Restore NOTREACHED() in WebKitClientImpl::loadResource()

Restore NOTREACHED() in WebKitClientImpl::loadResource()

BUG=50675
TEST=none


Committed: http://src.chromium.org/viewvc/chrome?view=rev&revision=91767

[Ilya Sherman, 2011-07-07 08:11:31]

If you don't like returning early for an empty string arg, I can go ahead and
clean up https://bugs.webkit.org/show_bug.cgi?id=64044.  But, as Adam points out
there, it's probably cleaner to handle this pattern uniformly in one place,
rather than having to worry about it at each callsite.

[James Hawkins, 2011-07-07 16:24:07]

On 2011/07/07 08:11:31, Ilya Sherman wrote:
> If you don't like returning early for an empty string arg, I can go ahead and
> clean up https://bugs.webkit.org/show_bug.cgi?id=64044.  But, as Adam points
out
> there, it's probably cleaner to handle this pattern uniformly in one place,
> rather than having to worry about it at each callsite.

I don't see how you get "handle this pattern uniformly in one place" from Adam's
comments on that WK bug.

[Ilya Sherman, 2011-07-07 18:50:21]

On 2011/07/07 16:24:07, James Hawkins wrote:
> On 2011/07/07 08:11:31, Ilya Sherman wrote:
> > If you don't like returning early for an empty string arg, I can go ahead
and
> > clean up https://bugs.webkit.org/show_bug.cgi?id=64044.  But, as Adam points
> out
> > there, it's probably cleaner to handle this pattern uniformly in one place,
> > rather than having to worry about it at each callsite.
> 
> I don't see how you get "handle this pattern uniformly in one place" from
Adam's
> comments on that WK bug.

The relevant comment is "Why doesn't loadPlatformResource just return a null ptr
if icon is empty?"

[James Hawkins, 2011-07-07 20:47:37]

I still don't see how one implies the other, but I'm not that concerned about
it.

LGTM with comment.

http://codereview.chromium.org/7316019/diff/1/webkit/glue/webkitclient_impl.cc
File webkit/glue/webkitclient_impl.cc (right):

http://codereview.chromium.org/7316019/diff/1/webkit/glue/webkitclient_impl.c...
webkit/glue/webkitclient_impl.cc:387: if (!strlen(name))
Add a comment for why this may happen.

[Ilya Sherman, 2011-07-07 20:51:49]

http://codereview.chromium.org/7316019/diff/1/webkit/glue/webkitclient_impl.cc
File webkit/glue/webkitclient_impl.cc (right):

http://codereview.chromium.org/7316019/diff/1/webkit/glue/webkitclient_impl.c...
webkit/glue/webkitclient_impl.cc:387: if (!strlen(name))
On 2011/07/07 20:47:37, James Hawkins wrote:
> Add a comment for why this may happen.

Done.

[commit-bot: I haz the power, 2011-07-07 22:09:19]

Change committed as 91767