[webkit-reviews] review granted: [Bug 226912] Check for "xn--" in any subdomain when parsing URL hosts : [Attachment 431360] Patch

[bugzilla-daemon at webkit.org]

Darin Adler <darin at apple.com> has granted Alex Christensen
<achristensen at apple.com>'s request for review:
Bug 226912: Check for "xn--" in any subdomain when parsing URL hosts
https://bugs.webkit.org/show_bug.cgi?id=226912

Attachment 431360: Patch

https://bugs.webkit.org/attachment.cgi?id=431360&action=review




--- Comment #14 from Darin Adler <darin at apple.com> ---
Comment on attachment 431360
  --> https://bugs.webkit.org/attachment.cgi?id=431360
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=431360&action=review

> Source/WTF/wtf/URL.cpp:-406
> -    // For host names bigger than this, we won't do IDN encoding, which is
almost certainly OK.

I feel like it’s a problem to lose this comment. Could write a better one, but
just removing it seems wrong.

Without the comment it’s not obvious why it’s OK to leave the hostname as-is
when it’s super long, and the check just below is mysterious.

> Source/WTF/wtf/URLHelpers.cpp:-43
> -// For host names bigger than this, we won't do IDN encoding, which is
almost certainly OK.

Again, losing this comment.

> Source/WTF/wtf/URLHelpers.cpp:575
> +    if (hostName.length() > URLParser::hostnameBufferLength)

Mysterious code without a comment. Why is it safe to just do nothing for long
strings.