Bug 132753 – Subpixel rendering: Empty rects should remain empty after integrally enclosing them.

• Status: RESOLVED
• Resolution: FIXED
• Component: Layout and Rendering

[alan]

When the full screen transition's final rect is empty, enclosingIntRect turns it into a 1px width/height non-empty rect and stretch occurs.

[alan]

<rdar://problem/16846801>

[alan]

Created attachment 231191 [details]
Patch

[Simon Fraser (smfr)]

Comment on attachment 231191 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=231191&action=review

> Source/WebCore/ChangeLog:3
> +        Subpixel rendering: Empty rects should remain empty after integral enclosing them.

integrally?

[Build Bot]

Comment on attachment 231191 [details]
Patch

Attachment 231191 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/6214286661648384

New failing tests:
editing/selection/vertical-rl-rtl-extend-line-forward-br.html
fast/css/h1-in-section-elements.html
fast/forms/fieldset-align.html
tables/mozilla/other/wa_table_tr_align.html
fast/inline/continuation-outlines.html
fast/invalid/nestedh3s.html
editing/selection/vertical-rl-rtl-extend-line-backward-br.html
fast/multicol/client-rects-spanners-complex.html
fast/text/international/bidi-LDB-2-HTML.html
mathml/presentation/mo-stretch.html
editing/selection/vertical-rl-rtl-extend-line-backward-p.html
fast/encoding/utf-16-little-endian.html
fast/multicol/span/span-as-immediate-columns-child.html
fast/css/word-space-extra.html
fast/forms/textAreaLineHeight.html
fast/inline/outline-continuations.html
editing/selection/vertical-rl-rtl-extend-line-forward-p.html
fast/multicol/client-rects.html
fast/multicol/span/span-as-immediate-columns-child-dynamic.html
fast/css/css1_forward_compatible_parsing.html
svg/wicd/rightsizing-grid.xhtml
fast/inline/continuation-outlines-with-layers.html
fast/encoding/utf-16-big-endian.html
fast/css/first-line-text-decoration.html
fast/multicol/newmulticol/client-rects.html
fast/multicol/client-rects-spanners.html
tables/mozilla/other/wa_table_thtd_rowspan.html
fast/multicol/span/span-margin-collapsing.html
fast/overflow/clip-rects-fixed-ancestor.html
fast/table/014.html

[Build Bot]

Created attachment 231198 [details]
Archive of layout-test-results from webkit-ews-02 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-02  Port: mac-mountainlion  Platform: Mac OS X 10.8.5

[alan]

(In reply to comment #4)
> (From update of attachment 231191 [details])
> Attachment 231191 [details] did not pass mac-ews (mac):
> Output: http://webkit-queues.appspot.com/results/6214286661648384
> 
> New failing tests:
> editing/selection/vertical-rl-rtl-extend-line-forward-br.html
> fast/css/h1-in-section-elements.html
> fast/forms/fieldset-align.html
> tables/mozilla/other/wa_table_tr_align.html
> fast/inline/continuation-outlines.html
> fast/invalid/nestedh3s.html
> editing/selection/vertical-rl-rtl-extend-line-backward-br.html
> fast/multicol/client-rects-spanners-complex.html
> fast/text/international/bidi-LDB-2-HTML.html
> mathml/presentation/mo-stretch.html
> editing/selection/vertical-rl-rtl-extend-line-backward-p.html
> fast/encoding/utf-16-little-endian.html
> fast/multicol/span/span-as-immediate-columns-child.html
> fast/css/word-space-extra.html
> fast/forms/textAreaLineHeight.html
> fast/inline/outline-continuations.html
> editing/selection/vertical-rl-rtl-extend-line-forward-p.html
> fast/multicol/client-rects.html
> fast/multicol/span/span-as-immediate-columns-child-dynamic.html
> fast/css/css1_forward_compatible_parsing.html
> svg/wicd/rightsizing-grid.xhtml
> fast/inline/continuation-outlines-with-layers.html
> fast/encoding/utf-16-big-endian.html
> fast/css/first-line-text-decoration.html
> fast/multicol/newmulticol/client-rects.html
> fast/multicol/client-rects-spanners.html
> tables/mozilla/other/wa_table_thtd_rowspan.html
> fast/multicol/span/span-margin-collapsing.html
> fast/overflow/clip-rects-fixed-ancestor.html
> fast/table/014.html
These are all progressions (1px->0px). They need rebaselining.

[alan]

Created attachment 231200 [details]
Patch

[alan]

Comment on attachment 231200 [details]
Patch

EWS testing.

[Build Bot]

Comment on attachment 231200 [details]
Patch

Attachment 231200 [details] did not pass mac-wk2-ews (mac-wk2):
Output: http://webkit-queues.appspot.com/results/5490639164669952

New failing tests:
fast/forms/fieldset-align.html
fast/forms/textAreaLineHeight.html

[Build Bot]

Created attachment 231204 [details]
Archive of layout-test-results from webkit-ews-16 for mac-mountainlion-wk2

The attached test failures were seen while running run-webkit-tests on the mac-wk2-ews.
Bot: webkit-ews-16  Port: mac-mountainlion-wk2  Platform: Mac OS X 10.8.5

[Build Bot]

Comment on attachment 231200 [details]
Patch

Attachment 231200 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/5972646164430848

New failing tests:
fast/forms/fieldset-align.html
fast/forms/textAreaLineHeight.html

[Build Bot]

Created attachment 231206 [details]
Archive of layout-test-results from webkit-ews-07 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-07  Port: mac-mountainlion  Platform: Mac OS X 10.8.5

[Build Bot]

Comment on attachment 231200 [details]
Patch

Attachment 231200 [details] did not pass mac-ews (mac):
Output: http://webkit-queues.appspot.com/results/5039919122612224

New failing tests:
fast/forms/fieldset-align.html
fast/forms/textAreaLineHeight.html

[Build Bot]

Created attachment 231208 [details]
Archive of layout-test-results from webkit-ews-08 for mac-mountainlion

The attached test failures were seen while running run-webkit-tests on the mac-ews.
Bot: webkit-ews-08  Port: mac-mountainlion  Platform: Mac OS X 10.8.5

[alan]

Created attachment 231210 [details]
Patch

[alan]

Created attachment 231218 [details]
Patch

[WebKit Commit Bot]

Comment on attachment 231218 [details]
Patch

Clearing flags on attachment: 231218

Committed r168575: <http://trac.webkit.org/changeset/168575>

[WebKit Commit Bot]

All reviewed patches have been landed.  Closing bug.

[Darin Adler]

Comment on attachment 231218 [details]
Patch

View in context: https://bugs.webkit.org/attachment.cgi?id=231218&action=review

> Source/WebCore/platform/graphics/LayoutRect.cpp:136
> +    // Empty rects with fractional x, y values turn into non-empty rects when converting to enclosing.
> +    // We need to ensure that empty rects stay empty after the conversion, because the selection code expects them to be empty.
>      IntPoint location = flooredIntPoint(rect.minXMinYCorner());
> -    IntPoint maxPoint = ceiledIntPoint(rect.maxXMaxYCorner());
> -
> +    IntPoint maxPoint = IntPoint(rect.width() ? rect.maxX().ceil() : location.x(), rect.height() ? rect.maxY().ceil() : location.y());
>      return IntRect(location, maxPoint - location);

This defines an empty rect as "width or height is zero" and independently applies this rule to each dimension. That’s one way of doing it.

Another would be to use isEmpty() to define emptiness:

    IntPoint location = flooredIntPoint(rect.minXMinYCorner());
    if (rect.isEmpty())
        return IntRect(location, IntSize());
    return IntRect(location, ceiledIntPoint(rect.maxXMaxYCorner()) - location);