2 |
<html lang=en> |
<html lang=en> |
3 |
<head> |
<head> |
4 |
<title>Live Scripting HTML Parser</title> |
<title>Live Scripting HTML Parser</title> |
5 |
|
<link rel=author href="http://suika.fam.cx/~wakaba/who?"> |
6 |
|
<link rel=license href="http://suika.fam.cx/c/gnu/gpl" |
7 |
|
title="GNU GPL2 or later"> |
8 |
<style> |
<style> |
9 |
h1, h2 { |
h1 { |
10 |
|
margin: 0; |
11 |
|
font-size: 150%; |
12 |
|
} |
13 |
|
h2 { |
14 |
margin: 0; |
margin: 0; |
15 |
font-size: 100%; |
font-size: 100%; |
16 |
} |
} |
17 |
p, pre { |
p { |
18 |
margin: 0; |
margin: 0 1em; |
19 |
} |
} |
20 |
textarea { |
textarea { |
21 |
width: 100%; |
width: 100%; |
86 |
this.input = i; |
this.input = i; |
87 |
this.scriptsExecutedAfterParsing = []; |
this.scriptsExecutedAfterParsing = []; |
88 |
this.scriptsExecutedSoon = []; |
this.scriptsExecutedSoon = []; |
89 |
|
this.scriptsExecutedAsynchronously = []; |
90 |
} // Parser |
} // Parser |
91 |
|
|
92 |
Parser.prototype.getNextToken = function () { |
Parser.prototype.getNextToken = function () { |
93 |
var p = this; |
var p = this; |
94 |
var i = this.input; |
var i = this.input; |
95 |
if (this.parseMode == 'script') { |
if (this.parseMode == 'cdata') { |
96 |
|
var tagName = this.endTagName; |
97 |
var token; |
var token; |
98 |
if (p.insertionPoint <= 0) { |
if (p.insertionPoint <= 0) { |
99 |
return {type: 'abort'}; |
return {type: 'abort'}; |
111 |
return ''; |
return ''; |
112 |
}); |
}); |
113 |
if (token) return token; |
if (token) return token; |
114 |
i.s = i.s.replace (/^<\/[Ss][Cc][Rr][Ii][Pp][Tt]>/, function (s) { |
var pattern = new RegExp ('^</' + tagName + '>', 'i'); |
115 |
|
i.s = i.s.replace (pattern, function (s) { |
116 |
if (p.insertionPoint < s.length) { |
if (p.insertionPoint < s.length) { |
117 |
token = {type: 'abort'}; |
token = {type: 'abort'}; |
118 |
return s; |
return s; |
119 |
} |
} |
120 |
token = {type: 'end-tag', value: 'script'}; |
token = {type: 'end-tag', value: tagName}; |
121 |
p.insertionPoint -= s.length; |
p.insertionPoint -= s.length; |
122 |
return ''; |
return ''; |
123 |
}); |
}); |
124 |
if (token) return token; |
if (token) return token; |
125 |
var m; |
var m; |
126 |
if ((p.insertionPoint < '</script'.length) && |
if ((p.insertionPoint < ('</' + tagName).length) && |
127 |
(m = i.s.match (/^<\/([SCRIPTscript]+)/))) { |
(m = i.s.match (/^<\/([A-Za-z]+)/))) { |
128 |
var v = m[1].substring (0, p.insertionPoint).toLowerCase (); |
var v = m[1].substring (0, p.insertionPoint).toLowerCase (); |
129 |
if (v == 'script'.substring (0, p.insertionPoint - '</'.length)) { |
if (v == tagName.substring (0, p.insertionPoint - '</'.length)) { |
130 |
return {type: 'abort'}; |
return {type: 'abort'}; |
131 |
} |
} |
132 |
} |
} |
231 |
el.manakaiParserInserted = true; |
el.manakaiParserInserted = true; |
232 |
|
|
233 |
// 3. Switch the tokeniser's content model flag to the CDATA state. |
// 3. Switch the tokeniser's content model flag to the CDATA state. |
234 |
this.parseMode = 'script'; |
this.parseMode = 'cdata'; |
235 |
|
this.endTagName = 'script'; |
236 |
|
|
237 |
// 4.1. Collect all the character tokens. |
// 4.1. Collect all the character tokens. |
238 |
while (true) { |
while (true) { |
246 |
// 4.2. Until it returns a token that is not a character token, or |
// 4.2. Until it returns a token that is not a character token, or |
247 |
// until it stops tokenising. |
// until it stops tokenising. |
248 |
} else if (token.type == 'eof' || |
} else if (token.type == 'eof' || |
249 |
(token.type == 'end-tag' && token.value == 'script') || |
token.type == 'end-tag' || |
250 |
token.type == 'abort') { |
token.type == 'abort') { |
251 |
// 6. Switched back to the PCDATA state. |
// 6. Switched back to the PCDATA state. |
252 |
this.parseMode = 'pcdata'; |
this.parseMode = 'pcdata'; |
253 |
|
|
254 |
// 7.1. If the next token is not an end tag token with ... |
// 7.1. If the next token is not an end tag token with ... |
255 |
if (token.type != 'end-tag') { |
if (!(token.type == 'end-tag' && token.value == 'script')) { |
256 |
// 7.2. This is a parse error. |
// 7.2. This is a parse error. |
257 |
log ('Parse error: no </' + 'script>'); |
log ('Parse error: no </' + 'script>'); |
258 |
|
|
316 |
// |
// |
317 |
} |
} |
318 |
} |
} |
319 |
|
} else if (token.value == 'style' || |
320 |
|
token.value == 'noscript' || |
321 |
|
token.value == 'xmp') { |
322 |
|
// 1. Create an element for the token in the HTML namespace. |
323 |
|
var el = new JSElement (this.doc, token.value); |
324 |
|
|
325 |
|
// 2. Append the new element to the current node. |
326 |
|
this.openElements[this.openElements.length - 1].appendChild (el); |
327 |
|
|
328 |
|
// 3. Switch the tokeniser's content model flag to the CDATA state. |
329 |
|
this.parseMode = 'cdata'; |
330 |
|
this.endTagName = token.value; |
331 |
|
|
332 |
|
// 4.1. Collect all the character tokens. |
333 |
|
while (true) { |
334 |
|
var token = this.getNextToken (); |
335 |
|
log ('token: ' + token.type + ' "' + token.value + '"'); |
336 |
|
|
337 |
|
if (token.type == 'char') { |
338 |
|
// 5. Append a single Text node to the script element node. |
339 |
|
el.manakaiAppendText (token.value); |
340 |
|
|
341 |
|
// 4.2. Until it returns a token that is not a character token, or |
342 |
|
// until it stops tokenising. |
343 |
|
} else if (token.type == 'eof' || |
344 |
|
token.type == 'end-tag' || |
345 |
|
token.type == 'abort') { |
346 |
|
// 6. Switched back to the PCDATA state. |
347 |
|
this.parseMode = 'pcdata'; |
348 |
|
|
349 |
|
// 7.1. If the next token is not an end tag token with ... |
350 |
|
if (!(token.type == 'end-tag' && |
351 |
|
token.value == this.endTagName)) { |
352 |
|
// 7.2. This is a parse error. |
353 |
|
log ('Parse error: no </' + this.endTagName + '>'); |
354 |
|
|
355 |
|
// 7.3. Mark the script element as "already executed". |
356 |
|
el.manakaiAlreadyExecuted = true; |
357 |
|
} else { |
358 |
|
// 7.4. Ignore it. |
359 |
|
// |
360 |
|
} |
361 |
|
break; |
362 |
|
} |
363 |
|
} |
364 |
} else { |
} else { |
365 |
var el = new JSElement (this.doc, token.value); |
var el = new JSElement (this.doc, token.value); |
366 |
this.openElements[this.openElements.length - 1].appendChild (el); |
this.openElements[this.openElements.length - 1].appendChild (el); |
391 |
|
|
392 |
// "When a script completes loading" rules start applying. |
// "When a script completes loading" rules start applying. |
393 |
|
|
394 |
// List of scripts that will execute as soon as possible |
while (this.scriptsExecutedSoon.length > 0 || |
395 |
for (var i = 0; i < this.scriptsExecutedSoon.length; i++) { |
this.scriptsExecutedAsynchronously.length > 0) { |
396 |
var e = this.scriptsExecutedSoon[i]; |
// Handle "list of scripts that will execute as soon as possible". |
397 |
|
while (this.scriptsExecutedSoon.length > 0) { |
398 |
// If it has completed loading |
var e = this.scriptsExecutedSoon.shift (); |
399 |
log ('Execute an external script not inserted by parser...'); |
|
400 |
executeScript (this.doc, e); |
// If it has completed loading |
401 |
|
log ('Execute an external script not inserted by parser...'); |
402 |
|
executeScript (this.doc, e); |
403 |
|
|
404 |
|
// NOTE: It MAY be executed before the end of the parsing, according |
405 |
|
// to the spec. |
406 |
|
this.hasAsyncScript = true; |
407 |
|
} |
408 |
|
|
409 |
|
// Handle "list of scripts that will execute asynchronously". |
410 |
|
while (this.scriptsExecutedAsynchronously.length > 0) { |
411 |
|
var e = this.scriptsExecutedAsynchronously.shift (); |
412 |
|
|
413 |
|
// Step 1. |
414 |
|
// We assume that all scripts have been loaded at this time. |
415 |
|
|
416 |
|
// Step 2. |
417 |
|
log ('Execute an asynchronous script...'); |
418 |
|
executeScript (this.doc, e); |
419 |
|
|
420 |
|
// Step 3. |
421 |
|
// |
422 |
|
|
423 |
|
// Step 4. |
424 |
|
// |
425 |
|
|
426 |
// NOTE: It MAY be executed before the end of the parsing, according |
this.hasAsyncScript = true; |
427 |
// to the spec. |
} |
|
this.hasAsyncScript = true; |
|
428 |
} |
} |
429 |
|
|
|
// TODO: Handles |
|
|
// "list of scripts that will execute asynchronously" |
|
|
|
|
430 |
// Handle "list of scripts that will execute when the document has finished |
// Handle "list of scripts that will execute when the document has finished |
431 |
// parsing". |
// parsing". |
432 |
var list = this.scriptsExecutedAfterParsing; |
var list = this.scriptsExecutedAfterParsing; |
445 |
|
|
446 |
log ('DOMContentLoaded event fired'); |
log ('DOMContentLoaded event fired'); |
447 |
|
|
448 |
// "delays tha load event" things has completed: |
// "delays the load event" things has completed: |
449 |
// readyState = 'complete' |
// readyState = 'complete' |
450 |
log ('load event fired'); |
log ('load event fired'); |
451 |
|
|
520 |
p.scriptsExecutedAfterParsing.push (e); |
p.scriptsExecutedAfterParsing.push (e); |
521 |
log ('Running a script: aborted (defer)'); |
log ('Running a script: aborted (defer)'); |
522 |
} else if (e.async && e.src != null) { |
} else if (e.async && e.src != null) { |
523 |
// TODO |
p.scriptsExecutedAsynchronously.push (e); |
524 |
} else if (e.async && e.src == null |
log ('Running a script: aborted (async src)'); |
525 |
/* && list of scripts that will execute asynchronously is not empty */) { |
} else if (e.async && e.src == null && |
526 |
// TODO |
p.scriptsExecutedAsynchronously.length > 0) { |
527 |
|
p.scriptsExecutedAsynchronously.push (e); |
528 |
|
log ('Running a script: aborted (async)'); |
529 |
|
// ISSUE: What is the difference with the case above? |
530 |
} else if (e.src != null && e.manakaiParserInserted) { |
} else if (e.src != null && e.manakaiParserInserted) { |
531 |
if (p.scriptExecutedWhenParserResumes) { |
if (p.scriptExecutedWhenParserResumes) { |
532 |
log ('Error: There is a script that will execute as soon as the parser resumes.'); |
log ('Error: There is a script that will execute as soon as the parser resumes.'); |
846 |
algorithm, and so on. |
algorithm, and so on. |
847 |
<li>Does not raise parse errors for invalid attribute specifications in start |
<li>Does not raise parse errors for invalid attribute specifications in start |
848 |
or end tags. |
or end tags. |
849 |
<li>Does not support CDATA/PCDATA element other than <code>script</code>. |
<li>Does not support PCDATA elements (<code>title</code> and |
850 |
|
<code>textarea</code>). |
851 |
|
<li>Does not strip the first newline in <code>pre</code> elements. |
852 |
<li>Does not support <code><!--</code>..<code>--></code> parsing rule |
<li>Does not support <code><!--</code>..<code>--></code> parsing rule |
853 |
in <code>script</code> element. |
in <code>script</code> element. |
854 |
<li>Does not support foreign (SVG or MathML) elements. |
<li>Does not support foreign (SVG or MathML) elements. |
873 |
the regular expression <code>^javascript:\s*(?:"[^"]*"|'[^']*')\s*$</code>. |
the regular expression <code>^javascript:\s*(?:"[^"]*"|'[^']*')\s*$</code>. |
874 |
<li>Only supports <code>\u<var>HHHH</var></code> escapes in JavaScript |
<li>Only supports <code>\u<var>HHHH</var></code> escapes in JavaScript |
875 |
string literals. |
string literals. |
876 |
|
<li>Does not handle <i>stop parsing</i> phase correctly if the document is |
877 |
|
replaced by <code>document.open ()</code> call. In other word, delayed |
878 |
|
(deferred or asynchronous) script executions and event firings might be |
879 |
|
treated in a wrong way if a <code>document.open ()</code> invocation |
880 |
|
is implicitly done by <code>document.write ()</code> in a delayed script. |
881 |
</ul> |
</ul> |
882 |
|
|
883 |
<p>For some reason, this parser does not work in browsers that do |
<p>For some reason, this parser does not work in browsers that do |
884 |
not support JavaScript 1.5. |
not support JavaScript 1.5. |
885 |
|
|
886 |
<!-- TODO: license --> |
<!-- TODO: |src| attribute value should refer the value at the time |
887 |
|
when it is inserted into the document, not the value when the script is |
888 |
|
executed. Currently it does not matter, since we don't allow dynamic |
889 |
|
modification to the |src| content/DOM attribute value yet. --> |
890 |
|
|
891 |
</body> |
</body> |
|
</html> |
|
892 |
|
</html> |
893 |
|
<!-- $Date$ --> |
894 |
|
<!-- |
895 |
|
|
896 |
|
Copyright 2008 Wakaba <w@suika.fam.cx> |
897 |
|
|
898 |
|
This program is free software; you can redistribute it and/or |
899 |
|
modify it under the terms of the GNU General Public License |
900 |
|
as published by the Free Software Foundation; either version 2 |
901 |
|
of the License, or (at your option) any later version. |
902 |
|
|
903 |
|
This program is distributed in the hope that it will be useful, |
904 |
|
but WITHOUT ANY WARRANTY; without even the implied warranty of |
905 |
|
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the |
906 |
|
GNU General Public License for more details. |
907 |
|
|
908 |
|
You should have received a copy of the GNU General Public License |
909 |
|
along with this program; if not, write to the Free Software |
910 |
|
Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301, USA. |
911 |
|
|
912 |
|
--> |