Bug 103453 – LayoutTests/fast/table/css-table-max-height.html failing on Qt Linux

• Component: Tables
• Status: RESOLVED FIXED

[Pravin D]

The test case LayoutTests/fast/table/css-table-max-height.html is failing on Qt linux.
Diff
http://build.webkit.org/results/Qt%20Linux%20Release/r135895%20%2854971%29/fast/table/css-table-max-height-pretty-diff.html

The test passes on other platforms.

[Julien Chaffraix]

Test disabled in http://trac.webkit.org/changeset/135917

Failures:

This sub-test checks that when content height is greater than max-height, content height is applied to the table with auto layout.
FILLER TEXT TO INCREASE CONTENT HEIGHT.
FAIL:
Expected 192 for height, but got 204. 

<div class="container">
    <div class="child" style="max-height:100px;" data-expected-height="192">
        This sub-test checks that when content height is greater than max-height, content height is applied to the table with auto layout.
        <br><br>FILLER TEXT TO INCREASE CONTENT HEIGHT.
    </div>
</div>
[snip]
This sub-test checks that when content height is greater than max-height, content height is applied to a table with fixed layout. 

FILLER TEXT TO INCREASE CONTENT HEIGHT.
FAIL:
Expected 192 for height, but got 204. 

<div class="container">
    <div class="child fixed-table" style="max-height:100px;" data-expected-height="192">
        This sub-test checks that when content height is greater than max-height, content height is applied to a table with fixed layout.
        <br><br>FILLER TEXT TO INCREASE CONTENT HEIGHT.
    </div>
</div>
This test checks that when height of a fixed table is auto, there is are no asserts. No crash means the test passed.
PASS

[Pravin D]

Created attachment 176420 [details]
Patch

[Pravin D]

Patch
https://bugs.webkit.org/attachment.cgi?id=176420

The default line-height used by Qt-linux is different from the other platforms. This was causing the height of the sub-test cases to be different from expected value.

Have also fixed the font-size value, to be on a safer side.

@Julien: thanks for the hunch :)

[Simon Hausmann]

Julien, can you review this patch? It looks sane to me, but I don't feel comfortable messing with the table tests that I don't know at all :)

Thanks :)

[Julien Chaffraix]

Comment on attachment 176420 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=176420&action=review

> LayoutTests/fast/table/css-table-max-height.html:11
>      font-family:ahem;
> +    line-height:1;
> +    font-size:16px;

You can use the shorthand here:

font: 16px/1 Ahem;

It's more compact thus better.

[Pravin D]

Created attachment 176527 [details]
Patch

[Pravin D]

(In reply to comment #5)
> (From update of attachment 176420 [details])
> View in context: https://bugs.webkit.org/attachment.cgi?id=176420&action=review
> 
> > LayoutTests/fast/table/css-table-max-height.html:11
> >      font-family:ahem;
> > +    line-height:1;
> > +    font-size:16px;
> 
> You can use the shorthand here:
> 
> font: 16px/1 Ahem;
> 
> It's more compact thus better.
> 
Done..

[WebKit Review Bot]

Comment on attachment 176527 [details]
Patch

Clearing flags on attachment: 176527

Committed r136041: <http://trac.webkit.org/changeset/136041>

[WebKit Review Bot]

All reviewed patches have been landed.  Closing bug.

[Darin Adler]

(In reply to comment #3)
> The default line-height used by Qt-linux is different from the other platforms.

Really? Why?

[Julien Chaffraix]

(In reply to comment #10)
> (In reply to comment #3)
> > The default line-height used by Qt-linux is different from the other platforms.
> 
> Really? Why?

It seems to be a matter of used font which impacts the height of a line. I remember hitting this issue myself and bug 74219 confirms it too. Regardless of what I just said, it's always good to limit the test variation on the font / line-height selection as it reduces platform-specificity.