Live Scripting HTML Parser

Markup to test (permalink, Live DOM Viewer)

<html> <head></head><body> <p> <script> document.write ('aaaaaaa</p><script>document.write("cccccc");</', 'script>bbbbbb'); </script> <p>

Log

parse: start token: start-tag "html" token: char " " token: start-tag "head" token: end-tag "head" token: start-tag "body" token: char " " token: start-tag "p" token: char " " token: start-tag "script" token: char " document.write ('aaaaaaa" token: char "<" token: char "/p>" token: char "<" token: char "script>document.write("cccccc");" token: char "<" token: char "/', 'script>bbbbbb'); " token: end-tag "script" insertion point: set to 0 (before " <p> ") Running a script: start executing a script block: start load event fired at the script element document.write: start

Notes

This is a simplified implementation of HTML5 Parsing Algorithm. It only implements script-related part of the algorithm. Especially, this parser:

Does not support DOCTYPE and comment tokens.
Does not support entities except for ", ', and & in script src attribute value.
Does not support omissions of start or end tags, the AAA algorithm, and so on.
Does not raise parse errors for invalid attribute specifications in start or end tags.
Does not support PCDATA elements (title and textarea).
Does not strip the first newline in pre elements.
Does not support  parsing rule in script element.
Does not support foreign (SVG or MathML) elements.
Only supports script type text/javascript. type and language attributes are ignored.
Only supports limited statements. It must consist of zero or more of statements looking similar to the following statements, possibly introduced, followed, or separated by white space characters:
- document.write ("string", ["string", ...]);.
- var s = document.createElement ("script"); s.src = "string"; document.documentElement.appendChild (s);
- w (document.documentElement.innerHTML); (This statement can be used to dump the document, even when the document has no document element. The output format is the tree dump format used in html5lib test data, not HTML.)
Note that strings may be delimited by 's instead of "s.
Only supports javascript: URI scheme in the src attribute of the script element. In addition, the URI must be conform to the regular expression ^javascript:\s*(?:"[^"]*"|'[^']*')\s*$.
Only supports \uHHHH escapes in JavaScript string literals.
Does not handle stop parsing phase correctly if the document is replaced by document.open () call. In other word, delayed (deferred or asynchronous) script executions and event firings might be treated in a wrong way if a document.open () invocation is implicitly done by document.write () in a delayed script.

For some reason, this parser does not work in browsers that do not support JavaScript 1.5.