[Webkit-unassigned] [Bug 48578] HTML5 Conformance Test failure: approved/canvas/canvas_text_font_001.htm

bugzilla-daemon at webkit.org bugzilla-daemon at webkit.org
Thu Jun 2 02:03:15 PDT 2011


https://bugs.webkit.org/show_bug.cgi?id=48578





--- Comment #14 from Mustafizur Rahaman <mustaf.here at gmail.com>  2011-06-02 02:03:15 PST ---
(In reply to comment #13)
> The approach seems fine.  We just need better testing.  "inherit 20px" and "inherit" and "INHERIT" and "Inheritance 20px", etc.

I have prepared a spreadsheet with different font string & comparing their behavior in Winlauncher, FF & Chrome. I will attach it soon.

> 
> We should also consider testing for invalid values like "default" and "initial" as well.  I assume we already handle those correctly?

If you see the result, the tests in winlauncher/Chrome are failing if the font string contains "inherit" (Case insensitive) OR "initial". Therefor I am checking for both "initial Or "initial" in the new font string.

Currently "default" works perfectly fine. For "20px initial" it works fine but for "initial" it fails. The reason being in CSSParser::parseValue(), we have the following code

    else if (id == CSSValueInitial) {
        if (num != 1)
            return false;
        addProperty(propId, CSSInitialValue::createExplicit(), important);
        return true;
    }

For "20px initial" num!=1 & therefore it returns false & ignores the parsed properties. For "initial" it adds the property(num=1) with value=0 & therefore I assume it is rendering with default font (10 Sans Serif)

> 
> If this is an unmodified test from a test suite, it's fine to add it as-is, but we should add it in whatever directory that imported test suite is.

I could not find any directory inside LayoutTests to contains test cases from http://w3c-test.org/html/tests/approved/canvas/ & therefore I have kept the test inside fast/

-- 
Configure bugmail: https://bugs.webkit.org/userprefs.cgi?tab=email
------- You are receiving this mail because: -------
You are the assignee for the bug.



More information about the webkit-unassigned mailing list