masayuki@d-toybox.com
|
cb5c20d581
|
Bug 392785 overflowed underline sometimes is not repainted at scrolling r+sr=roc b1.9=dbaron
|
2008-02-16 07:31:27 -08:00 |
|
roc+@cs.cmu.edu
|
b278a54a54
|
Bug 400813. Text with font-size:0 should not make a line be treated as empty. r+sr=dbaron
|
2008-02-14 01:08:33 -08:00 |
|
roc+@cs.cmu.edu
|
79bbde3785
|
backing out bug 400813 again due to reftest failure...
|
2008-02-11 18:48:17 -08:00 |
|
roc+@cs.cmu.edu
|
4251201001
|
Bug 400813. Lines with font-size:0 content should not be treated as empty. r+sr=dbaron
|
2008-02-11 17:42:14 -08:00 |
|
roc+@cs.cmu.edu
|
dada6c8a8d
|
Backing out 400813, sigh
|
2008-01-29 23:35:05 -08:00 |
|
roc+@cs.cmu.edu
|
81d839b21e
|
Bug 400813. Lines with actual content that just happens to be zero-width should be treated as non-empty. r+sr=dbaron
|
2008-01-29 23:12:26 -08:00 |
|
roc+@cs.cmu.edu
|
926a7142d9
|
Bug 411792. Never put a line break before a preformatted newline. r=smontagu
|
2008-01-26 23:48:39 -08:00 |
|
mats.palmgren@bredband.net
|
5eb8d1615d
|
Update the content offset for existing non-fluid continuations when creating a next-in-flow. b=406380 r+sr=roc a=blocking1.9
|
2008-01-04 09:09:35 -08:00 |
|
roc+@cs.cmu.edu
|
fb28ffc853
|
Bug 407889. Reduce dynamic memory allocations during event handling by having nested invocations of HitTest share a single buffer for reordering display items; this buffer is an nsAutoTArray so it's mostly stack-allocated. r+sr=mats
|
2008-01-03 18:08:29 -08:00 |
|
mats.palmgren@bredband.net
|
56d920eba2
|
Backing out bug 406380 to fix reftest failure.
|
2007-12-31 03:32:07 -08:00 |
|
mats.palmgren@bredband.net
|
4aa25eb7a7
|
Update content offset for all next-continuations, not just fluid ones. b=406380 r+sr=roc a=blocking1.9
|
2007-12-31 02:29:20 -08:00 |
|
Evan.Yan@Sun.COM
|
bc99bfe327
|
Bug 405248 - Space between embedded objects characters is missing r=aaronleventhal a=mtschrep
|
2007-12-19 04:57:08 -08:00 |
|
vladimir@pobox.com
|
1f7f3e59ed
|
b=406458; code cleanup: add new ThebesContext() call as opposed to NATIVE_THEBES_CONTEXT hackery, r=stuart
|
2007-12-18 15:01:15 -08:00 |
|
smontagu@smontagu.org
|
89837357fe
|
Make sure that the offsets in a textframe are up-to-date before creating a bidi continuation. Bug 406675, r=uriber, sr=roc, blocking1.9=schrep
|
2007-12-12 23:51:27 -08:00 |
|
roc+@cs.cmu.edu
|
8a54485b51
|
Bug 407078. Don't allow a trailing break in a text run that is followed by a <br>. r=smontagu
|
2007-12-10 17:19:20 -08:00 |
|
roc+@cs.cmu.edu
|
991719200d
|
Bug 407155. Backing out 392809 to fix regressions. a=schrep
|
2007-12-07 01:35:44 -08:00 |
|
roc+@cs.cmu.edu
|
deaf474141
|
Bug 407227. Include end-of-linebreaker-textrun break opportunity in min-width calculation. r+sr=dbaron,a=schrep
|
2007-12-06 19:27:01 -08:00 |
|
roc+@cs.cmu.edu
|
079aadfbec
|
Bug 407227. Include end-of-linebreaker-textrun break opportunity in min-width calculation. r+sr=dbaron,a=schrep
|
2007-12-06 19:16:07 -08:00 |
|
roc+@cs.cmu.edu
|
3d895dcd14
|
Bug 405577. Allow available width to go negative for inlines/text, so we can tell whether zero-width content fits or not. Also allow textframes to honour a forced break at the end of the text. r+sr=dbaron
|
2007-12-05 17:24:03 -08:00 |
|
roc+@cs.cmu.edu
|
264008d558
|
Backing out unintentionally-checked in stuff
|
2007-12-04 18:12:40 -08:00 |
|
roc+@cs.cmu.edu
|
ea37d818bc
|
Bug 392809. Break between whitespace and following punctuation even when stop_at_punctuation is false. Don't stop between punctuation and following non-punctuatino if the preceding punctuation follows white-space. r=smontagu
|
2007-12-04 18:11:17 -08:00 |
|
dmose@mozilla.org
|
a7aaa3bf1f
|
Backing out fix for bug 392785 as per roc + mconnor.
|
2007-12-04 15:32:08 -08:00 |
|
roc+@cs.cmu.edu
|
3577d8dad0
|
Bug 403426. We should clear whitespace status when we reset the linebreaker. If a final break opportunity exists we should save it and forward it to the line layout. Relanding with a fix so hopefully we won't crash Tp this time. r=smontagu
|
2007-12-03 00:22:07 -08:00 |
|
masayuki@d-toybox.com
|
db9a1f0734
|
Bug 392785 overflowed underline sometimes is not repainted at scrolling r+sr=roc, b1.9=dbaron
|
2007-12-02 22:51:57 -08:00 |
|
roc+@cs.cmu.edu
|
d77a6b0058
|
Relanding bug 404624 because it was not, in fact, responsible for that assert
|
2007-12-02 20:18:37 -08:00 |
|
roc+@cs.cmu.edu
|
b2e3b8803a
|
Backing out fix for bug 404624 because I think it's causing a mochitest to assert
|
2007-12-02 18:20:22 -08:00 |
|
roc+@cs.cmu.edu
|
26f2048d33
|
Relanding fix for bug 404624 because it had no effect on Tp
|
2007-12-02 17:42:52 -08:00 |
|
roc+@cs.cmu.edu
|
202376d9be
|
Backing out bug 404624 to see if it fixed Tp regression on bl-bldxp01
|
2007-12-02 12:47:41 -08:00 |
|
roc+@cs.cmu.edu
|
1a6bd11ee8
|
Bug 399531. Rework TrimTrailingWhitespace so that we recompute the overflow area for trimmed textframes and so that soft hyphens in otherwise-empty textframes are activated. r+sr=dbaron
|
2007-12-01 01:24:24 -08:00 |
|
roc+@cs.cmu.edu
|
41d0d00457
|
Backing out bug 403426
|
2007-11-30 12:52:24 -08:00 |
|
roc+@cs.cmu.edu
|
a2464227d0
|
Bug 403426. Clear break-after-whitespace flags when we reset the linebreaker. Also rescue a trailing break opportunity when we reset the line breaker and notify the line layout that the opportunity exists. r=smontagu
|
2007-11-30 12:03:33 -08:00 |
|
roc+@cs.cmu.edu
|
6d826063d8
|
Bug 404624. Silence assertion failures by allowing PropertyProvider's content length to be PR_INT32_MAX. r=smontagu
|
2007-11-30 11:59:03 -08:00 |
|
smontagu@smontagu.org
|
5e7ee1a7dc
|
Prevent duplicate entries in the line-break array. Bug 404204, b+r+sr=roc
|
2007-11-26 00:27:06 -08:00 |
|
Olli.Pettay@helsinki.fi
|
2976ce8f18
|
Bug 368760, r=mats, sr=roc, a=blocking1.9+
|
2007-11-22 12:47:33 -08:00 |
|
smontagu@smontagu.org
|
eb7e623bfe
|
Use Unicode escapes instead of octal escapes to print non-ASCII text in frame tree dumps. Bug 403871, r+sr=roc, a=dsicore
|
2007-11-21 02:37:40 -08:00 |
|
roc+@cs.cmu.edu
|
5a449a18d5
|
Bug 394752. Skip over preformatted newlines when moving caret; the caret will appear at the end of a line when it is positioned at the start of the next line with a HINTLEFT. r=smontagu
|
2007-11-15 18:25:59 -08:00 |
|
roc+@cs.cmu.edu
|
4dcead5d4b
|
Bug 403455. Allow prewrap text to record 'last break opportunity'. r=smontagu
|
2007-11-15 17:48:26 -08:00 |
|
bzbarsky@mit.edu
|
ac8234d604
|
Don't compare lines from different blockframes' lists to each other. Also fix handling of text-decorations in in-flows of blocks with text-indent. Bug 393330, r+sr=roc, a=schrep
|
2007-11-14 22:12:17 -08:00 |
|
roc+@cs.cmu.edu
|
27f5bebbe9
|
Bug 389707. Rework text-transform:capitalize by putting logic to decide what to capitalize into nsLineBreaker. r=smontagu
|
2007-11-11 17:51:31 -08:00 |
|
dholbert@cs.stanford.edu
|
3c53007c53
|
Bug 398181: Use NSCoordSaturatingAdd/Subtract in 3 places where we might have unconstrained heights. r+sr+a=roc
|
2007-11-09 09:16:36 -08:00 |
|
reed@reedloden.com
|
948ef26c2b
|
Bug 400081 - "Leak gfxFont and gfxTextRun with font-size: 0" [p=cpearce r+sr=roc a1.9=stuart]
|
2007-11-09 00:04:19 -08:00 |
|
reed@reedloden.com
|
f968a1c4dc
|
Fixing bustage from landing of bug 397510.
|
2007-11-08 23:37:31 -08:00 |
|
reed@reedloden.com
|
27a0532bd8
|
Bug 397510 - Limit the number of lines we build textruns in BuildTextRuns (speeds up painting of huge pages) [p=roc r=smontagu a=blocking1.9+]
|
2007-11-08 23:27:18 -08:00 |
|
reed@reedloden.com
|
ec9b5160d0
|
Bug 385417 - Rework textrun glyph representation so we can handle clusters containing glyphs in different fonts [p=roc r=stuart r=smontagu a=blocking1.9+]
|
2007-11-08 22:27:23 -08:00 |
|
roc+@cs.cmu.edu
|
a1e0ed3605
|
Bug 403004. Let the text-transform style setting loop cross non-fluid continuation boundaries. r+sr=bz
|
2007-11-08 15:37:34 -08:00 |
|
reed@reedloden.com
|
50d128a352
|
Backout roc's patch for bug 385417, as it most likely caused bug 402990 and the numerous test failures.
|
2007-11-07 23:41:06 -08:00 |
|
reed@reedloden.com
|
00ceaa5363
|
Bug 391584 - Pass around a context string so we can detect word breaks at the beginning or end of frames [p=roc r=smontagu a=blocking1.9+]
|
2007-11-07 20:33:28 -08:00 |
|
reed@reedloden.com
|
2fc6943fb7
|
Bug 385417 - Rework textrun glyph representation so we can handle clusters containing glyphs in different fonts [p=roc r=stuart r=smontagu a=blocking1.9+]
|
2007-11-07 20:31:33 -08:00 |
|
reed@reedloden.com
|
5dbce5ddb5
|
Bug 336408 - Allow the caret to be positioned at the end of trimmed whitespace, as if the whitespace wasn't trimmed [p=roc r=smontagu sr=mrbkap a=blocking1.9+]
|
2007-11-07 20:07:00 -08:00 |
|
reed@reedloden.com
|
c6d0b4440c
|
Bug 400826 - We shouldn't build the styles array for transformed textruns assuming every frame in a flow has the same style; instead we should build it frame by frame. [p=roc r=smontagu a=blocking1.9+]
|
2007-11-07 19:52:33 -08:00 |
|
roc+@cs.cmu.edu
|
23a4a840a3
|
Bug 389666. The mTransformedTextOffset field of MappedFlow wasn't always being set up correctly, so let's just stop using it altogether. r=smontagu,a=vlad
|
2007-10-23 17:07:26 -07:00 |
|
masayuki@d-toybox.com
|
aad8b02ec5
|
Bug 399159 left edge is not aligned when text-align: justify; in Japanese paragraph (relanding) r+sr+a=roc
|
2007-10-21 01:10:34 -07:00 |
|
roc+@cs.cmu.edu
|
7a31877544
|
Bug 393096. Allow an element containing breakable whitespace to introduce a break opportunity no matter what the context. Also cleans up some trimming stuff and adds comprehensive whitespace breaking and trimming reftests. r+sr=dbaron
|
2007-10-20 00:30:26 -07:00 |
|
roc+@cs.cmu.edu
|
7210b8c4a3
|
Backing out bug 397510 again because again it looks like it might have caused a regression, and we don't need it for M9
|
2007-10-18 21:56:05 -07:00 |
|
dietrich@mozilla.com
|
ba0e963628
|
re-landing Bug 393758
|
2007-10-18 19:01:05 -07:00 |
|
reed@reedloden.com
|
e6ba209193
|
Fixing bustage from landing of bug 397510.
|
2007-10-18 18:33:52 -07:00 |
|
dietrich@mozilla.com
|
cd47b16d5a
|
Bug 397510 Scrolling in large file is very slow due to textrun reconstruction (re-landing for roc)
|
2007-10-18 18:19:11 -07:00 |
|
roc+@cs.cmu.edu
|
523af98b9b
|
Backing out bustage fix
|
2007-10-17 23:48:44 -07:00 |
|
gavin@gavinsharp.com
|
3d5815da5b
|
Back out the patch for bug 399159 because it's suspected of having caused the Tp regression
|
2007-10-17 10:03:09 -07:00 |
|
roc+@cs.cmu.edu
|
3d8ce0f47f
|
Backing out 393758 just in case it caused perf regression
|
2007-10-16 16:56:27 -07:00 |
|
roc+@cs.cmu.edu
|
62761abd1f
|
Backing out fix for bug 397510 to see if it fixes the Tp regression
|
2007-10-16 01:20:22 -07:00 |
|
roc+@cs.cmu.edu
|
c9fff8055b
|
Fixing bustage
|
2007-10-15 14:20:36 -07:00 |
|
roc+@cs.cmu.edu
|
85116646be
|
Bug 393758. Detect when we're about to create two flows with the same element in the same textrun (which violates our invariants), and prevent it by starting a new textrun for the second flow. r=smontagu
|
2007-10-15 14:15:35 -07:00 |
|
roc+@cs.cmu.edu
|
4ad51c6b43
|
Bug 397510. Limit the number of lines below the initial line for which we reconstruct textruns. Speeds up line painting near the top of huge documents. r=smontagu
|
2007-10-15 14:11:37 -07:00 |
|
masayuki@d-toybox.com
|
96a98b686f
|
Bug 399159 left edge is not aligned when text-align: justify; in Japanese paragraph r+sr+a=roc
|
2007-10-15 05:37:15 -07:00 |
|
roc+@cs.cmu.edu
|
32b36ae07b
|
Bug 397518. Wrap inline non-MathML children of MathML frames in anonymous blocks. r+sr=bzbarsky
|
2007-10-12 01:30:54 -07:00 |
|
roc+@cs.cmu.edu
|
cc8a388fa5
|
Bug 391584. Don't miss word-break opportunities at points where whitespace has collapsed away. Also, note that a word-break opportunity exists at the start of a frame when it starts with a space.
|
2007-10-10 15:32:51 -07:00 |
|
roc+@cs.cmu.edu
|
416b58ab83
|
Backing out bug 391584 because of test failures on non-Mac
|
2007-10-09 21:25:42 -07:00 |
|
roc+@cs.cmu.edu
|
f21c86924d
|
Fixing bustage
|
2007-10-09 19:30:48 -07:00 |
|
roc+@cs.cmu.edu
|
fca6dd9463
|
Bug 387414. Detect situations where the textrun doesn't map enough text to fully reflow the frame (which can happen during style changes to preformatted whitespace, for example) and rebuild textruns in that case. r=smontagu
|
2007-10-09 19:25:57 -07:00 |
|
roc+@cs.cmu.edu
|
0ced6f7aa5
|
Bug 391584. Don't miss word-break opportunities at points where whitespace has collapsed away. Also, note that a word-break opportunity exists at the start of a frame when it starts with a space.
|
2007-10-09 19:21:33 -07:00 |
|
roc+@cs.cmu.edu
|
f8fe365101
|
Bug 385635. Make text min-width calculation clear out old value of trimmable whitespace when the text contains non-collapsible whitespace. Also, set skipWhitespace to skip subsequent whitespace only if the trailing whitespace in the text is not preformatted. r=smontagu
|
2007-10-02 01:52:09 -07:00 |
|
roc+@cs.cmu.edu
|
ccb6bf3bdb
|
Bug 392829. Clean up nsLineLayout code that's not used (or at least used differently). r+sr+a=dbaron
|
2007-10-01 19:36:26 -07:00 |
|
roc+@cs.cmu.edu
|
1fb7f38610
|
Bug 397734. More parsimonious storage of glyph extents ... trying to fix private-bytes regression. r=vlad,a=pavlov
|
2007-10-01 18:47:40 -07:00 |
|
roc+@cs.cmu.edu
|
aa2f1bb4bf
|
Backing out
|
2007-10-01 15:11:23 -07:00 |
|
roc+@cs.cmu.edu
|
07952e214b
|
Bug 397734. More parsimonious storage of glyph extents ... trying to fix private-bytes regression. r=vlad,a=pavlov
|
2007-10-01 14:29:19 -07:00 |
|
roc+@cs.cmu.edu
|
7e025b85ef
|
Bug 393906. Make AdjustOffsetsForBidi maintain textframe invariants across the entire continuation chain, not just the fluid chain. r=smontagu
|
2007-09-27 16:37:46 -07:00 |
|
roc+@cs.cmu.edu
|
bfd939088e
|
Bug 96041. Compute precise glyph extents for high-quality text, and when requested. r=vlad,r+sr=dbaron
|
2007-09-23 19:19:14 -07:00 |
|
roc+@cs.cmu.edu
|
c11d00e32d
|
Backing out due to test bustage
|
2007-09-22 06:28:16 -07:00 |
|
roc+@cs.cmu.edu
|
d72a839c93
|
Bug 96041. Compute true glyph bounding metrics. r=vlad,dbaron,sr=dbaron
|
2007-09-22 05:43:52 -07:00 |
|
mozilla@weilbacher.org
|
6fb8876814
|
[OS/2] Bug 396661: fix OS/2 build break in nsTextFrameThebes by removing the unnecessary cast, r/sr/a=roc
|
2007-09-19 16:01:44 -07:00 |
|
roc+@cs.cmu.edu
|
03a7855701
|
Bug 393796. Be more careful about verifying that an existing textrun is okay to be reused again after a change to the frame tree. r=smontagu
|
2007-09-17 21:23:14 -07:00 |
|
roc+@cs.cmu.edu
|
5d758d9550
|
Bug 385607. Ensure that whitespace is collapsed across text-run boundaries across incremental updates like adding/removing frames in the middle of the textrun. r+sr+a=dbaron
|
2007-09-17 20:00:16 -07:00 |
|
roc+@cs.cmu.edu
|
6dca4e8957
|
Bug 394832. Remove overeager rounding that causes underlines to appear too close to the text. r=smontagu,a=dbaron
|
2007-09-17 19:48:19 -07:00 |
|
sayrer@gmail.com
|
3c534c5634
|
Bug 394542. we leak the blink timer and its timer in a cycle. Patch by David Baron. r/sr=roc, a=dsicore/mconnor
|
2007-09-11 13:06:44 -07:00 |
|
roc+@cs.cmu.edu
|
4f3ef06d64
|
Bug 386807. Remove duplicate elements from the linebreak array textrun construction parameter. r=smontagu,a-M8=damon
|
2007-09-06 14:54:33 -07:00 |
|
roc+@cs.cmu.edu
|
17a86da0cd
|
Bug 392435. Make ClearTextRun work no matter which frame of a continuation chain we call it on. r=smontagu,a=dbaron
|
2007-09-05 16:06:05 -07:00 |
|
roc+@cs.cmu.edu
|
fa88299620
|
Bug 393923. Fix text-offset calculation for first-letter. r=smontagu,a=dbaron
|
2007-09-05 16:03:31 -07:00 |
|
roc+@cs.cmu.edu
|
d832331217
|
Bug 394510. Don't leak the hyphen textrun when we display a soft hyphen. r=smontagu
|
2007-09-05 15:59:44 -07:00 |
|
roc+@cs.cmu.edu
|
5901370c90
|
Bug 391621. Remove unnecessary trimmed-space counting. r=smontagu,a=dbaron
|
2007-09-03 21:02:22 -07:00 |
|
masayuki@d-toybox.com
|
5a84442569
|
Bug 394109 Inconsistent text-decoration positioning. r+sr+a=roc
|
2007-09-03 08:27:26 -07:00 |
|
masayuki@d-toybox.com
|
bfe2436be6
|
Oops, backing-out previous checked-in, because tinderbox has RED.
|
2007-09-03 04:03:49 -07:00 |
|
masayuki@d-toybox.com
|
8541f8f7f0
|
Bug 394109 Inconsistent text-decoration positioning. r+sr+a=roc
|
2007-09-03 04:01:48 -07:00 |
|
roc+@cs.cmu.edu
|
f467e1f991
|
Bug 389421. Rework word selection, in particular so that layout.word_select.stop_at_punctuation is applied to boundaries between punctuation and non-punctuation, and all Unicode punctuation is treated as punctuation.
|
2007-08-29 20:10:19 -07:00 |
|
roc+@cs.cmu.edu
|
03e055e67f
|
Bug 385426. Be careful about looking for a previous soft-hyphen when checking for hyphen breaks. r=smontagu
|
2007-08-29 02:11:39 -07:00 |
|
smontagu@smontagu.org
|
e1a3ea87ca
|
Blow away text runs after Bidi resolution. Bug 382429, r+sr+a=roc
|
2007-08-26 10:20:28 -07:00 |
|
masayuki@d-toybox.com
|
a17a03f2ea
|
Bug 392252 need new APIs for converting AppUnits to gfx usable pixels r=sharparrow1, sr+a=roc
|
2007-08-22 04:13:46 -07:00 |
|
roc+@cs.cmu.edu
|
592a593857
|
Bug 390050. Scan all blocks in a flow-chain, and their overflow lines, when iterating through lines to build text runs. r=smontagu,mats,sr=mats,a=bz
|
2007-08-22 02:08:13 -07:00 |
|
philringnalda@gmail.com
|
48c8d7714f
|
Bug 390423 - Simplify the signature of nsIFrame::GetPointFromOffset, patch by Ben Karel <web+moz@eschew.org>, r=aaronlev/elif, sr=roc, a1.9=dbaron
|
2007-08-16 20:43:00 -07:00 |
|
masayuki@d-toybox.com
|
06c667051a
|
Bug 367614 [Win Only] The composition string for IME should have dashed underline r+sr+a=roc
|
2007-08-16 13:35:18 -07:00 |
|