Bug 45240 – [Qt] utf8 encoding of console() messages

• Status: CLOSED
• Resolution: FIXED
• Component: WebKit API

[Robert Hogan]

http/tests/security/xssAuditor/embed-tag-null-char.html
http/tests/security/xssAuditor/object-embed-tag-null-char.html

both fail because ChromeClientQt::addMessageToConsole() is casting String to QString rather than String::utf8().data().

[Robert Hogan]

Created attachment 66596 [details]
Patch

[Antonio Gomes]

Comment on attachment 66596 [details]
Patch

What about bug 35263 ?

[Robert Hogan]

(In reply to comment #2)
> (From update of attachment 66596 [details])
> What about bug 35263 ?

That ref in the Skipped list is misleading - bugs are unrelated. I think it was me who added that misleading reference!

[WebKit Commit Bot]

Comment on attachment 66596 [details]
Patch

Clearing flags on attachment: 66596

Committed r66801: <http://trac.webkit.org/changeset/66801>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[WebKit Review Bot]

http://trac.webkit.org/changeset/66801 might have broken Qt Linux Release

[Antonio Gomes]

it was rolled out. reopenning

[Robert Hogan]

Created attachment 66599 [details]
Patch

[Robert Hogan]

(In reply to comment #7)
> it was rolled out. reopenning

I think I understand the problem now. However the solution involves creating a QString, casting to a QByteArray, then char*, then String, then back to QString, then a QByteArray, then const *char - at which point something finally gets printed out!

[Andreas Kling]

Comment on attachment 66599 [details]
Patch

> +    return StringImpl::create(buffer.toUtf8().constData(), buffer.toUtf8().length());

This will cause the UTF-8 conversion to be done twice.

r=me, but please fix that before landing.

[Robert Hogan]

(In reply to comment #10)
> (From update of attachment 66599 [details])
> > +    return StringImpl::create(buffer.toUtf8().constData(), buffer.toUtf8().length());
> 
> This will cause the UTF-8 conversion to be done twice.
> 
> r=me, but please fix that before landing.

heh, right enough. i think i get it now!

[Robert Hogan]

Committed r66843: <http://trac.webkit.org/changeset/66843>

[WebKit Review Bot]

http://trac.webkit.org/changeset/66843 might have broken Qt Linux Release minimal, Qt Linux ARMv5 Release, and Qt Linux ARMv7 Release

[Robert Hogan]

(In reply to comment #10)
> (From update of attachment 66599 [details])
> > +    return StringImpl::create(buffer.toUtf8().constData(), buffer.toUtf8().length());
> 
> This will cause the UTF-8 conversion to be done twice.
> 
> r=me, but please fix that before landing.

I thought I had followed your advice but made the change during a compile and mistakenly thought it had been caught - but it hadn't.

I don't see a way of fixing this without using toUtf8().constData().

[Antonio Gomes]

(In reply to comment #14)
> (In reply to comment #10)
> > (From update of attachment 66599 [details] [details])
> > > +    return StringImpl::create(buffer.toUtf8().constData(), buffer.toUtf8().length());
> > 
> > This will cause the UTF-8 conversion to be done twice.
> > 
> > r=me, but please fix that before landing.
> 
> I thought I had followed your advice but made the change during a compile and mistakenly thought it had been caught - but it hadn't.
> 
> I don't see a way of fixing this without using toUtf8().constData().

maybe he was referring you called buffer.toUtf8() twice?