Commit Graph

20 Commits

Author SHA1 Message Date
Boris Zbarsky
de62e00f8f Bug 491415. Need to initialize this member so we don't report bogus ascents. r+sr=roc 2009-05-05 01:01:47 -04:00
Benjamin Smedberg
24e7cb1f78 Bug 396185 - Make nsIFrame derivatives and helper abstract classes use a different dynamic-cast system than nsISupports:
* we know all types frames may be cast to at compile time, so instead of extensible GUID IIDs, use a big enum (see nsQueryFrame::FrameIID)
* eliminate all vestiges of refcounting, since frames aren't refcounted

Some frames (SVG frames in particular) still implement nsISupports-derived interfaces, for example nsISVGValue. There is a FrameIID for nsISVGValue that lets you go from a frame to the XPCOM interface, but you can't query back.

r+sr=roc

nsITextControlFrame didn't have an IID the first time around, but this wasn't a compile error because nsITextControlFrame::kFrameIID inherited from nsIFormControlFrame::kFrameIID. I've added a static analysis pass to verify the correct behavior, since I can't figure out a way to make the compiler do it.

--HG--
extra : rebase_source : 4894a2ca0278e2ab92f27459db77165f8348cf41
2009-01-12 14:20:59 -05:00
Benjamin Smedberg
50621e4f9f Backed out changeset 4c4df6ed1b41 - Bug 396185 - Make nsIFrame not inherit from nsISupports due to mochitest failures... these appear to be crashes in nsGenericHTMLElement::GetEditorInternal. 2009-01-09 11:35:24 -05:00
Benjamin Smedberg
d0b948876f Bug 396185 - Make nsIFrame derivatives and helper abstract classes use a different dynamic-cast system than nsISupports:
* we know all types frames may be cast to at compile time, so instead of extensible GUID IIDs, use a big enum (see nsQueryFrame::FrameIID)
* eliminate all vestiges of refcounting, since frames aren't refcounted

Some frames (SVG frames in particular) still implement nsISupports-derived interfaces, for example nsISVGValue. There is a FrameIID for nsISVGValue that lets you go from a frame to the XPCOM interface, but you can't query back.

r+sr=roc

This patch locally causes two REFTEST-UNEXPECTED-PASS for Bidi stuff. It's possible that I accidentally fixed a bug, but I'm not sure, so I'm going to wait for the tinderboxes to confirm my local results.
2008-11-05 14:25:30 -05:00
L. David Baron
eee42cfdd7 Replace space manager with a more limited float manager. (Bug 191448) r+sr=roc
--HG--
rename : layout/generic/nsSpaceManager.cpp => layout/generic/nsFloatManager.cpp
rename : layout/generic/nsSpaceManager.h => layout/generic/nsFloatManager.h
2009-01-04 19:39:54 -05:00
Benjamin Smedberg
9696613a50 Relanding bug 461410 - nsILineIterator is never used outside of layout - set nsresults to replicate the existing code more correctly, which fixes the mochitest failures, r+sr=roc with nits noted by Mats Palmgren 2008-10-30 12:17:59 -07:00
Benjamin Smedberg
4214b72dd5 Backed out changeset d4c9a0776667 (deCOM nsILineEnumerator) due to regression from it or bug 461212 2008-10-28 02:49:14 -04:00
Benjamin Smedberg
dd4c950cff Bug 461410 - nsILineIterator is never used outside of layout, and doesn't need to be refcounted: the callers can just destroy it when they're done with it. We can't do this with a virtual destructor, however, because nsTableRowGroupFrame implements the interface directly, while nsLineIterator is a separately-allocated class. So clients are expected to call DisposeLineIterator when they're done with it.
Instead of accessing nsILineIterator using QueryInterface, we add a nsIFrame::GetLineIterator API, which is cleaner and more efficient all at the same time!

r+sr=roc
2008-10-28 00:47:19 -04:00
roc+@cs.cmu.edu
2032b542bf Bug 411870. nsBidiPresUtils needs to track the current line across block continuation boundaries. r=smontagu 2008-01-29 15:39:39 -08:00
roc+@cs.cmu.edu
dea0868d56 Bug 411870 backout again 2008-01-28 11:10:26 -08:00
roc+@cs.cmu.edu
6ac78d0700 Bug 411870. Bidi resolution needs to scan through block continuations when tracking the current line for frames. r=smontagu 2008-01-28 10:16:22 -08:00
roc+@cs.cmu.edu
d0b43fea97 backing out 411870 2008-01-27 22:28:31 -08:00
roc+@cs.cmu.edu
1233a70df6 Bug 411870. Make bidi resolution track lines across block continuation boundaries. r=smontagu 2008-01-27 21:07:14 -08:00
mats.palmgren@bredband.net
151fc8f9ab [DEBUG-only] Print a warning if we find a null float in the float cache. b=387651 r+sr=roc 2007-07-22 06:37:36 -07:00
mrbkap@gmail.com
4287a9b897 Fix printf format-specifier and args mismatch to prevent potential DEBUG-only crashes. bug 388358, r+sr=roc 2007-07-16 15:26:53 -07:00
jwalden@mit.edu
e3c4baccae Bug 348748 - Replace all instances of NS_STATIC_CAST and friends with C++ casts (and simultaneously bitrot nearly every patch in existence). r=bsmedberg on the script that did this. Tune in next time for Macro Wars: Episode II: Attack on the LL_* Macros. 2007-07-08 00:08:04 -07:00
roc+@cs.cmu.edu
26f439cf9b Bug 383551. Clean up IsInlineFrame, nsStyleDisplay::IsBlockLevel and related methods. r+sr=dbaron 2007-06-26 19:31:35 -07:00
sharparrow1@yahoo.com
78559aa94e Bug 381385: get rid of unneeded members of nsFloatCache (saves about 50 bytes footprint per float). r+sr=roc. 2007-06-16 13:27:46 -07:00
bzbarsky@mit.edu
2fd17c94d8 Rename nsIFrame::GetPresContext to nsIFrame::PresContext. Bug 376042, patch byTaras Glek <tglek@mozilla.com>, rs=roc. 2007-03-30 14:11:41 -07:00
hg@mozilla.com
465265d0d4 Free the (distributed) Lizard! Automatic merge from CVS: Module mozilla: tag HG_REPO_INITIAL_IMPORT at 22 Mar 2007 10:30 PDT, 2007-03-22 10:30:00 -07:00