...and rather live with macro calls not terminated by semicolons until those
macros are gone for good anyway. Completely remove unused
DECLARE_UNO3_XCOMPONENT_DEFAULTS.
Change-Id: I0d4eb5a91736004b0da2b09fce860e57f6d1ddbd
...mostly done with a rewriting Clang plugin, with just some manual tweaking
necessary to fix poor macro usage.
Change-Id: I71fa20213e86be10de332ece0aa273239df7b61a
...which reveals that some uses are not derived from XEventListener, so re-
introduce BaseMutexHelper as an alternative (which originally got superseded by
DisambiguationHelper in 09b546cf9c "Fix
overloaded-virtual warning in canvas").
Change-Id: I1418370733b72526ec481b70fa69f03e7c28b7ce
...which reveals that SpriteCanvasBase is not derived from XBitmapCanvas, so
split up BitmapCanvasBase (just XBitmap) and BitmapCanvasBase2 (plus
XBitmapCanvas) and make IntegerBitmapBase directly abstract over those two as
arguments.
Change-Id: Ieae87fe466bb9f1a861c8a95162ef25235ae5799
With enabled -Woverloaded-virtual gcc warning (see
http://lists.freedesktop.org/archives/libreoffice/2011-March/009567.html),
canvas exposed a nasty clash between WeakComponentImplHelper::disposing
and XEventListener::disposing. Fixed by overriding *once* in baseclass,
and then calling disambiguated, renamed methods.
2008/04/21 07:32:16 thb 1.2.94.4: RESYNC: (1.2-1.3); FILE MERGED
2008/04/02 22:56:27 thb 1.2.94.3: Reworked Surface class to abstract interface; changed all manual refcount handling to RAII
2008/03/18 22:00:55 thb 1.2.94.2: Implementing non-backbuffered canvas for cairocanvas as well - reworked to share most of the code
2007/12/20 22:18:56 thb 1.2.94.1: #i81092# #i78888# #i78925# #i79258# #i79437# #i84784# Large canvas rework, completing various areas such as color spaces, bitmap data access, true sprite and non-sprite implementations, and upstreaming the canvas parts of rodos emf+ rendering