[webkit-reviews] review granted: [Bug 27895] [XSSAuditor] Inline Event Handler with single-line JavaScript comment can bypass XSSAuditor : [Attachment 107229] Patch simplified.

[bugzilla-daemon at webkit.org]

Adam Barth <abarth at webkit.org> has granted Thomas Sepez <tsepez at chromium.org>'s
request for review:
Bug 27895: [XSSAuditor] Inline Event Handler with single-line JavaScript
comment can bypass XSSAuditor
https://bugs.webkit.org/show_bug.cgi?id=27895

Attachment 107229: Patch simplified.
https://bugs.webkit.org/attachment.cgi?id=107229&action=review

------- Additional Comments from Adam Barth <abarth at webkit.org>
View in context: https://bugs.webkit.org/attachment.cgi?id=107229&action=review


This looks like a great start.	We might need to iterate on the ordering
between the truncation and the canonicalization, but we can do that in a
separate patch.  I'd like to hear your answers to the questions in my last
review before landing.

> Source/WebCore/html/parser/XSSAuditor.cpp:492
> +    // Limit the length of the fragment to avoid comparing very long
strings.
> +    snippet.truncate(kMaximumSnippetLength);

Will this cause a problem if we truncate in the middle of a %-escape sequence
of a UTF-16 surrogate?	I'm worried that the canonicalization step will end up
with a different result.  Maybe we should canonicalize before trimming the
snippet?