /* U+003E GREATER-THAN SIGN (>) | |
1111 | Set the self-closing flag of the current tag token. | |
1112 | Emit the current tag token. Switch to the data state. */ | |
1113 | // not sure if this is the name we want | |
1114 | $this->token['self-closing'] = true; | |
1115 | $this->emitToken($this->token); | |
1116 | $state = 'data'; | |
1117 | ||
1118 | } elseif ($char === false) { | |
1119 | /* EOF | |
1120 | Parse error. Reconsume the EOF character in the data state. */ | |
1121 | $this->emitToken(array( | |
1122 | 'type' => self::PARSEERROR, | |
1123 | 'data' => 'unexpected-eof-after-self-closing' | |
1124 | )); | |
1125 | $this->stream->unget(); | |
1126 | $state = 'data'; | |
1127 | ||
1128 | } else { | |
1129 | /* Anything else | |
1130 | Parse error. Reconsume the character in the before attribute name state. */ | |
1131 | $this->emitToken(array( | |
1132 | 'type' => self::PARSEERROR, | |
1133 | 'data' => 'unexpected-character-after-self-closing' | |
1134 | )); | |
1135 | $this->stream->unget(); | |
1136 | $state = 'before attribute name'; | |
1137 | } | |
1138 | break; | |
1139 | ||
1140 | case 'bogus comment': | |
1141 | /* (This can only happen if the content model flag is set to the PCDATA state.) */ | |
1142 | /* Consume every character up to the first U+003E GREATER-THAN SIGN | |
1143 | character (>) or the end of the file (EOF), whichever comes first. Emit | |
1144 | a comment token whose data is the concatenation of all the characters | |
1145 | starting from and including the character that caused the state machine | |
1146 | to switch into the bogus comment state, up to and including the last | |
1147 | consumed character before the U+003E character, if any, or up to the | |
1148 | end of the file otherwise. (If the comment was started by the end of | |
1149 | the file (EOF), the token is empty.) */ | |
1150 | $this->token['data'] .= (string) $this->stream->charsUntil('>'); | |
1151 | $this->stream->char(); | |
1152 | ||
1153 | $this->emitToken($this->token); | |
1154 | ||
1155 | /* Switch to the data state. */ | |
1156 | $state = 'data'; | |
1157 | break; | |
1158 | ||
1159 | case 'markup declaration open': | |
1160 | // Consume for below | |
1161 | $hyphens = $this->stream->charsWhile('-', 2); | |
1162 | if ($hyphens === '-') { | |
1163 | $this->stream->unget(); | |
1164 | } | |
1165 | if ($hyphens !== '--') { | |
1166 | $alpha = $this->stream->charsWhile(self::ALPHA, 7); | |
1167 | } | |
1168 | ||
1169 | /* If the next two characters are both U+002D HYPHEN-MINUS (-) | |
1170 | characters, consume those two characters, create a comment token whose | |
1171 | data is the empty string, and switch to the comment state. */ | |
1172 | if($hyphens === '--') { | |
1173 | $state = 'comment start'; | |
1174 | $this->token = array( | |
1175 | 'data' => '', | |
1176 | 'type' => self::COMMENT | |
1177 | ); | |
1178 | ||
1179 | /* Otherwise if the next seven characters are a case-insensitive match | |
1180 | for the word "DOCTYPE", then consume those characters and switch to the | |
1181 | DOCTYPE state. */ | |
1182 | } elseif(strtoupper($alpha) === 'DOCTYPE') { | |
1183 | $state = 'DOCTYPE'; | |
1184 | ||
1185 | // XXX not implemented | |
1186 | /* Otherwise, if the insertion mode is "in foreign content" | |
1187 | and the current node is not an element in the HTML namespace | |
1188 | and the next seven characters are an ASCII case-sensitive | |
1189 | match for the string "[CDATA[" (the five uppercase letters | |
1190 | "CDATA" with a U+005B LEFT SQUARE BRACKET character before | |
1191 | and after), then consume those characters and switch to the | |
1192 | CDATA section state (which is unrelated to the content model | |
1193 | flag's CDATA state). */ | |
1194 | ||
1195 | /* Otherwise, is is a parse error. Switch to the bogus comment state. | |
1196 | The next character that is consumed, if any, is the first character | |
1197 | that will be in the comment. */ | |
1198 | } else { | |
1199 | $this->emitToken(array( | |
1200 | 'type' => self::PARSEERROR, | |
1201 | 'data' => 'expected-dashes-or-doctype' | |
1202 | )); | |
1203 | $this->token = array( | |
1204 | 'data' => (string) $alpha, | |
1205 | 'type' => self::COMMENT | |
1206 | ); | |
1207 | $state = 'bogus comment'; | |
1208 | } | |
1209 | break; | |
1210 | ||
1211 | case 'comment start': | |
1212 | /* Consume the next input character: */ | |
1213 | $char = $this->stream->char(); | |
1214 | ||
1215 | if ($char === '-') { | |
1216 | /* U+002D HYPHEN-MINUS (-) | |
1217 | Switch to the comment start dash state. */ | |
1218 | $state = 'comment start dash'; | |
1219 | } elseif ($char === '>') { | |
1220 | /* U+003E GREATER-THAN SIGN (>) | |
1221 | Parse error. Emit the comment token. Switch to the | |
1222 | data state. */ | |
1223 | $this->emitToken(array( | |
1224 | 'type' => self::PARSEERROR, | |
1225 | 'data' => 'incorrect-comment' | |
1226 | )); | |
1227 | $this->emitToken($this->token); | |
1228 | $state = 'data'; | |
1229 | } elseif ($char === false) { | |
1230 | /* EOF | |
1231 | Parse error. Emit the comment token. Reconsume the | |
1232 | EOF character in the data state. */ | |
1233 | $this->emitToken(array( | |
1234 | 'type' => self::PARSEERROR, | |
1235 | 'data' => 'eof-in-comment' | |
1236 | )); | |
1237 | $this->emitToken($this->token); | |
1238 | $this->stream->unget(); | |
1239 | $state = 'data'; | |
1240 | } else { | |
1241 | /* Anything else | |
1242 | Append the input character to the comment token's | |
1243 | data. Switch to the comment state. */ | |
1244 | $this->token['data'] .= $char; | |
1245 | $state = 'comment'; | |
1246 | } | |
1247 | break; | |
1248 | ||
1249 | case 'comment start dash': | |
1250 | /* Consume the next input character: */ | |
1251 | $char = $this->stream->char(); | |
1252 | if ($char === '-') { | |
1253 | /* U+002D HYPHEN-MINUS (-) | |
1254 | Switch to the comment end state */ | |
1255 | $state = 'comment end'; | |
1256 | } elseif ($char === '>') { | |
1257 | /* U+003E GREATER-THAN SIGN (>) | |
1258 | Parse error. Emit the comment token. Switch to the | |
1259 | data state. */ | |
1260 | $this->emitToken(array( | |
1261 | 'type' => self::PARSEERROR, | |
1262 | 'data' => 'incorrect-comment' | |
1263 | )); | |
1264 | $this->emitToken($this->token); | |
1265 | $state = 'data'; | |
1266 | } elseif ($char === false) { | |
1267 | /* Parse error. Emit the comment token. Reconsume the | |
1268 | EOF character in the data state. */ | |
1269 | $this->emitToken(array( | |
1270 | 'type' => self::PARSEERROR, | |
1271 | 'data' => 'eof-in-comment' | |
1272 | )); | |
1273 | $this->emitToken($this->token); | |
1274 | $this->stream->unget(); | |
1275 | $state = 'data'; | |
1276 | } else { | |
1277 | $this->token['data'] .= '-' . $char; | |
1278 | $state = 'comment'; | |
1279 | } | |
1280 | break; | |
1281 | ||
1282 | case 'comment': | |
1283 | /* Consume the next input character: */ | |
1284 | $char = $this->stream->char(); | |
1285 | ||
1286 | if($char === '-') { | |
1287 | /* U+002D HYPHEN-MINUS (-) | |
1288 | Switch to the comment end dash state */ | |
1289 | $state = 'comment end dash'; | |
1290 | ||
1291 | } elseif($char === false) { | |
1292 | /* EOF | |
1293 | Parse error. Emit the comment token. Reconsume the EOF character | |
1294 | in the data state. */ | |
1295 | $this->emitToken(array( | |
1296 | 'type' => self::PARSEERROR, | |
1297 | 'data' => 'eof-in-comment' | |
1298 | )); | |
1299 | $this->emitToken($this->token); | |
1300 | $this->stream->unget(); | |
1301 | $state = 'data'; | |
1302 | ||
1303 | } else { | |
1304 | /* Anything else | |
1305 | Append the input character to the comment token's data. Stay in | |
1306 | the comment state. */ | |
1307 | $chars = $this->stream->charsUntil('-'); | |
1308 | ||
1309 | $this->token['data'] .= $char . $chars; | |
1310 | } | |
1311 | break; | |
1312 | ||
1313 | case 'comment end dash': | |
1314 | /* Consume the next input character: */ | |
1315 | $char = $this->stream->char(); | |
1316 | ||
1317 | if($char === '-') { | |
1318 | /* U+002D HYPHEN-MINUS (-) | |
1319 | Switch to the comment end state */ | |
1320 | $state = 'comment end'; | |
1321 | ||
1322 | } elseif($char === false) { | |
1323 | /* EOF | |
1324 | Parse error. Emit the comment token. Reconsume the EOF character | |
1325 | in the data state. */ | |
1326 | $this->emitToken(array( | |
1327 | 'type' => self::PARSEERROR, | |
1328 | 'data' => 'eof-in-comment-end-dash' | |
1329 | )); | |
1330 | $this->emitToken($this->token); | |
1331 | $this->stream->unget(); | |
1332 | $state = 'data'; | |
1333 | ||
1334 | } else { | |
1335 | /* Anything else | |
1336 | Append a U+002D HYPHEN-MINUS (-) character and the input | |
1337 | character to the comment token's data. Switch to the comment state. */ | |
1338 | $this->token['data'] .= '-'.$char; | |
1339 | $state = 'comment'; | |
1340 | } | |
1341 | break; | |
1342 | ||
1343 | case 'comment end': | |
1344 | /* Consume the next input character: */ | |
1345 | $char = $this->stream->char(); | |
1346 | ||
1347 | if($char === '>') { | |
1348 | /* U+003E GREATER-THAN SIGN (>) | |
1349 | Emit the comment token. Switch to the data state. */ | |
1350 | $this->emitToken($this->token); | |
1351 | $state = 'data'; | |
1352 | ||
1353 | } elseif($char === '-') { | |
1354 | /* U+002D HYPHEN-MINUS (-) | |
1355 | Parse error. Append a U+002D HYPHEN-MINUS (-) character | |
1356 | to the comment token's data. Stay in the comment end | |
1357 | state. */ | |
1358 | $this->emitToken(array( | |
1359 | 'type' => self::PARSEERROR, | |
1360 | 'data' => 'unexpected-dash-after-double-dash-in-comment' | |
1361 | )); | |
1362 | $this->token['data'] .= '-'; | |
1363 | ||
1364 | } elseif($char === "\t" || $char === "\n" || $char === "\x0a" || $char === ' ') { | |
1365 | $this->emitToken(array( | |
1366 | 'type' => self::PARSEERROR, | |
1367 | 'data' => 'unexpected-space-after-double-dash-in-comment' | |
1368 | )); | |
1369 | $this->token['data'] .= '--' . $char; | |
1370 | $state = 'comment end space'; | |
1371 | ||
1372 | } elseif($char === '!') { | |
1373 | $this->emitToken(array( | |
1374 | 'type' => self::PARSEERROR, | |
1375 | 'data' => 'unexpected-bang-after-double-dash-in-comment' | |
1376 | )); | |
1377 | $state = 'comment end bang'; | |
1378 | ||
1379 | } elseif($char === false) { | |
1380 | /* EOF | |
1381 | Parse error. Emit the comment token. Reconsume the | |
1382 | EOF character in the data state. */ | |
1383 | $this->emitToken(array( | |
1384 | 'type' => self::PARSEERROR, | |
1385 | 'data' => 'eof-in-comment-double-dash' | |
1386 | )); | |
1387 | $this->emitToken($this->token); | |
1388 | $this->stream->unget(); | |
1389 | $state = 'data'; | |
1390 | ||
1391 | } else { | |
1392 | /* Anything else | |
1393 | Parse error. Append two U+002D HYPHEN-MINUS (-) | |
1394 | characters and the input character to the comment token's | |
1395 | data. Switch to the comment state. */ | |
1396 | $this->emitToken(array( | |
1397 | 'type' => self::PARSEERROR, | |
1398 | 'data' => 'unexpected-char-in-comment' | |
1399 | )); | |
1400 | $this->token['data'] .= '--'.$char; | |
1401 | $state = 'comment'; | |
1402 | } | |
1403 | break; | |
1404 | ||
1405 | case 'comment end bang': | |
1406 | $char = $this->stream->char(); | |
1407 | if ($char === '>') { | |
1408 | $this->emitToken($this->token); | |
1409 | $state = 'data'; | |
1410 | } elseif ($char === "-") { | |
1411 | $this->token['data'] .= '--!'; | |
1412 | $state = 'comment end dash'; | |
1413 | } elseif ($char === false) { | |
1414 | $this->emitToken(array( | |
1415 | 'type' => self::PARSEERROR, | |
1416 | 'data' => 'eof-in-comment-end-bang' | |
1417 | )); | |
1418 | $this->emitToken($this->token); | |
1419 | $this->stream->unget(); | |
1420 | $state = 'data'; | |
1421 | } else { | |
1422 | $this->token['data'] .= '--!' . $char; | |
1423 | $state = 'comment'; | |
1424 | } | |
1425 | break; | |
1426 | ||
1427 | case 'comment end space': | |
1428 | $char = $this->stream->char(); | |
1429 | if ($char === '>') { | |
1430 | $this->emitToken($this->token); | |
1431 | $state = 'data'; | |
1432 | } elseif ($char === '-') { | |
1433 | $state = 'comment end dash'; | |
1434 | } elseif ($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1435 | $this->token['data'] .= $char; | |
1436 | } elseif ($char === false) { | |
1437 | $this->emitToken(array( | |
1438 | 'type' => self::PARSEERROR, | |
1439 | 'data' => 'unexpected-eof-in-comment-end-space', | |
1440 | )); | |
1441 | $this->emitToken($this->token); | |
1442 | $this->stream->unget(); | |
1443 | $state = 'data'; | |
1444 | } else { | |
1445 | $this->token['data'] .= $char; | |
1446 | $state = 'comment'; | |
1447 | } | |
1448 | break; | |
1449 | ||
1450 | case 'DOCTYPE': | |
1451 | /* Consume the next input character: */ | |
1452 | $char = $this->stream->char(); | |
1453 | ||
1454 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1455 | /* U+0009 CHARACTER TABULATION | |
1456 | U+000A LINE FEED (LF) | |
1457 | U+000C FORM FEED (FF) | |
1458 | U+0020 SPACE | |
1459 | Switch to the before DOCTYPE name state. */ | |
1460 | $state = 'before DOCTYPE name'; | |
1461 | ||
1462 | } elseif($char === false) { | |
1463 | /* EOF | |
1464 | Parse error. Create a new DOCTYPE token. Set its | |
1465 | force-quirks flag to on. Emit the token. Reconsume the | |
1466 | EOF character in the data state. */ | |
1467 | $this->emitToken(array( | |
1468 | 'type' => self::PARSEERROR, | |
1469 | 'data' => 'need-space-after-doctype-but-got-eof' | |
1470 | )); | |
1471 | $this->emitToken(array( | |
1472 | 'name' => '', | |
1473 | 'type' => self::DOCTYPE, | |
1474 | 'force-quirks' => true, | |
1475 | 'error' => true | |
1476 | )); | |
1477 | $this->stream->unget(); | |
1478 | $state = 'data'; | |
1479 | ||
1480 | } else { | |
1481 | /* Anything else | |
1482 | Parse error. Reconsume the current character in the | |
1483 | before DOCTYPE name state. */ | |
1484 | $this->emitToken(array( | |
1485 | 'type' => self::PARSEERROR, | |
1486 | 'data' => 'need-space-after-doctype' | |
1487 | )); | |
1488 | $this->stream->unget(); | |
1489 | $state = 'before DOCTYPE name'; | |
1490 | } | |
1491 | break; | |
1492 | ||
1493 | case 'before DOCTYPE name': | |
1494 | /* Consume the next input character: */ | |
1495 | $char = $this->stream->char(); | |
1496 | ||
1497 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1498 | /* U+0009 CHARACTER TABULATION | |
1499 | U+000A LINE FEED (LF) | |
1500 | U+000C FORM FEED (FF) | |
1501 | U+0020 SPACE | |
1502 | Stay in the before DOCTYPE name state. */ | |
1503 | ||
1504 | } elseif($char === '>') { | |
1505 | /* U+003E GREATER-THAN SIGN (>) | |
1506 | Parse error. Create a new DOCTYPE token. Set its | |
1507 | force-quirks flag to on. Emit the token. Switch to the | |
1508 | data state. */ | |
1509 | $this->emitToken(array( | |
1510 | 'type' => self::PARSEERROR, | |
1511 | 'data' => 'expected-doctype-name-but-got-right-bracket' | |
1512 | )); | |
1513 | $this->emitToken(array( | |
1514 | 'name' => '', | |
1515 | 'type' => self::DOCTYPE, | |
1516 | 'force-quirks' => true, | |
1517 | 'error' => true | |
1518 | )); | |
1519 | ||
1520 | $state = 'data'; | |
1521 | ||
1522 | } elseif('A' <= $char && $char <= 'Z') { | |
1523 | /* U+0041 LATIN CAPITAL LETTER A through to U+005A LATIN CAPITAL LETTER Z | |
1524 | Create a new DOCTYPE token. Set the token's name to the | |
1525 | lowercase version of the input character (add 0x0020 to | |
1526 | the character's code point). Switch to the DOCTYPE name | |
1527 | state. */ | |
1528 | $this->token = array( | |
1529 | 'name' => strtolower($char), | |
1530 | 'type' => self::DOCTYPE, | |
1531 | 'error' => true | |
1532 | ); | |
1533 | ||
1534 | $state = 'DOCTYPE name'; | |
1535 | ||
1536 | } elseif($char === false) { | |
1537 | /* EOF | |
1538 | Parse error. Create a new DOCTYPE token. Set its | |
1539 | force-quirks flag to on. Emit the token. Reconsume the | |
1540 | EOF character in the data state. */ | |
1541 | $this->emitToken(array( | |
1542 | 'type' => self::PARSEERROR, | |
1543 | 'data' => 'expected-doctype-name-but-got-eof' | |
1544 | )); | |
1545 | $this->emitToken(array( | |
1546 | 'name' => '', | |
1547 | 'type' => self::DOCTYPE, | |
1548 | 'force-quirks' => true, | |
1549 | 'error' => true | |
1550 | )); | |
1551 | ||
1552 | $this->stream->unget(); | |
1553 | $state = 'data'; | |
1554 | ||
1555 | } else { | |
1556 | /* Anything else | |
1557 | Create a new DOCTYPE token. Set the token's name to the | |
1558 | current input character. Switch to the DOCTYPE name state. */ | |
1559 | $this->token = array( | |
1560 | 'name' => $char, | |
1561 | 'type' => self::DOCTYPE, | |
1562 | 'error' => true | |
1563 | ); | |
1564 | ||
1565 | $state = 'DOCTYPE name'; | |
1566 | } | |
1567 | break; | |
1568 | ||
1569 | case 'DOCTYPE name': | |
1570 | /* Consume the next input character: */ | |
1571 | $char = $this->stream->char(); | |
1572 | ||
1573 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1574 | /* U+0009 CHARACTER TABULATION | |
1575 | U+000A LINE FEED (LF) | |
1576 | U+000C FORM FEED (FF) | |
1577 | U+0020 SPACE | |
1578 | Switch to the after DOCTYPE name state. */ | |
1579 | $state = 'after DOCTYPE name'; | |
1580 | ||
1581 | } elseif($char === '>') { | |
1582 | /* U+003E GREATER-THAN SIGN (>) | |
1583 | Emit the current DOCTYPE token. Switch to the data state. */ | |
1584 | $this->emitToken($this->token); | |
1585 | $state = 'data'; | |
1586 | ||
1587 | } elseif('A' <= $char && $char <= 'Z') { | |
1588 | /* U+0041 LATIN CAPITAL LETTER A through to U+005A LATIN CAPITAL LETTER Z | |
1589 | Append the lowercase version of the input character | |
1590 | (add 0x0020 to the character's code point) to the current | |
1591 | DOCTYPE token's name. Stay in the DOCTYPE name state. */ | |
1592 | $this->token['name'] .= strtolower($char); | |
1593 | ||
1594 | } elseif($char === false) { | |
1595 | /* EOF | |
1596 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1597 | to on. Emit that DOCTYPE token. Reconsume the EOF | |
1598 | character in the data state. */ | |
1599 | $this->emitToken(array( | |
1600 | 'type' => self::PARSEERROR, | |
1601 | 'data' => 'eof-in-doctype-name' | |
1602 | )); | |
1603 | $this->token['force-quirks'] = true; | |
1604 | $this->emitToken($this->token); | |
1605 | $this->stream->unget(); | |
1606 | $state = 'data'; | |
1607 | ||
1608 | } else { | |
1609 | /* Anything else | |
1610 | Append the current input character to the current | |
1611 | DOCTYPE token's name. Stay in the DOCTYPE name state. */ | |
1612 | $this->token['name'] .= $char; | |
1613 | } | |
1614 | ||
1615 | // XXX this is probably some sort of quirks mode designation, | |
1616 | // check tree-builder to be sure. In general 'error' needs | |
1617 | // to be specc'ified, this probably means removing it at the end | |
1618 | $this->token['error'] = ($this->token['name'] === 'HTML') | |
1619 | ? false | |
1620 | : true; | |
1621 | break; | |
1622 | ||
1623 | case 'after DOCTYPE name': | |
1624 | /* Consume the next input character: */ | |
1625 | $char = $this->stream->char(); | |
1626 | ||
1627 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1628 | /* U+0009 CHARACTER TABULATION | |
1629 | U+000A LINE FEED (LF) | |
1630 | U+000C FORM FEED (FF) | |
1631 | U+0020 SPACE | |
1632 | Stay in the after DOCTYPE name state. */ | |
1633 | ||
1634 | } elseif($char === '>') { | |
1635 | /* U+003E GREATER-THAN SIGN (>) | |
1636 | Emit the current DOCTYPE token. Switch to the data state. */ | |
1637 | $this->emitToken($this->token); | |
1638 | $state = 'data'; | |
1639 | ||
1640 | } elseif($char === false) { | |
1641 | /* EOF | |
1642 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1643 | to on. Emit that DOCTYPE token. Reconsume the EOF | |
1644 | character in the data state. */ | |
1645 | $this->emitToken(array( | |
1646 | 'type' => self::PARSEERROR, | |
1647 | 'data' => 'eof-in-doctype' | |
1648 | )); | |
1649 | $this->token['force-quirks'] = true; | |
1650 | $this->emitToken($this->token); | |
1651 | $this->stream->unget(); | |
1652 | $state = 'data'; | |
1653 | ||
1654 | } else { | |
1655 | /* Anything else */ | |
1656 | ||
1657 | $nextSix = strtoupper($char . $this->stream->charsWhile(self::ALPHA, 5)); | |
1658 | if ($nextSix === 'PUBLIC') { | |
1659 | /* If the next six characters are an ASCII | |
1660 | case-insensitive match for the word "PUBLIC", then | |
1661 | consume those characters and switch to the before | |
1662 | DOCTYPE public identifier state. */ | |
1663 | $state = 'before DOCTYPE public identifier'; | |
1664 | ||
1665 | } elseif ($nextSix === 'SYSTEM') { | |
1666 | /* Otherwise, if the next six characters are an ASCII | |
1667 | case-insensitive match for the word "SYSTEM", then | |
1668 | consume those characters and switch to the before | |
1669 | DOCTYPE system identifier state. */ | |
1670 | $state = 'before DOCTYPE system identifier'; | |
1671 | ||
1672 | } else { | |
1673 | /* Otherwise, this is the parse error. Set the DOCTYPE | |
1674 | token's force-quirks flag to on. Switch to the bogus | |
1675 | DOCTYPE state. */ | |
1676 | $this->emitToken(array( | |
1677 | 'type' => self::PARSEERROR, | |
1678 | 'data' => 'expected-space-or-right-bracket-in-doctype' | |
1679 | )); | |
1680 | $this->token['force-quirks'] = true; | |
1681 | $this->token['error'] = true; | |
1682 | $state = 'bogus DOCTYPE'; | |
1683 | } | |
1684 | } | |
1685 | break; | |
1686 | ||
1687 | case 'before DOCTYPE public identifier': | |
1688 | /* Consume the next input character: */ | |
1689 | $char = $this->stream->char(); | |
1690 | ||
1691 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1692 | /* U+0009 CHARACTER TABULATION | |
1693 | U+000A LINE FEED (LF) | |
1694 | U+000C FORM FEED (FF) | |
1695 | U+0020 SPACE | |
1696 | Stay in the before DOCTYPE public identifier state. */ | |
1697 | } elseif ($char === '"') { | |
1698 | /* U+0022 QUOTATION MARK (") | |
1699 | Set the DOCTYPE token's public identifier to the empty | |
1700 | string (not missing), then switch to the DOCTYPE public | |
1701 | identifier (double-quoted) state. */ | |
1702 | $this->token['public'] = ''; | |
1703 | $state = 'DOCTYPE public identifier (double-quoted)'; | |
1704 | } elseif ($char === "'") { | |
1705 | /* U+0027 APOSTROPHE (') | |
1706 | Set the DOCTYPE token's public identifier to the empty | |
1707 | string (not missing), then switch to the DOCTYPE public | |
1708 | identifier (single-quoted) state. */ | |
1709 | $this->token['public'] = ''; | |
1710 | $state = 'DOCTYPE public identifier (single-quoted)'; | |
1711 | } elseif ($char === '>') { | |
1712 | /* Parse error. Set the DOCTYPE token's force-quirks flag | |
1713 | to on. Emit that DOCTYPE token. Switch to the data state. */ | |
1714 | $this->emitToken(array( | |
1715 | 'type' => self::PARSEERROR, | |
1716 | 'data' => 'unexpected-end-of-doctype' | |
1717 | )); | |
1718 | $this->token['force-quirks'] = true; | |
1719 | $this->emitToken($this->token); | |
1720 | $state = 'data'; | |
1721 | } elseif ($char === false) { | |
1722 | /* Parse error. Set the DOCTYPE token's force-quirks | |
1723 | flag to on. Emit that DOCTYPE token. Reconsume the EOF | |
1724 | character in the data state. */ | |
1725 | $this->emitToken(array( | |
1726 | 'type' => self::PARSEERROR, | |
1727 | 'data' => 'eof-in-doctype' | |
1728 | )); | |
1729 | $this->token['force-quirks'] = true; | |
1730 | $this->emitToken($this->token); | |
1731 | $this->stream->unget(); | |
1732 | $state = 'data'; | |
1733 | } else { | |
1734 | /* Parse error. Set the DOCTYPE token's force-quirks flag | |
1735 | to on. Switch to the bogus DOCTYPE state. */ | |
1736 | $this->emitToken(array( | |
1737 | 'type' => self::PARSEERROR, | |
1738 | 'data' => 'unexpected-char-in-doctype' | |
1739 | )); | |
1740 | $this->token['force-quirks'] = true; | |
1741 | $state = 'bogus DOCTYPE'; | |
1742 | } | |
1743 | break; | |
1744 | ||
1745 | case 'DOCTYPE public identifier (double-quoted)': | |
1746 | /* Consume the next input character: */ | |
1747 | $char = $this->stream->char(); | |
1748 | ||
1749 | if ($char === '"') { | |
1750 | /* U+0022 QUOTATION MARK (") | |
1751 | Switch to the after DOCTYPE public identifier state. */ | |
1752 | $state = 'after DOCTYPE public identifier'; | |
1753 | } elseif ($char === '>') { | |
1754 | /* U+003E GREATER-THAN SIGN (>) | |
1755 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1756 | to on. Emit that DOCTYPE token. Switch to the data state. */ | |
1757 | $this->emitToken(array( | |
1758 | 'type' => self::PARSEERROR, | |
1759 | 'data' => 'unexpected-end-of-doctype' | |
1760 | )); | |
1761 | $this->token['force-quirks'] = true; | |
1762 | $this->emitToken($this->token); | |
1763 | $state = 'data'; | |
1764 | } elseif ($char === false) { | |
1765 | /* EOF | |
1766 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1767 | to on. Emit that DOCTYPE token. Reconsume the EOF | |
1768 | character in the data state. */ | |
1769 | $this->emitToken(array( | |
1770 | 'type' => self::PARSEERROR, | |
1771 | 'data' => 'eof-in-doctype' | |
1772 | )); | |
1773 | $this->token['force-quirks'] = true; | |
1774 | $this->emitToken($this->token); | |
1775 | $this->stream->unget(); | |
1776 | $state = 'data'; | |
1777 | } else { | |
1778 | /* Anything else | |
1779 | Append the current input character to the current | |
1780 | DOCTYPE token's public identifier. Stay in the DOCTYPE | |
1781 | public identifier (double-quoted) state. */ | |
1782 | $this->token['public'] .= $char; | |
1783 | } | |
1784 | break; | |
1785 | ||
1786 | case 'DOCTYPE public identifier (single-quoted)': | |
1787 | /* Consume the next input character: */ | |
1788 | $char = $this->stream->char(); | |
1789 | ||
1790 | if ($char === "'") { | |
1791 | /* U+0027 APOSTROPHE (') | |
1792 | Switch to the after DOCTYPE public identifier state. */ | |
1793 | $state = 'after DOCTYPE public identifier'; | |
1794 | } elseif ($char === '>') { | |
1795 | /* U+003E GREATER-THAN SIGN (>) | |
1796 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1797 | to on. Emit that DOCTYPE token. Switch to the data state. */ | |
1798 | $this->emitToken(array( | |
1799 | 'type' => self::PARSEERROR, | |
1800 | 'data' => 'unexpected-end-of-doctype' | |
1801 | )); | |
1802 | $this->token['force-quirks'] = true; | |
1803 | $this->emitToken($this->token); | |
1804 | $state = 'data'; | |
1805 | } elseif ($char === false) { | |
1806 | /* EOF | |
1807 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1808 | to on. Emit that DOCTYPE token. Reconsume the EOF | |
1809 | character in the data state. */ | |
1810 | $this->emitToken(array( | |
1811 | 'type' => self::PARSEERROR, | |
1812 | 'data' => 'eof-in-doctype' | |
1813 | )); | |
1814 | $this->token['force-quirks'] = true; | |
1815 | $this->emitToken($this->token); | |
1816 | $this->stream->unget(); | |
1817 | $state = 'data'; | |
1818 | } else { | |
1819 | /* Anything else | |
1820 | Append the current input character to the current | |
1821 | DOCTYPE token's public identifier. Stay in the DOCTYPE | |
1822 | public identifier (double-quoted) state. */ | |
1823 | $this->token['public'] .= $char; | |
1824 | } | |
1825 | break; | |
1826 | ||
1827 | case 'after DOCTYPE public identifier': | |
1828 | /* Consume the next input character: */ | |
1829 | $char = $this->stream->char(); | |
1830 | ||
1831 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1832 | /* U+0009 CHARACTER TABULATION | |
1833 | U+000A LINE FEED (LF) | |
1834 | U+000C FORM FEED (FF) | |
1835 | U+0020 SPACE | |
1836 | Stay in the after DOCTYPE public identifier state. */ | |
1837 | } elseif ($char === '"') { | |
1838 | /* U+0022 QUOTATION MARK (") | |
1839 | Set the DOCTYPE token's system identifier to the | |
1840 | empty string (not missing), then switch to the DOCTYPE | |
1841 | system identifier (double-quoted) state. */ | |
1842 | $this->token['system'] = ''; | |
1843 | $state = 'DOCTYPE system identifier (double-quoted)'; | |
1844 | } elseif ($char === "'") { | |
1845 | /* U+0027 APOSTROPHE (') | |
1846 | Set the DOCTYPE token's system identifier to the | |
1847 | empty string (not missing), then switch to the DOCTYPE | |
1848 | system identifier (single-quoted) state. */ | |
1849 | $this->token['system'] = ''; | |
1850 | $state = 'DOCTYPE system identifier (single-quoted)'; | |
1851 | } elseif ($char === '>') { | |
1852 | /* U+003E GREATER-THAN SIGN (>) | |
1853 | Emit the current DOCTYPE token. Switch to the data state. */ | |
1854 | $this->emitToken($this->token); | |
1855 | $state = 'data'; | |
1856 | } elseif ($char === false) { | |
1857 | /* Parse error. Set the DOCTYPE token's force-quirks | |
1858 | flag to on. Emit that DOCTYPE token. Reconsume the EOF | |
1859 | character in the data state. */ | |
1860 | $this->emitToken(array( | |
1861 | 'type' => self::PARSEERROR, | |
1862 | 'data' => 'eof-in-doctype' | |
1863 | )); | |
1864 | $this->token['force-quirks'] = true; | |
1865 | $this->emitToken($this->token); | |
1866 | $this->stream->unget(); | |
1867 | $state = 'data'; | |
1868 | } else { | |
1869 | /* Anything else | |
1870 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1871 | to on. Switch to the bogus DOCTYPE state. */ | |
1872 | $this->emitToken(array( | |
1873 | 'type' => self::PARSEERROR, | |
1874 | 'data' => 'unexpected-char-in-doctype' | |
1875 | )); | |
1876 | $this->token['force-quirks'] = true; | |
1877 | $state = 'bogus DOCTYPE'; | |
1878 | } | |
1879 | break; | |
1880 | ||
1881 | case 'before DOCTYPE system identifier': | |
1882 | /* Consume the next input character: */ | |
1883 | $char = $this->stream->char(); | |
1884 | ||
1885 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
1886 | /* U+0009 CHARACTER TABULATION | |
1887 | U+000A LINE FEED (LF) | |
1888 | U+000C FORM FEED (FF) | |
1889 | U+0020 SPACE | |
1890 | Stay in the before DOCTYPE system identifier state. */ | |
1891 | } elseif ($char === '"') { | |
1892 | /* U+0022 QUOTATION MARK (") | |
1893 | Set the DOCTYPE token's system identifier to the empty | |
1894 | string (not missing), then switch to the DOCTYPE system | |
1895 | identifier (double-quoted) state. */ | |
1896 | $this->token['system'] = ''; | |
1897 | $state = 'DOCTYPE system identifier (double-quoted)'; | |
1898 | } elseif ($char === "'") { | |
1899 | /* U+0027 APOSTROPHE (') | |
1900 | Set the DOCTYPE token's system identifier to the empty | |
1901 | string (not missing), then switch to the DOCTYPE system | |
1902 | identifier (single-quoted) state. */ | |
1903 | $this->token['system'] = ''; | |
1904 | $state = 'DOCTYPE system identifier (single-quoted)'; | |
1905 | } elseif ($char === '>') { | |
1906 | /* Parse error. Set the DOCTYPE token's force-quirks flag | |
1907 | to on. Emit that DOCTYPE token. Switch to the data state. */ | |
1908 | $this->emitToken(array( | |
1909 | 'type' => self::PARSEERROR, | |
1910 | 'data' => 'unexpected-char-in-doctype' | |
1911 | )); | |
1912 | $this->token['force-quirks'] = true; | |
1913 | $this->emitToken($this->token); | |
1914 | $state = 'data'; | |
1915 | } elseif ($char === false) { | |
1916 | /* Parse error. Set the DOCTYPE token's force-quirks | |
1917 | flag to on. Emit that DOCTYPE token. Reconsume the EOF | |
1918 | character in the data state. */ | |
1919 | $this->emitToken(array( | |
1920 | 'type' => self::PARSEERROR, | |
1921 | 'data' => 'eof-in-doctype' | |
1922 | )); | |
1923 | $this->token['force-quirks'] = true; | |
1924 | $this->emitToken($this->token); | |
1925 | $this->stream->unget(); | |
1926 | $state = 'data'; | |
1927 | } else { | |
1928 | /* Parse error. Set the DOCTYPE token's force-quirks flag | |
1929 | to on. Switch to the bogus DOCTYPE state. */ | |
1930 | $this->emitToken(array( | |
1931 | 'type' => self::PARSEERROR, | |
1932 | 'data' => 'unexpected-char-in-doctype' | |
1933 | )); | |
1934 | $this->token['force-quirks'] = true; | |
1935 | $state = 'bogus DOCTYPE'; | |
1936 | } | |
1937 | break; | |
1938 | ||
1939 | case 'DOCTYPE system identifier (double-quoted)': | |
1940 | /* Consume the next input character: */ | |
1941 | $char = $this->stream->char(); | |
1942 | ||
1943 | if ($char === '"') { | |
1944 | /* U+0022 QUOTATION MARK (") | |
1945 | Switch to the after DOCTYPE system identifier state. */ | |
1946 | $state = 'after DOCTYPE system identifier'; | |
1947 | } elseif ($char === '>') { | |
1948 | /* U+003E GREATER-THAN SIGN (>) | |
1949 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1950 | to on. Emit that DOCTYPE token. Switch to the data state. */ | |
1951 | $this->emitToken(array( | |
1952 | 'type' => self::PARSEERROR, | |
1953 | 'data' => 'unexpected-end-of-doctype' | |
1954 | )); | |
1955 | $this->token['force-quirks'] = true; | |
1956 | $this->emitToken($this->token); | |
1957 | $state = 'data'; | |
1958 | } elseif ($char === false) { | |
1959 | /* EOF | |
1960 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1961 | to on. Emit that DOCTYPE token. Reconsume the EOF | |
1962 | character in the data state. */ | |
1963 | $this->emitToken(array( | |
1964 | 'type' => self::PARSEERROR, | |
1965 | 'data' => 'eof-in-doctype' | |
1966 | )); | |
1967 | $this->token['force-quirks'] = true; | |
1968 | $this->emitToken($this->token); | |
1969 | $this->stream->unget(); | |
1970 | $state = 'data'; | |
1971 | } else { | |
1972 | /* Anything else | |
1973 | Append the current input character to the current | |
1974 | DOCTYPE token's system identifier. Stay in the DOCTYPE | |
1975 | system identifier (double-quoted) state. */ | |
1976 | $this->token['system'] .= $char; | |
1977 | } | |
1978 | break; | |
1979 | ||
1980 | case 'DOCTYPE system identifier (single-quoted)': | |
1981 | /* Consume the next input character: */ | |
1982 | $char = $this->stream->char(); | |
1983 | ||
1984 | if ($char === "'") { | |
1985 | /* U+0027 APOSTROPHE (') | |
1986 | Switch to the after DOCTYPE system identifier state. */ | |
1987 | $state = 'after DOCTYPE system identifier'; | |
1988 | } elseif ($char === '>') { | |
1989 | /* U+003E GREATER-THAN SIGN (>) | |
1990 | Parse error. Set the DOCTYPE token's force-quirks flag | |
1991 | to on. Emit that DOCTYPE token. Switch to the data state. */ | |
1992 | $this->emitToken(array( | |
1993 | 'type' => self::PARSEERROR, | |
1994 | 'data' => 'unexpected-end-of-doctype' | |
1995 | )); | |
1996 | $this->token['force-quirks'] = true; | |
1997 | $this->emitToken($this->token); | |
1998 | $state = 'data'; | |
1999 | } elseif ($char === false) { | |
2000 | /* EOF | |
2001 | Parse error. Set the DOCTYPE token's force-quirks flag | |
2002 | to on. Emit that DOCTYPE token. Reconsume the EOF | |
2003 | character in the data state. */ | |
2004 | $this->emitToken(array( | |
2005 | 'type' => self::PARSEERROR, | |
2006 | 'data' => 'eof-in-doctype' | |
2007 | )); | |
2008 | $this->token['force-quirks'] = true; | |
2009 | $this->emitToken($this->token); | |
2010 | $this->stream->unget(); | |
2011 | $state = 'data'; | |
2012 | } else { | |
2013 | /* Anything else | |
2014 | Append the current input character to the current | |
2015 | DOCTYPE token's system identifier. Stay in the DOCTYPE | |
2016 | system identifier (double-quoted) state. */ | |
2017 | $this->token['system'] .= $char; | |
2018 | } | |
2019 | break; | |
2020 | ||
2021 | case 'after DOCTYPE system identifier': | |
2022 | /* Consume the next input character: */ | |
2023 | $char = $this->stream->char(); | |
2024 | ||
2025 | if($char === "\t" || $char === "\n" || $char === "\x0c" || $char === ' ') { | |
2026 | /* U+0009 CHARACTER TABULATION | |
2027 | U+000A LINE FEED (LF) | |
2028 | U+000C FORM FEED (FF) | |
2029 | U+0020 SPACE | |
2030 | Stay in the after DOCTYPE system identifier state. */ | |
2031 | } elseif ($char === '>') { | |
2032 | /* U+003E GREATER-THAN SIGN (>) | |
2033 | Emit the current DOCTYPE token. Switch to the data state. */ | |
2034 | $this->emitToken($this->token); | |
2035 | $state = 'data'; | |
2036 | } elseif ($char === false) { | |
2037 | /* Parse error. Set the DOCTYPE token's force-quirks | |
2038 | flag to on. Emit that DOCTYPE token. Reconsume the EOF | |
2039 | character in the data state. */ | |
2040 | $this->emitToken(array( | |
2041 | 'type' => self::PARSEERROR, | |
2042 | 'data' => 'eof-in-doctype' | |
2043 | )); | |
2044 | $this->token['force-quirks'] = true; | |
2045 | $this->emitToken($this->token); | |
2046 | $this->stream->unget(); | |
2047 | $state = 'data'; | |
2048 | } else { | |
2049 | /* Anything else | |
2050 | Parse error. Switch to the bogus DOCTYPE state. | |
2051 | (This does not set the DOCTYPE token's force-quirks | |
2052 | flag to on.) */ | |
2053 | $this->emitToken(array( | |
2054 | 'type' => self::PARSEERROR, | |
2055 | 'data' => 'unexpected-char-in-doctype' | |
2056 | )); | |
2057 | $state = 'bogus DOCTYPE'; | |
2058 | } | |
2059 | break; | |
2060 | ||
2061 | case 'bogus DOCTYPE': | |
2062 | /* Consume the next input character: */ | |
2063 | $char = $this->stream->char(); | |
2064 | ||
2065 | if ($char === '>') { | |
2066 | /* U+003E GREATER-THAN SIGN (>) | |
2067 | Emit the DOCTYPE token. Switch to the data state. */ | |
2068 | $this->emitToken($this->token); | |
2069 | $state = 'data'; | |
2070 | ||
2071 | } elseif($char === false) { | |
2072 | /* EOF | |
2073 | Emit the DOCTYPE token. Reconsume the EOF character in | |
2074 | the data state. */ | |
2075 | $this->emitToken($this->token); | |
2076 | $this->stream->unget(); | |
2077 | $state = 'data'; | |
2078 | ||
2079 | } else { | |
2080 | /* Anything else | |
2081 | Stay in the bogus DOCTYPE state. */ | |
2082 | } | |
2083 | break; | |
2084 | ||
2085 | // case 'cdataSection': | |
2086 | ||
2087 | } | |
2088 | } | |
2089 | } | |
2090 | ||
2091 | /** | |
2092 | * Returns a serialized representation of the tree. | |
2093 | */ | |
2094 | public function save() { | |
2095 | return $this->tree->save(); | |
2096 | } | |
2097 | ||
2098 | /** | |
2099 | * Returns the input stream. | |
2100 | */ | |
2101 | public function stream() { | |
2102 | return $this->stream; | |
2103 | } | |
2104 | ||
2105 | private function consumeCharacterReference($allowed = false, $inattr = false) { | |
2106 | // This goes quite far against spec, and is far closer to the Python | |
2107 | // impl., mainly because we don't do the large unconsuming the spec | |
2108 | // requires. | |
2109 | ||
2110 | // All consumed characters. | |
2111 | $chars = $this->stream->char(); | |
2112 | ||
2113 | /* This section defines how to consume a character | |
2114 | reference. This definition is used when parsing character | |
2115 | references in text and in attributes. | |
2116 | ||
2117 | The behavior depends on the identity of the next character | |
2118 | (the one immediately after the U+0026 AMPERSAND character): */ | |
2119 | ||
2120 | if ( | |
2121 | $chars[0] === "\x09" || | |
2122 | $chars[0] === "\x0A" || | |
2123 | $chars[0] === "\x0C" || | |
2124 | $chars[0] === "\x20" || | |
2125 | $chars[0] === '<' || | |
2126 | $chars[0] === '&' || | |
2127 | $chars === false || | |
2128 | $chars[0] === $allowed | |
2129 | ) { | |
2130 | /* U+0009 CHARACTER TABULATION | |
2131 | U+000A LINE FEED (LF) | |
2132 | U+000C FORM FEED (FF) | |
2133 | U+0020 SPACE | |
2134 | U+003C LESS-THAN SIGN | |
2135 | U+0026 AMPERSAND | |
2136 | EOF | |
2137 | The additional allowed character, if there is one | |
2138 | Not a character reference. No characters are consumed, | |
2139 | and nothing is returned. (This is not an error, either.) */ | |
2140 | // We already consumed, so unconsume. | |
2141 | $this->stream->unget(); | |
2142 | return '&'; | |
2143 | } elseif ($chars[0] === '#') { | |
2144 | /* Consume the U+0023 NUMBER SIGN. */ | |
2145 | // Um, yeah, we already did that. | |
2146 | /* The behavior further depends on the character after | |
2147 | the U+0023 NUMBER SIGN: */ | |
2148 | $chars .= $this->stream->char(); | |
2149 | if (isset($chars[1]) && ($chars[1] === 'x' || $chars[1] === 'X')) { | |
2150 | /* U+0078 LATIN SMALL LETTER X | |
2151 | U+0058 LATIN CAPITAL LETTER X */ | |
2152 | /* Consume the X. */ | |
2153 | // Um, yeah, we already did that. | |
2154 | /* Follow the steps below, but using the range of | |
2155 | characters U+0030 DIGIT ZERO through to U+0039 DIGIT | |
2156 | NINE, U+0061 LATIN SMALL LETTER A through to U+0066 | |
2157 | LATIN SMALL LETTER F, and U+0041 LATIN CAPITAL LETTER | |
2158 | A, through to U+0046 LATIN CAPITAL LETTER F (in other | |
2159 | words, 0123456789, ABCDEF, abcdef). */ | |
2160 | $char_class = self::HEX; | |
2161 | /* When it comes to interpreting the | |
2162 | number, interpret it as a hexadecimal number. */ | |
2163 | $hex = true; | |
2164 | } else { | |
2165 | /* Anything else */ | |
2166 | // Unconsume because we shouldn't have consumed this. | |
2167 | $chars = $chars[0]; | |
2168 | $this->stream->unget(); | |
2169 | /* Follow the steps below, but using the range of | |
2170 | characters U+0030 DIGIT ZERO through to U+0039 DIGIT | |
2171 | NINE (i.e. just 0123456789). */ | |
2172 | $char_class = self::DIGIT; | |
2173 | /* When it comes to interpreting the number, | |
2174 | interpret it as a decimal number. */ | |
2175 | $hex = false; | |
2176 | } | |
2177 | ||
2178 | /* Consume as many characters as match the range of characters given above. */ | |
2179 | $consumed = $this->stream->charsWhile($char_class); | |
2180 | if ($consumed === '' || $consumed === false) { | |
2181 | /* If no characters match the range, then don't consume | |
2182 | any characters (and unconsume the U+0023 NUMBER SIGN | |
2183 | character and, if appropriate, the X character). This | |
2184 | is a parse error; nothing is returned. */ | |
2185 | $this->emitToken(array( | |
2186 | 'type' => self::PARSEERROR, | |
2187 | 'data' => 'expected-numeric-entity' | |
2188 | )); | |
2189 | return '&' . $chars; | |
2190 | } else { | |
2191 | /* Otherwise, if the next character is a U+003B SEMICOLON, | |
2192 | consume that too. If it isn't, there is a parse error. */ | |
2193 | if ($this->stream->char() !== ';') { | |
2194 | $this->stream->unget(); | |
2195 | $this->emitToken(array( | |
2196 | 'type' => self::PARSEERROR, | |
2197 | 'data' => 'numeric-entity-without-semicolon' | |
2198 | )); | |
2199 | } | |
2200 | ||
2201 | /* If one or more characters match the range, then take | |
2202 | them all and interpret the string of characters as a number | |
2203 | (either hexadecimal or decimal as appropriate). */ | |
2204 | $codepoint = $hex ? hexdec($consumed) : (int) $consumed; | |
2205 | ||
2206 | /* If that number is one of the numbers in the first column | |
2207 | of the following table, then this is a parse error. Find the | |
2208 | row with that number in the first column, and return a | |
2209 | character token for the Unicode character given in the | |
2210 | second column of that row. */ | |
2211 | $new_codepoint = HTML5_Data::getRealCodepoint($codepoint); | |
2212 | if ($new_codepoint) { | |
2213 | $this->emitToken(array( | |
2214 | 'type' => self::PARSEERROR, | |
2215 | 'data' => 'illegal-windows-1252-entity' | |
2216 | )); | |
2217 | return HTML5_Data::utf8chr($new_codepoint); | |
2218 | } else { | |
2219 | /* Otherwise, if the number is greater than 0x10FFFF, then | |
2220 | * this is a parse error. Return a U+FFFD REPLACEMENT | |
2221 | * CHARACTER. */ | |
2222 | if ($codepoint > 0x10FFFF) { | |
2223 | $this->emitToken(array( | |
2224 | 'type' => self::PARSEERROR, | |
2225 | 'data' => 'overlong-character-entity' // XXX probably not correct | |
2226 | )); | |
2227 | return "\xEF\xBF\xBD"; | |
2228 | } | |
2229 | /* Otherwise, return a character token for the Unicode | |
2230 | * character whose code point is that number. If the | |
2231 | * number is in the range 0x0001 to 0x0008, 0x000E to | |
2232 | * 0x001F, 0x007F to 0x009F, 0xD800 to 0xDFFF, 0xFDD0 to | |
2233 | * 0xFDEF, or is one of 0x000B, 0xFFFE, 0xFFFF, 0x1FFFE, | |
2234 | * 0x1FFFF, 0x2FFFE, 0x2FFFF, 0x3FFFE, 0x3FFFF, 0x4FFFE, | |
2235 | * 0x4FFFF, 0x5FFFE, 0x5FFFF, 0x6FFFE, 0x6FFFF, 0x7FFFE, | |
2236 | * 0x7FFFF, 0x8FFFE, 0x8FFFF, 0x9FFFE, 0x9FFFF, 0xAFFFE, | |
2237 | * 0xAFFFF, 0xBFFFE, 0xBFFFF, 0xCFFFE, 0xCFFFF, 0xDFFFE, | |
2238 | * 0xDFFFF, 0xEFFFE, 0xEFFFF, 0xFFFFE, 0xFFFFF, 0x10FFFE, | |
2239 | * or 0x10FFFF, then this is a parse error. */ | |
2240 | // && has higher precedence than || | |
2241 | if ( | |
2242 | $codepoint >= 0x0000 && $codepoint <= 0x0008 || | |
2243 | $codepoint === 0x000B || | |
2244 | $codepoint >= 0x000E && $codepoint <= 0x001F || | |
2245 | $codepoint >= 0x007F && $codepoint <= 0x009F || | |
2246 | $codepoint >= 0xD800 && $codepoint <= 0xDFFF || | |
2247 | $codepoint >= 0xFDD0 && $codepoint <= 0xFDEF || | |
2248 | ($codepoint & 0xFFFE) === 0xFFFE || | |
2249 | $codepoint == 0x10FFFF || $codepoint == 0x10FFFE | |
2250 | ) { | |
2251 | $this->emitToken(array( | |
2252 | 'type' => self::PARSEERROR, | |
2253 | 'data' => 'illegal-codepoint-for-numeric-entity' | |
2254 | )); | |
2255 | } | |
2256 | return HTML5_Data::utf8chr($codepoint); | |
2257 | } | |
2258 | } | |
2259 | ||
2260 | } else { | |
2261 | /* Anything else */ | |
2262 | ||
2263 | /* Consume the maximum number of characters possible, | |
2264 | with the consumed characters matching one of the | |
2265 | identifiers in the first column of the named character | |
2266 | references table (in a case-sensitive manner). */ | |
2267 | // What we actually do here is consume as much as we can while it | |
2268 | // matches the start of one of the identifiers in the first column. | |
2269 | ||
2270 | $refs = HTML5_Data::getNamedCharacterReferences(); | |
2271 | ||
2272 | // Get the longest string which is the start of an identifier | |
2273 | // ($chars) as well as the longest identifier which matches ($id) | |
2274 | // and its codepoint ($codepoint). | |
2275 | $codepoint = false; | |
2276 | $char = $chars; | |
2277 | while ($char !== false && isset($refs[$char])) { | |
2278 | $refs = $refs[$char]; | |
2279 | if (isset($refs['codepoint'])) { | |
2280 | $id = $chars; | |
2281 | $codepoint = $refs['codepoint']; | |
2282 | } | |
2283 | $chars .= $char = $this->stream->char(); | |
2284 | } | |
2285 | ||
2286 | // Unconsume the one character we just took which caused the while | |
2287 | // statement to fail. This could be anything and could cause state | |
2288 | // changes (as if it matches the while loop it must be | |
2289 | // alphanumeric so we can just concat it to whatever we get later). | |
2290 | $this->stream->unget(); | |
2291 | if ($char !== false) { | |
2292 | $chars = substr($chars, 0, -1); | |
2293 | } | |
2294 | ||
2295 | /* If no match can be made, then this is a parse error. | |
2296 | No characters are consumed, and nothing is returned. */ | |
2297 | if (!$codepoint) { | |
2298 | $this->emitToken(array( | |
2299 | 'type' => self::PARSEERROR, | |
2300 | 'data' => 'expected-named-entity' | |
2301 | )); | |
2302 | return '&' . $chars; | |
2303 | } | |
2304 | ||
2305 | /* If the last character matched is not a U+003B SEMICOLON | |
2306 | (;), there is a parse error. */ | |
2307 | $semicolon = true; | |
2308 | if (substr($id, -1) !== ';') { | |
2309 | $this->emitToken(array( | |
2310 | 'type' => self::PARSEERROR, | |
2311 | 'data' => 'named-entity-without-semicolon' | |
2312 | )); | |
2313 | $semicolon = false; | |
2314 | } | |
2315 | ||
2316 | /* If the character reference is being consumed as part of | |
2317 | an attribute, and the last character matched is not a | |
2318 | U+003B SEMICOLON (;), and the next character is in the | |
2319 | range U+0030 DIGIT ZERO to U+0039 DIGIT NINE, U+0041 | |
2320 | LATIN CAPITAL LETTER A to U+005A LATIN CAPITAL LETTER Z, | |
2321 | or U+0061 LATIN SMALL LETTER A to U+007A LATIN SMALL LETTER Z, | |
2322 | then, for historical reasons, all the characters that were | |
2323 | matched after the U+0026 AMPERSAND (&) must be unconsumed, | |
2324 | and nothing is returned. */ | |
2325 | if ($inattr && !$semicolon) { | |
2326 | // The next character is either the next character in $chars or in the stream. | |
2327 | if (strlen($chars) > strlen($id)) { | |
2328 | $next = substr($chars, strlen($id), 1); | |
2329 | } else { | |
2330 | $next = $this->stream->char(); | |
2331 | $this->stream->unget(); | |
2332 | } | |
2333 | if ( | |
2334 | '0' <= $next && $next <= '9' || | |
2335 | 'A' <= $next && $next <= 'Z' || | |
2336 | 'a' <= $next && $next <= 'z' | |
2337 | ) { | |
2338 | return '&' . $chars; | |
2339 | } | |
2340 | } | |
2341 | ||
2342 | /* Otherwise, return a character token for the character | |
2343 | corresponding to the character reference name (as given | |
2344 | by the second column of the named character references table). */ | |
2345 | return HTML5_Data::utf8chr($codepoint) . substr($chars, strlen($id)); | |
2346 | } | |
2347 | } | |
2348 | ||
2349 | private function characterReferenceInAttributeValue($allowed = false) { | |
2350 | /* Attempt to consume a character reference. */ | |
2351 | $entity = $this->consumeCharacterReference($allowed, true); | |
2352 | ||
2353 | /* If nothing is returned, append a U+0026 AMPERSAND | |
2354 | character to the current attribute's value. | |
2355 | ||
2356 | Otherwise, append the returned character token to the | |
2357 | current attribute's value. */ | |
2358 | $char = (!$entity) | |
2359 | ? '&' | |
2360 | : $entity; | |
2361 | ||
2362 | $last = count($this->token['attr']) - 1; | |
2363 | $this->token['attr'][$last]['value'] .= $char; | |
2364 | ||
2365 | /* Finally, switch back to the attribute value state that you | |
2366 | were in when were switched into this state. */ | |
2367 | } | |
2368 | ||
2369 | /** | |
2370 | * Emits a token, passing it on to the tree builder. | |
2371 | */ | |
2372 | protected function emitToken($token, $checkStream = true, $dry = false) { | |
2373 | if ($checkStream) { | |
2374 | // Emit errors from input stream. | |
2375 | while ($this->stream->errors) { | |
2376 | $this->emitToken(array_shift($this->stream->errors), false); | |
2377 | } | |
2378 | } | |
2379 | if($token['type'] === self::ENDTAG && !empty($token['attr'])) { | |
2380 | for ($i = 0; $i < count($token['attr']); $i++) { | |
2381 | $this->emitToken(array( | |
2382 | 'type' => self::PARSEERROR, | |
2383 | 'data' => 'attributes-in-end-tag' | |
2384 | )); | |
2385 | } | |
2386 | } | |
2387 | if($token['type'] === self::ENDTAG && !empty($token['self-closing'])) { | |
2388 | $this->emitToken(array( | |
2389 | 'type' => self::PARSEERROR, | |
2390 | 'data' => 'self-closing-flag-on-end-tag', | |
2391 | )); | |
2392 | } | |
2393 | if($token['type'] === self::STARTTAG) { | |
2394 | // This could be changed to actually pass the tree-builder a hash | |
2395 | $hash = array(); | |
2396 | foreach ($token['attr'] as $keypair) { | |
2397 | if (isset($hash[$keypair['name']])) { | |
2398 | $this->emitToken(array( | |
2399 | 'type' => self::PARSEERROR, | |
2400 | 'data' => 'duplicate-attribute', | |
2401 | )); | |
2402 | } else { | |
2403 | $hash[$keypair['name']] = $keypair['value']; | |
2404 | } | |
2405 | } | |
2406 | } | |
2407 | ||
2408 | if(!$dry) { | |
2409 | // the current structure of attributes is not a terribly good one | |
2410 | $this->tree->emitToken($token); | |
2411 | } | |
2412 | ||
2413 | if(!$dry && is_int($this->tree->content_model)) { | |
2414 | $this->content_model = $this->tree->content_model; | |
2415 | $this->tree->content_model = null; | |
2416 | ||
2417 | } elseif($token['type'] === self::ENDTAG) { | |
2418 | $this->content_model = self::PCDATA; | |
2419 | } | |
2420 | } | |
2421 | } | |
2422 |