Граф коммитов

10726 Коммитов

Автор SHA1 Сообщение Дата
Jonathan Kew f8c40475ef Bug 1105268 - part 1 - Fix the logical/physical confusion with min- and max- dimensions in nsFrame::ComputeSize(). r=smontagu 2014-11-27 15:01:11 +00:00
Daniel Holbert 0f8922ac4a Bug 1103184 part 1: Honor "object-fit" & "object-position" on <canvas> element. r=roc 2014-11-26 19:43:36 -08:00
Ryan VanderMeulen 792273e467 Merge inbound to m-c. a=merge 2014-11-26 20:34:29 -05:00
Daniel Holbert 621d56400d Bug 1104354: Give nsHTMLCanvasFrame its own GetIntrinsicSize() impl. r=mats 2014-11-26 11:29:49 -08:00
Jonathan Kew 1c2aa200a8 Bug 1096224 - part 2 - Fix writing mode mismatch in margin calculation in PlaceBlock. r=smontagu 2014-11-26 10:24:16 +00:00
Jonathan Kew 6ce2710325 Bug 1096224 - part 1 - CalculateBlockSideMargins should be working in the containing block's writing mode. r=smontagu 2014-11-26 10:24:16 +00:00
Kartikaya Gupta baccad357d Bug 1104809 - When building display lists for non-stacking children don't discard the event regions. r=mstange,tn 2014-11-25 17:05:58 -05:00
Kartikaya Gupta 4a94a25a0d Bug 1104266 - Avoid needless temporary nsACString. r=mattwoodrow 2014-11-25 08:45:19 -05:00
Seth Fowler 626d39ccdc Bug 1102617 - Replace imgIContainer::FrameIsOpaque with imgIContainer::IsOpaque. r=tn 2014-11-24 23:42:43 -08:00
Xidorn Quan c23c15eb67 Bug 1084183 - Propagate text decoration to ruby frames. r=dbaron 2014-11-26 15:53:18 +11:00
Xidorn Quan e464df4013 Bug 1052924 - Implement basic line breaking for ruby. r=dbaron
Known problem:
It would cause infinite loop if there is any line break happens inside
ruby base or annotation, or the width of container is not enough for
the widest pair/span. This might be fixed in bug 1098272.
2014-11-26 15:52:50 +11:00
Xidorn Quan 5f4cd8dd0b Bug 1052924 - Separate reflow code. r=dbaron 2014-11-26 15:52:50 +11:00
Xidorn Quan 43f5510c25 Bug 1052924 - Give rbc & rtc correct position. r=dbaron
Important changes:
  * Use line layout to reflow ruby base container
2014-11-26 15:52:49 +11:00
Xidorn Quan d25178b0db Bug 1052924 - Resolve some warnings. r=dbaron
Important changes:
  * Change base class of nsRuby{Base,Text}Frame to nsInlineFrame
  * Make ComputeSize of nsRubyFrame and nsRubyBaseContainerFrame behavior like inline frames
2014-11-26 15:52:49 +11:00
Xidorn Quan c25136319a Bug 1052924 - Rewrite reflow code. r=dbaron
Important changes:
  * Avoid using GetPrefISize on the ruby texts in
    nsRubyBaseContainerFrame::Reflow, since the size it produces might
    not match the size produced by Reflow.  The old code calls that on
    all the ruby texts to determine how big they are, then reflows all
    the ruby bases, and then reflows all the ruby texts.  The new code
    instead processes one pair at a time, and for each pair reflows the
    ruby texts and then the ruby base.
  * Change the base class of nsRubyTextContainerFrame from nsBlockFrame
    to nsContainerFrame, and stop constructing an nsBlockReflowState for
    its reflow.
  * Move the code for reflowing ruby texts from nsRubyTextContainerFrame
    and to nsRubyBaseContainerFrame.
  * Fix the regression that ruby text containers contain span are not
    reflowed properly. It is the regression introduced in patch 0.

Known regression:
  * This patch drops centering ruby base and annotation in pairs. This
    should be fixed in bug 1055676 (ruby-align).
2014-11-26 15:52:49 +11:00
Xidorn Quan e254b3d884 Bug 1052924 - Add assertion and comments pointing out that ruby base container frames do not have an mBlockRS, but that's OK since they are never the containing block for floats. r=dbaron 2014-11-26 15:52:49 +11:00
Xidorn Quan 1fbb3167e6 Bug 1052924 - Treat ruby text containers as inline elements for the purposes of finding a font inflation container so that ruby text gets the same font inflation as its ruby bases. r=dbaron 2014-11-26 15:52:48 +11:00
Xidorn Quan 458a202ef8 Bug 1052924 - Use unified enumerators to simplify code. r=dbaron 2014-11-26 15:52:48 +11:00
Xidorn Quan 9b0f833e9c Bug 1052924 - Rewrite pref isize and min isize computation. r=dbaron
Known regression:
This patch changes AppendTextContainer, and put aside text containers
which contain spanning annotations. This changes makes those text
containers not be reflowed by the current code. It will be fixed in
some later patch.
2014-11-26 15:52:48 +11:00
Daniel Holbert 82b8f74f2a Bug 1098417 part 2: Make nsImageFrame pass "object-position"-determined anchor-point to DrawSingleImage. r=seth 2014-11-25 16:46:14 -08:00
Daniel Holbert a03d3e37d5 Bug 1093316 part 1: Backout changeset aece7f9f944c (i.e. backout bug 1032922 part 2), to reflect CSSWG removing "flex-basis: main-size" from the flexbox spec. 2014-11-25 11:28:15 -08:00
Daniel Holbert 4d0b72d499 Bug 624647 followup: Minor cleanup in nsVideoFrame::BuildLayer(); restore a local variable which shouldn't have been removed. (no review)
An earlier patch for this bug (changeset 466d3ff030e6) removed local variable
"nsRect area", but then ended up effectively re-creating it further down in the
function as "contentBoxRect". (I added contentBoxRect as a late-breaking
change, as part of addressing a review comment.)

I've now realized that contentBoxRect is the same rect as "area", and it's
simpler to just declare the original "area" rect at the function's start
instead of dynamically creating it further down.

(Not requesting review, since this is a no-functional-change tweak, which is
following up on a change that I'd made in response to review feedback on a r+'d
patch.)
2014-11-25 15:40:02 -08:00
Jonathan Kew 9a4427f454 Bug 1103613 - Use better inline-size for an orthogonal flow within a block of specified dimensions. r=smontagu 2014-11-24 21:44:06 +00:00
Mats Palmgren 98fa65732a Bug 1102791 - Always return NS_FRAME_COMPLETE status from nsHTMLButtonControlFrame::Reflow. Don't create overflow container continuations unless the delegating frame has eCanContainOverflowContainers. r=roc 2014-11-24 16:42:31 +00:00
Jonathan Watt 9910919cdd Bug 1103623 - Port most remaining gfxContext::Fill() calls to Moz2D. r=mattwoodrow 2014-11-24 00:04:33 +00:00
Jonathan Watt 8cf9f138a9 Bug 1101571 - List-style discs are all wobbly during vertical transitions. r=mattwoodrow 2014-11-21 15:48:05 +00:00
Bob Clary a08e4446e4 Bug 622596 - crashtest. 2014-11-23 15:09:51 +00:00
Jesse Ruderman 5919efd1e0 Bug 472957 - crashtest. 2014-11-23 15:09:51 +00:00
Jesse Ruderman 4dbf029a32 Bug 430991 - crashtest. 2014-11-23 15:09:51 +00:00
Jesse Ruderman 47b0568d7c Bug 429960 - crashtests. 2014-11-23 15:09:51 +00:00
Jesse Ruderman 310e5bae93 Bug 429458 - crashtest. 2014-11-23 15:09:50 +00:00
Jonathan Kew 74d961eaf3 Bug 1077515 - part 6 - Create a new nsISelectionController::PhysicalMove command. r=roc 2014-11-22 14:39:03 +00:00
Jonathan Kew 9a00763a7b Bug 1077515 - part 3 - Change desiredX (nscoord) to desiredPos (nsPoint) in nsPeekOffsetStruct, to support maintaining either vertical or horizontal position on inter-line moves. r=roc 2014-11-22 14:39:03 +00:00
Jonathan Kew d4bd3503ab Bug 1077515 - part 1 - Eliminate use of keyCode parameters and values in nsFrameSelection. r=roc 2014-11-22 14:39:02 +00:00
David Anderson a7ce0b61d2 Put meta viewport support behind a pref (bug 1092450, r=kats). 2014-11-21 14:32:32 -05:00
Vladimir Vukicevic 01202baa97 Bug 1036604 - add VR display item and VR frame state bit; r=roc 2014-07-09 12:25:31 -07:00
David Anderson e025120099 Make sure event regions propagate to more layers. (bug 1101260 part 2, r=roc) 2014-11-20 16:58:19 -08:00
David Anderson 0965889407 Cache the current animated geometry root in nsDisplayListBuilder. (bug 1101260 part 1, r=roc) 2014-11-20 16:58:18 -08:00
Mats Palmgren 8f6f1c70f4 Bug 907396 - DEBUG only; make frame tree dumps print up to 2 style context parents in the unusual case when they are not the parent frame's style context (e.g. display:contents). r=bzbarsky 2014-11-20 18:24:10 +00:00
Mats Palmgren debd2f3e4d Bug 907396 - Replace GetParentStyleContextFrame with GetParentStyleContext which can return frame-less display:contents style contexts. r=bzbarsky 2014-11-20 18:24:10 +00:00
Mats Palmgren 96ad0f2c46 Bug 907396 - Frame constructor changes for display:contents. r=bzbarsky 2014-11-20 18:24:09 +00:00
Mats Palmgren 88b2e61eb5 Bug 399262, part 2 - Exclude <select> because ::first-letter is not supported there. r=roc 2014-11-20 13:07:58 +00:00
Mats Palmgren 426662ee0a Bug 399262, part 1 - The -moz-button-content pseudo is also allowed to have a ::first-letter. r=roc 2014-11-20 13:07:57 +00:00
Simon Montagu 2f9f1b57d2 Bug 1100071 patch 3: don't pass around bidi levels when we only need the direction, r=dholbert 2014-11-20 12:45:23 +02:00
Simon Montagu ad766a1fcb Bug 1100071 patch 2: add macros for common tests whether bidi level is odd and whether two bidi levels have the same parity, r=dholbert 2014-11-20 12:45:22 +02:00
Simon Montagu aafee71bb6 Bug 1100071 patch 1: change uint8_t to nsBidiLevel in a few more places, r=dholbert 2014-11-20 12:45:22 +02:00
Simon Montagu 39cdd85ae0 Bug 1097894: if the embedding level of a frame is unset, get the direction from style, r=dholbert 2014-11-20 12:45:22 +02:00
David Anderson 397b1a3520 Add inactive subframes to the dispatch-to-content region. (bug 1082594, r=tn,kats) 2014-11-19 20:24:15 -08:00
Daniel Holbert bacb520ce5 Bug 1101128 part 2: Make nsDisplayImage::GetOpaqueRegion() take object-fit & object-position into account. r=seth 2014-11-19 18:34:52 -08:00
Daniel Holbert 58dc4ebd39 Bug 1101128 part 1: Move nsDisplayImage::GetOpaqueRegion() to .cpp file, and label it as MOZ_OVERRIDE. r=seth 2014-11-19 18:34:50 -08:00