This moves restyling management out of nsCSSFrameConstructor (thus
reducing its size), and keeps the restyling code closer together.
This is the first of two big chunks of code moved in this patch series.
A later patch in this series will move related code from nsFrameManager
into the same destination file.
The fixes to the miniflush code
(nsTransitionManager::UpdateThrottledStyle and UpdateAllThrottledStyles)
fix the case where we constructed totally incorrect style contexts for
outer table frames (which have special style contexts inheriting from
the table frame) during the miniflush, leading to inconsistent style
data and other bad things, when we should have been touching the style
on the table frame instead.
The fixes to the other OMTA codepaths lead to layer tests being
performed on the same frame that the styles will be applied to, and
probably fix real bugs (which would occur when animating opacity or
transform on a table).
The fixes to the miniflush code
(nsTransitionManager::UpdateThrottledStyle and UpdateAllThrottledStyles)
fix the case where we constructed totally incorrect style contexts for
outer table frames (which have special style contexts inheriting from
the table frame) during the miniflush, leading to inconsistent style
data and other bad things, when we should have been touching the style
on the table frame instead.
The fixes to the other OMTA codepaths lead to layer tests being
performed on the same frame that the styles will be applied to, and
probably fix real bugs (which would occur when animating opacity or
transform on a table).
Note that this removes the Reset method and the mURL and mDocument
members (and arguments to set them) from nsHTMLCSSStyleSheet.
On the other hand, from nsHTMLStyleSheet it only removes mURL (and
equivalent arguments), and nsHTMLStyleSheet keeps the SetOwningDocument
method that was previously part of nsIStyleSheet (but no longer virtual).
- Implemented WebVTTLoadListener to manage the webvtt parser.
- TextTrackCue now handles the conversion of webvtt nodes to
anonymous content which will be displayed on the video div
overlay.
- HTMLTrackElement manages the lifetime of the WebVTTLoadListener.
In nsRuleNode, when we parse a number of columns from a CSS style sheet, we enforce
a rule that the number of columns cannot exceed 1000. This patch simply enforces
the same rule when we decide that the column count will be generated based on
the available width (e.g. the column-count property has a value of 'auto').
When 'font-family' is secified with unquoted identifiers rather than a quoted string, a single space should be inserted between identifier tokens, even if they are separated by a comment rather than a whitespace token.
The code in nsHTMLStyleSheet implements LangRule to map xml:lang into
style and the code to manage its uniqueness.
The change to nsGenericHTMLElement fixes the mapping of the HTML lang
attribute to do cascading the way all other rule mapping does so that
the cascading works correctly.
The tests test that the correct style language is used for hyphenation
by copying over a set of hyphenation reftests that check its basic
response to languages. There are no specific tests for font selection,
but font selection is known to use the same language data from style.
I verified manually (see other attachments to bug) that the rule
uniqueness is being managed correctly.
--HG--
rename : layout/reftests/text/auto-hyphenation-1.html => layout/reftests/text/auto-hyphenation-xmllang-1.xhtml
rename : layout/reftests/text/auto-hyphenation-10.html => layout/reftests/text/auto-hyphenation-xmllang-10.xhtml
rename : layout/reftests/text/auto-hyphenation-1.html => layout/reftests/text/auto-hyphenation-xmllang-11a.xhtml
rename : layout/reftests/text/auto-hyphenation-1.html => layout/reftests/text/auto-hyphenation-xmllang-11b.xhtml
rename : layout/reftests/text/auto-hyphenation-4.html => layout/reftests/text/auto-hyphenation-xmllang-12a.xhtml
rename : layout/reftests/text/auto-hyphenation-4.html => layout/reftests/text/auto-hyphenation-xmllang-12b.xhtml
rename : layout/reftests/text/auto-hyphenation-1.html => layout/reftests/text/auto-hyphenation-xmllang-13a.xhtml
rename : layout/reftests/text/auto-hyphenation-1.html => layout/reftests/text/auto-hyphenation-xmllang-13b.xhtml
rename : layout/reftests/text/auto-hyphenation-4.html => layout/reftests/text/auto-hyphenation-xmllang-14a.xhtml
rename : layout/reftests/text/auto-hyphenation-4.html => layout/reftests/text/auto-hyphenation-xmllang-14b.xhtml
rename : layout/reftests/text/auto-hyphenation-1a.html => layout/reftests/text/auto-hyphenation-xmllang-1a.xhtml
rename : layout/reftests/text/auto-hyphenation-2.html => layout/reftests/text/auto-hyphenation-xmllang-2.xhtml
rename : layout/reftests/text/auto-hyphenation-3.html => layout/reftests/text/auto-hyphenation-xmllang-3.xhtml
rename : layout/reftests/text/auto-hyphenation-4.html => layout/reftests/text/auto-hyphenation-xmllang-4.xhtml
rename : layout/reftests/text/auto-hyphenation-5.html => layout/reftests/text/auto-hyphenation-xmllang-5.xhtml
rename : layout/reftests/text/auto-hyphenation-6.html => layout/reftests/text/auto-hyphenation-xmllang-6.xhtml
rename : layout/reftests/text/auto-hyphenation-7.html => layout/reftests/text/auto-hyphenation-xmllang-7.xhtml
rename : layout/reftests/text/auto-hyphenation-8.html => layout/reftests/text/auto-hyphenation-xmllang-8.xhtml
rename : layout/reftests/text/auto-hyphenation-9.html => layout/reftests/text/auto-hyphenation-xmllang-9.xhtml
We set the class caption-box class on the anonymous div which
acts as the parent container for the captions overlayed on
the video frame.
caption-text isn't currently used, but will be applied to
child elements by when bug 833382 lands.
Note that this patch has a little bit of a belt-and-braces aspect to it.
In each file, either one of the changes should be sufficient, but one of
them prevents us from doing unneeded work and the other one ensures that
we never apply style resulting from transitions and animations even if
somehow we do that work.
Also note that the tests don't actually test anything usefully, since
the reftest harness doesn't currently make the pres context non-dynamic.
(Thus they're marked as failing.) I'm not sure what I should do about
that, though I'm considering just deleting the tests entirely.
