This change does the following:
* Introduce a new smart pointer called CounterStylePtr which either
holds an AnonymousCounterStyle strongly, or a named CounterStyle
managed by CounterStyleManager weakly, and use it to replace all
RefPtr<CounterStyle> around the codebase.
* Rename CounterStyleManager::mCacheTable to mStyles to reflect the
fact that it is used to manage all styles, not just for caching.
* Add a retired styles list which collect all named CounterStyle
evicted from mStyles, and post a PostRefreshObserver to destroy
objects in that list after next flush.
* Remove helper functions for counter style in nsStyleList and expose
mCounterStyle directly, to make code simpler with the new pointer.
Reason for adding a new smart pointer type rather than making their
AddRef/Release behave like BuiltinCounterStyle is that, it is possible
that after a flush, some stale style structs may still be alive. They
can contain pointer to destroyed CounterStyle objects. Although the
actual content may never be accessed anymore, RefPtr may still access
the object for refcounting during destruction.
MozReview-Commit-ID: xxegwSDhNb
--HG--
extra : rebase_source : bb5443f0eb56eee51cbdfd08e0400335648610e8
This avoids conflicts with mozilla::dom::FrameType.
MozReview-Commit-ID: 7aEMbHRaTFk
--HG--
extra : rebase_source : 2d01321f5ce0ec8c0e3f70984674f82678034b3c
To print log in nsBlockFrame, run "GECKO_BLOCK_DEBUG_FLAGS=reflow ./mach run".
Here is a sample nsReflowStatus string:
[Complete=Y,NIF=N,Truncated=N,Break=N,FirstLetter=Y].
MozReview-Commit-ID: 4voGcXfJlN7
--HG--
extra : rebase_source : f84ba1ef47e30513674e578fab37b66a8b2a1ab3
This patch is written by the following script with some manual adjustment to
the comment in nsRubyTextContainerFrame.cpp and nsRubyFrame.cpp, and
nsColumnSetFrame's constructor.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename "nsReflowStatus *([a-zA-Z0-9]*) = NS_FRAME_COMPLETE" "nsReflowStatus \1"
rename "([a-zA-Z0-9.*]*) *= NS_FRAME_COMPLETE;" "\1.Reset();"
rename "([a-zA-Z0-9.*]*) == NS_FRAME_COMPLETE" "\1.IsEmpty()"
MozReview-Commit-ID: 9tqQAHvdQex
--HG--
extra : rebase_source : 3119776946dc2c8350098b7bf9f3ceff29bdffb5
NS_INLINE_LINE_BREAK_BEFORE() was replaced by the help of the following script.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename " = NS_INLINE_LINE_BREAK_BEFORE\(\);" ".SetInlineLineBreakBeforeAndReset();"
MozReview-Commit-ID: mz6L8zay7q
--HG--
extra : rebase_source : fddffa1288c7a52c7245aec0839a65f8d1d3f008
Also, remove NS_FRAME_TRUNCATED and NS_FRAME_REFLOW_NEXTINFLOW because both
are used only by NS_MergeReflowStatusInto().
MozReview-Commit-ID: LsPOji9j2e
--HG--
extra : rebase_source : 7cf1421c066f0f8df3e8402b84ae0f584cad2d11
This patch is written by the help of the following script.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename "NS_FRAME_SET_OVERFLOW_INCOMPLETE\(\*([a-zA-Z0-9.*]*)\)" "\1->SetOverflowIncomplete()"
rename "NS_FRAME_SET_OVERFLOW_INCOMPLETE\(([a-zA-Z0-9.*]*)\)" "\1.SetOverflowIncomplete()"
MozReview-Commit-ID: EJOIs84vwev
--HG--
extra : rebase_source : 59a5ac827f57e7a0e50910f5a813c44560baeb00
This patch is written by the help of the following script.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename "NS_FRAME_SET_INCOMPLETE\(\*([a-zA-Z0-9.*]*)\)" "\1->SetIncomplete()"
rename "NS_FRAME_SET_INCOMPLETE\(([a-zA-Z0-9.*]*)\)" "\1.SetIncomplete()"
MozReview-Commit-ID: GOd4y2N6dcz
--HG--
extra : rebase_source : 185f5d6f5a3c8306761860c579eff10d931f3b35
This patch is written by the help of the following script.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename "NS_FRAME_OVERFLOW_IS_INCOMPLETE\(([a-zA-Z0-9.*]*)\)" "\1.IsOverflowIncomplete()"
MozReview-Commit-ID: GOd4y2N6dcz
--HG--
extra : rebase_source : 2945ff0274d91137249eb2b5a053eeec08864a25
This patch is written by the help of the following script.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename "NS_FRAME_IS_COMPLETE\(([a-zA-Z0-9.*]*)\)" "\1.IsComplete()"
MozReview-Commit-ID: GOd4y2N6dcz
--HG--
extra : rebase_source : aa8b11d3a756c9e7c521e6ffd70713af0174bd98
This patch is written by the help of the following script.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename "NS_FRAME_IS_FULLY_COMPLETE\(([a-zA-Z0-9.*]*)\)" "\1.IsFullyComplete()"
MozReview-Commit-ID: GOd4y2N6dcz
--HG--
extra : rebase_source : 200639e836cebe26fd77cde21f478fd027e1725f
This patch is written by the help of the following script.
function rename() {
find layout\
-type f\
\( -name "*.cpp" -or\
-name "*.h" \)\
-exec sed -i -r "s/$1/$2/g" "{}" \;
}
rename "NS_FRAME_IS_NOT_COMPLETE\(([a-zA-Z0-9.*]*)\)" "\1.IsIncomplete()"
MozReview-Commit-ID: GOd4y2N6dcz
--HG--
extra : rebase_source : 412b60e7954118d8443ebf2a786046d7cd38c516
From the "NS_ASSERTION(mFloatManager)" statement in BlockReflowInput's
constructor, we know that BlockReflowInput's mFloatManager is always valid
and equals to aReflowInput.mFloatManager. Therefore, we could just use
ReflowInput's float manager in BlockReflowInput.
Due to the removal of BlockReflowInput's mFloatManager, the logic which
resets mFloatManager near the end of nsBlockFrame::Reflow() is removed as
well. It's safe because beyond that point, no other logic involves floats,
and |state| (i.e. BlockReflowInput) lives only on the stack.
MozReview-Commit-ID: 3dwXMnWkEI6
--HG--
extra : rebase_source : 7f9af1af10fd54456450b23bc0004dd5f15db4e4
This patch makes ForceBreak() partially clear floats according to the
break type of the coming block.
MozReview-Commit-ID: 71Gl9lBoTJ5
--HG--
extra : rebase_source : 5ca01565f607241df0c63a7cd64c35ac7ff7648f
Per spec, float positioning and stacking is not affected by defining a float
area with a shape.
https://drafts.csswg.org/css-shapes/#relation-to-box-model-and-float-behavior
So all the call sites of GetFloatAvailableSpace() related to adding a
float are replaced by GetFloatAvailableSpaceForPlacingFloat().
<shape-box> with border-radius will be implemented in next part.
MozReview-Commit-ID: 1RXEeXDhdWo
--HG--
extra : rebase_source : 42cdb0c81b16168e4e30ee2261ceccb562e278cf
In nsBlockFrame::PlaceLine(), we query the float available space by
using the line's BSize(), which may cause the line to reflow again due
to available space shrunk.
The first issue lies in the second reflow. That is, we do not leverage
the line's BSize() computed in the first reflow to query the float
available space when updating the inline reflow engine in
BlockReflowInput::AddFloat(). So some tall inline elements could still
overlap the floats as in the first reflow.
To solve this, we cache current line's BSize so that it could be
used to update the inline reflow engine when redo the line.
Another issue is in nsBlockFrame::PlaceLine(). When determined whether
the available space is shrunk, we use the float manager's state *before*
placing the line. So if current line has floats, they're not considered.
To solve this, we use the current set of floats to get the float available
space for comparison, and leave the original aFloatAvailableSpace to provide
the information when redoing the line.
MozReview-Commit-ID: GqqNlphgxYS
--HG--
extra : rebase_source : e2c64ab1ac363c7a08e532dc043bee69d6455049
This patch is based on bug 1291110 comment 29, but |aCanGrow| is preserved,
since the set of float available spaces we'll be testing in Part 4 are
difference from the one we enforced here.
MozReview-Commit-ID: LswbIaRukVR
--HG--
extra : rebase_source : 007d55fd867d052163ea5bb4222baf201fe30bcf
Rename to make the variable name consistent with the writing-mode
terminology used in PlaceLine().
MozReview-Commit-ID: AvhwHyHz7hV
--HG--
extra : rebase_source : bcec67d9959c1b881b4d287dbbab47dd0d6e0de9
1. Rename these functions to agree with Mozilla coding style.
2. Use singular naming instead of plural naming since each of these functions
returns an iterator pointing to a singular line.
3. Rename line() and rline() to BeginLineFrom() and RBeginLineFrom(), which
shall improve the readability.
MozReview-Commit-ID: txZjVnv9Yb
--HG--
extra : rebase_source : b4dae99ab7bf751bd9974616f8f8e6dfe6248fc2
Since we're converting LineReflowStatus to an enum class, the "default:" branch
of LineReflowStatusToString() could be removed as well. In this way, we can add
protection at compile time (warning) instead of runtime (assertion).
MozReview-Commit-ID: 8Ul0eCxqSTR
--HG--
extra : rebase_source : 81cc36bf1ae566d7114466661b7d0db6d63e364f
This is a pre-patch that uses LineReflowStatusToString() function instead of
LineReflowStatusNames array to support debugging. With this patch, we could
easily regain debugging support from LineReflowStatusToString() while converting
LineReflowStatus to an enum class.
MozReview-Commit-ID: K2AzZ2zI69p
--HG--
extra : rebase_source : bfbb1221ffb59cd4c27c3fb147a3a3b64e86cf68
Per bug 1299753 comment 20, SetFlags() is designed for flag propagation
to continuations during block reflow. To avoid misuse, I expand
SetFlags() directly in the only reasonable caller nsBlackFrame::Init(),
and replace other usages by AddStateBits().
MozReview-Commit-ID: GsbE2Z0Rps1
--HG--
extra : rebase_source : 72a64e9218870d638f67d1b586f533cd7d16c491
nsCSSFrameConstructor::ConstructNonScrollableBlock() has logic to
determine whether to create a block formatting context for a block
frame. I refactor the function to make it reusable by
nsCSSFrameConstructor::ConstructDetailsFrame().
Also, make NS_NewBlockFrame() accept two arguments as other frame
factory functions so that it could be pointed by BlockFrameCreationFunc.
NS_NewBlockFormattingContext is changed accordingly.
The construction for a scrollable DetailsFrame will be further revised
in Part 3.
MozReview-Commit-ID: 8TwG9YMyGva
--HG--
extra : rebase_source : fffdd974df81a809a607491d2534aa8dd2d13ab1
After using enum class, a switch-case warning in CombineBreakType is caught.
This is one of such kind safty checks that we would like to gain.
Fix it by adding default case for switch-case in CombineBreakType.
MozReview-Commit-ID: BdS3LPN6qzX
--HG--
extra : rebase_source : 17f24a0d482ed6eb51b23e6942d0ac1c87375e0b
Xidorn Quan
Emilio Cobos Álvarez
Brad Werth
Mats Palmgren
cku
Ting-Yu Lin
Phil Ringnalda
Mats Palmgren
Jeremy Chen
Manish Goregaokar