The formatting rectangle is set with EM_SETRECT, and retrieved with
EM_GETRECT, so it corresponds to rcFormat in the code. This defines the
area that the richedit control should draw the text so that it is
offset by the top-left corner of the formatting rectangle, and clipped
so that it doesn't draw past the bottom or right hand side. Thus this
is important for implementing windowless richedit controls to not
interfere with the rest of the window.
This also reverts commit 2b52dd845097f16076c0185b02a003f63898dcab:
wordpad: Empty the richedit undo buffer on creation.
The reverted commit I created to fix an issue that only applied to Wine,
but it just masked the issue which was in richedit controls. The
default character format was set in two places while wordpad was
starting up, and caused wordpad to have two undo items at startup.
The bCaretAtEnd value in ME_TextEditor is used to identify that the
caret is at the end of a wrapped line instead of the start of the next
line in the paragraph since both these positions correspond to the
same position in the document. The bCaretAtEnd value was previously
being set back to FALSE whenever the window was resized.
Optimized reading large texts into RichEdit to be an O(n) order
algorythm instead of O(n^2) by removing extraneous conversions of
character offsets to run offsets.
should greatly improve stability of the wrap code and eliminate
regressions of the most recent versions)
- completely new scrollbar handling (much more reliable) and related
redraw fixes
- Page Down handler (no Page Up yet, fixing wrap/redraw/scrollbar bugs
was of higher priority)
unnecessary repaints)
- added unicode character support to RTF import (like: \u12345 ?)
- small fixes
- fixed whitespace identification bug
- removed drawing of paragraph marks
- improved stub implementations for IRichEditOle
multiple undo actions in one EM_STREAMIN).
- Related changes to undo code (umIgnore mode is now handled
correctly).
- Numerous improvements in the RTF reader: it reads some character
attributes now (you will have proper small print in license agreements
now).
- Fixed a memory overwrite bug in conversion from CHARFORMAT2A to
CHARFORMAT2W.
- ME_JoinRuns didn't mark the paragraph for rewrapping, fixed.
- Removed PostQuitMessage(0) from WM_DESTROY handler (duh!).
- Use of EM_GETOLEINTERFACE is reported with FIXME instead of TRACE
(any app using this message is likely to encounter major problems).
Dylan Smith