upload qian.html

lifajing398, can you provide more detailed instructions for how to reproduce this bug?

I think this may be an html parser bug, but I'm not yet sure from the description.

lifajing398, do you mean laoding qian.html reproduces the problem?  Is this reproducible Google Chrome browser, not WebView?

It is difficult to reproduce this bug by Google Chrome browser.

I also think this is an html parser bug.

print log:

E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=20,  end=0
E/HTMLDocumentParser( 8489): tokens:type=2,  textPosition start=20,  end=31
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=45,  end=9
E/HTMLDocumentParser( 8489): tokens:type=3,  textPosition start=45,  end=9
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): break  tokens:type=5,  textPosition start=48,  end=1 , locationChangePending=1
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=67,  end=55
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=67,  end=6899
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=75,  end=41
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=89,  end=59
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=97,  end=7061
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=123,  end=2108
E/HTMLDocumentParser( 8489): Inside --> HTMLDocumentParser::didReceiveParsedChunkFromBackgroundParser
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=130,  end=8
E/HTMLDocumentParser( 8489): tokens:type=3,  textPosition start=130,  end=8
E/HTMLDocumentParser( 8489): tokens:type=2,  textPosition start=130,  end=39
E/HTMLDocumentParser( 8489): tokens:type=5,  textPosition start=131,  end=2942

Thank you for providing more feedback. Adding requester "tkent@chromium.org" for another review and adding "Needs-Review" label for tracking.

For more details visit https://www.chromium.org/issue-tracking/autotriage - Your friendly Sheriffbot

kouhei@, any idea about this?

hi kouhei ,tkent

when location change is pending ,  break some start tag,

the html tree building mode become unpredictable.

if redirecting is finish and parse not end of file.

sometime ,the rest of html source show as string.

https://codereview.chromium.org/2107763002/

lifajing398@: can you clarify the expected & actual behavior a little clearer?  Does just opening qian.html in WebView reproduce the problem, or will the user also need to select 'show source code' manually at some point?  Also: what was the actual (wrong) behavior you observed?

kouhei@: would you be able to triage this / help him land the patch if it makes sense to you?

Tentatively me I guess but feel free to poach.

The log is interesting, the patch kind of makes sense, and the code does look a bit iffy--peek at the end of the tokens for EOF? Weird.

So maybe to trigger this you need a navigation, with background tokenization, but have the chunk *not* reach EOF token. This probably tickles one of those asserts in debug.

Bulk disowning per sshruthi's email about bug triage best practices.

We haven't received feedback since Jul 27 2016, so I'm just going to assume this has been fixed or isn't a big issue.