Age | Commit message (Collapse) | Author | Files | Lines |
|
|
|
|
|
|
|
SCI_GETPROPERTYINT, and SCI_SETPROPERTY as this is no longer supported.
|
|
|
|
|
|
|
|
on each subline.
|
|
|
|
|
|
|
|
Turn off CLANG_WARN_QUOTED_INCLUDE_IN_FRAMEWORK_HEADER
as headers also used on other platforms that do not build as a framework.
|
|
|
|
|
|
|
|
|
|
|
|
updates.
|
|
|
|
on Win32. This also avoids a header rectangle above the list.
|
|
Drop lexlib from include path as no longer part of Scintilla.
For clang on Windows, turn off deprecated function warning and disable
sanitizers as they are slow.
|
|
|
|
|
|
https://github.com/mneuroth/SciTEQt
|
|
which were considered word start / end positions even when first or last
character did not fit requirements.
|
|
This change does not affect behaviour.
|
|
to hold some number of lines. This is an optimization that can decrease
reallocation overhead.
|
|
prevents extra reallocations.
This change made 0-length gaps more likely and that caused attempts to std::move
by 0 which is undefined and failed with the g++ standard library. Fixed by not
moving for 0-length gap.
|
|
|
|
|
|
widely.
Ensure that case where all text is after the gap is as efficient as when all
text is before the gap.
|
|
pairs in backwards NextPosition.
|
|
before treating as a character.
|
|
|
|
Averaged 2x faster for forward search "needle" in Editor.cxx as UTF-8 or Latin1.
|
|
Increases performance of search by around 25%.
|
|
|
|
|
|
|
|
text.
|
|
char passed.
|
|
directions.
|
|
extra bytes.
|
|
WM_GETTEXT to not use CountUTF16 or GetRelativePositionUTF16, instead using
the Win32 API MultiByteToWideChar for consistency between reported length and
text as UTF-16.
|
|
removed.
|
|
invalid single bytes in multi-byte encodings.
|
|
|
|
Drop temporary test for IsDBCSTrailByteNoExcept.
|
|
instead of negated IsDBCSTrailByteInvalid.
|
|
ReplaceREInFile.
|