A user account is required in order to edit this wiki, but we've had to disable public user registrations due to spam.

To request an account, ask an autoconfirmed user on Chat (such as one of these permanent autoconfirmed members).

ParserIssues: Difference between revisions

From WHATWG Wiki
Jump to navigation Jump to search
(→‎Filed bugs: Add document.write() ignoring bug)
(→‎Filed bugs: </html> and </body> in "in foreign content" cause severe badness)
 
(9 intermediate revisions by 3 users not shown)
Line 6: Line 6:
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9580 Steps for handling end tags in the "in foreign" mode walk past the root node in the stack
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9580 Steps for handling end tags in the "in foreign" mode walk past the root node in the stack
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9767 Consider ignoring document.write() when IE ignores it if IE has a sane condition for ignoring it
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9767 Consider ignoring document.write() when IE ignores it if IE has a sane condition for ignoring it
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9829 <button> should be scoping for the purpose of implicitly closing <nowiki><p></nowiki>
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9831 Handling of EOF in foreignContent mode causes all tags to be popped of stack of open elements
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9843 <nowiki>Specced behavior for document.write("<link rel=stylesheet href=...><script>...</script>...") matches none of the top 4 engines</nowiki>
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9910 <nowiki><basefont>  and <bgsound> should not break out of 'in head', should be treated as a void elements within <head></nowiki>
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9983 Insertion point not saved for external scripts
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9984 Insertion point for script@onload doesn't match Firefox
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=9985 How to parse </foo </bar>
* http://www.w3.org/Bugs/Public/show_bug.cgi?id=10055 <nowiki></html> and </body> in "in foreign content" cause severe badness</nowiki>
== Emails ==
* http://lists.whatwg.org/pipermail/implementors-whatwg.org/2010-June/000273.html Tokenizer spec switches the tokenizer state back to the data state right after the tree builder may have set it to the state that it needs to be switched to


== Possible other issues ==
== Possible other issues ==


* <nowiki><!DOCTYPE html><p><b><i><u></p> <p>X</nowiki> - should spaceness of character have any effect in InBody mode?
* <nowiki><!DOCTYPE html><p><b><i><u></p> <p>X</nowiki> - should spaceness of character have any effect in InBody mode?
* To enable the parser to decide whether a given script can call document.write() (for optimizing speculative parsing), it is necessary for the execution properties of the script as seen at start tag to remain constant until the end tag is parsed. This can be achieved by freezing the effective values for the src, defer and async attributes when a parser-inserted script is inserted to the tree. There's no non-racy way for a script to detect this when the script comes from the network stream, so this is OK. However, if the parser implements the same script execution logic for the network stream case and the document.write case in the same way, there is a non-racy way to detect the implementation details in the document.write() case. For an implementation that does speculative off-the-main-thread parsing, it is a developer time-consuming theoretical purity exercise to make the document.write() case correct per spec. OTOH, if the spec enshrines the freezing behavior as correct, it'll be a developer time-consuming theoretical purity exercise for implementations that don't have the kind of speculative parsing that needs the freezing.
* Having <nowiki>-- ></nowiki> terminate comments breaks CNN Money https://bugzilla.mozilla.org/show_bug.cgi?id=570309

Latest revision as of 09:34, 30 June 2010

Filed bugs

Emails

Possible other issues

  • <!DOCTYPE html><p><b><i><u></p> <p>X - should spaceness of character have any effect in InBody mode?
  • To enable the parser to decide whether a given script can call document.write() (for optimizing speculative parsing), it is necessary for the execution properties of the script as seen at start tag to remain constant until the end tag is parsed. This can be achieved by freezing the effective values for the src, defer and async attributes when a parser-inserted script is inserted to the tree. There's no non-racy way for a script to detect this when the script comes from the network stream, so this is OK. However, if the parser implements the same script execution logic for the network stream case and the document.write case in the same way, there is a non-racy way to detect the implementation details in the document.write() case. For an implementation that does speculative off-the-main-thread parsing, it is a developer time-consuming theoretical purity exercise to make the document.write() case correct per spec. OTOH, if the spec enshrines the freezing behavior as correct, it'll be a developer time-consuming theoretical purity exercise for implementations that don't have the kind of speculative parsing that needs the freezing.
  • Having -- > terminate comments breaks CNN Money https://bugzilla.mozilla.org/show_bug.cgi?id=570309