aboutsummaryrefslogtreecommitdiffhomepage
AgeCommit message (Collapse)AuthorFilesLines
2015-06-23Casts to silence warnings for 64-bit builds.Neil1-2/+2
2015-06-22Remove some annoying messages from cppcheck.Neil1-1/+7
2015-06-20Updating download sizes.Neil1-2/+2
2015-06-20Fixed API name for selecting each.Neil1-1/+1
2015-06-20Added tag rel-3-5-7 for changeset 49ce1d4c2deaNeil1-0/+1
2015-06-19Updates for release 3.5.7.rel-3-5-7Neil6-13/+15
2015-06-19Updates for release 3.5.7.Neil1-2/+5
2015-06-19Fix crash in drag and drop.Neil Hodgson3-3/+16
2015-06-19Updated test program with Xcode recommended settings.Neil Hodgson1-5/+7
2015-06-16Clean up some selection operations. Commonly when changing selection modes,Neil4-30/+36
all of the selection needs to be redrawn so that is implmeneted in Editor::InvalidateWholeSelection. Any extra selections should be discarded with only the main remaining so that is Selection::DropAdditionalRanges. Some default parameters led to less clarity so they no longer have default values. Both Editor::MovePositionTo methods always returned 0 which was ignored so they are now void. Some variables were made const.
2015-06-13Fix bug where maximum two selections possible at once with Ctrl+click dueNeil Hodgson1-6/+6
to change in CancelModes.
2015-06-12Change log.Neil1-0/+4
2015-06-09When combining MarkerHandlerSets, prepend the other set instead of appendingJiří Techet1-2/+3
When undoing many lines with markers (e.g. "changebar" markers) LineMarkers::RemoveLine() is called for many lines and as a result combining markers from the next line for all the removed lines. This may cause the list contains many thousands of elements and traversing it becomes expensive. When lines are removed from the beginning to the end, it's better to prepend the markers from the next line to the current line instead of appending them because the current line "accumulates" all the markers from the following lines and walking the whole list takes more and more time.
2015-06-11Added explanation of SCI_MULTIPLESELECTADDEACH.Neil1-3/+7
2015-06-11Added MultipleSelectAddNext, MultipleSelectAddEach, IsRangeWord, andNeil9-10/+188
TargetWholeDocument.
2015-06-11Explain styling a little more.Neil1-4/+7
2015-06-10Change log.Neil Hodgson1-0/+8
2015-06-05Simplified retrieval of correct display for widget and ensure works on older ↵nyamatongwe2-15/+2
versions of GTK+.
2015-06-04Move FindText flag decoding from caller to function to minimize code and chanceNeil3-20/+6
of mistakes.
2015-06-04Remove cppcheck suppression that is no longer needed.Neil1-3/+0
2015-06-04Use static_cast instead of C-style casts to avoid warnings.Neil1-41/+41
2015-06-04Restored compatibility with GTK+ 2.18 for cursor calls.nyamatongwe2-2/+9
2015-06-03GTK: Don't use gdk_window_set_user_data() in a deprecated mannerColomban Wendling1-0/+4
Since GTK 3.8 gtk_widget_register_window() should be used instead of gdk_window_set_user_data() in a widget's ::realize() method. gtk_widget_unregister_window() is called by the GtkWidget's ::unrealize() handler we chain up to. This has no functional impact.
2015-06-03GTK: Use the proper GTK3 way to set widget window backgroundColomban Wendling1-6/+2
It also drops use of gtk_style_context_get_background_color() that has been deprecated in 3.16. This however has no functional changes.
2015-06-03GTK: Remove unused prototypeColomban Wendling1-1/+0
2015-06-02GTK: Fix runtime completion popup warningColomban Wendling1-1/+2
GtkScrolledWindow in GTK 3.16 might have a minimum and natural height of 0 if GTK_OVERLAY_SCROLLING is enabled (which is the default), so we need to fix our overridden minimal height to never be greater than the natural height. Do that by only changing the minimal height if it is greater than what we want. I don't know why we didn't use to set 0 here, but it might be that a height of 0 caused problems on older GTK versions, so keep 1 to be sure.
2015-06-02Added privacy policyNeil2-0/+71
2015-06-02Provide downloads from scintilla.org to avoid degradation of SourceForge.Neil2-6/+6
2015-05-29As of GTK+ 3.16, gdk_cursor_new is deprecated and has been replaced withNeil2-11/+20
gdk_cursor_new_for_display.
2015-05-29Change log.Neil1-0/+8
2015-05-29Fix warning from Visual C++.Neil1-1/+1
2015-05-21add support for protected regions, prevent styling and folding within ↵Joe Mueller1-53/+72
protected region, allow folding of protected region
2015-05-22GTK: Fill the scrollbars junctionColomban Wendling1-0/+18
Fill the scrollbars junction square with the appropriate style on GTK version having a style for it (GTK >= 3.4). Mostly fixes [bugs:#1611].
2015-05-29Remove wrongly duplicated text.Neil1-6/+0
2015-05-29Change log.Neil1-0/+4
2015-05-29Bug [#1728]. Initialise fields to nil when they may be released in ReleaseViewsNeil Hodgson1-3/+15
without ever being allocated by Create.
2015-05-29Document indexing of selection.Neil1-2/+2
2014-01-20Fix typos in comments (win32)Stefan Weil2-4/+4
2014-01-20Fix typo in commentStefan Weil1-1/+1
2014-01-20Fix typo in commentStefan Weil1-1/+1
2014-01-20Fix typo in string (cocoa)Stefan Weil1-1/+1
2015-05-29Add extra section for next release.Neil1-0/+14
2015-05-26Added tag rel-3-5-6 for changeset e9bfc7a0cb83Neil1-0/+1
2015-05-23Change log.Neil1-0/+5
2015-05-22GTK: Workaround ABI issue with Windows GTK2 bundle and GCC > 3Colomban Wendling1-1/+16
GtkScrolledWindow contains a bitfield, and GCC 3.4 and 4.8 don't agree on the size of the structure (regardless of -mms-bitfields): - GCC 3.4 has sizeof(GtkScrolledWindow)=88 - GCC 4.8 has sizeof(GtkScrolledWindow)=84 As Windows GTK2 bundle is built with GCC 3, it requires types derived from GtkScrolledWindow to be at least 88 bytes, which means we need to add some fake padding to fill in the extra 4 bytes. There is however no other issue with the layout difference as we never access any GtkScrolledWindow fields ourselves. See http://lists.geany.org/pipermail/devel/2015-April/thread.html#9379
2015-05-22Updates for 3.5.6.Neil7-16/+20
2015-05-21Fix bug with TentativeUndo where the undo history was being damaged on OS X.Neil Hodgson1-0/+2
For example the sequence ['a', 'c', left, 'b', undo] would produce "c" instead of "ac".
2015-05-20SCI_TARGETASUTF8 and SCI_ENCODEDFROMUTF8 implemented on Win32.Neil3-2/+65
From johnsonj.
2015-05-20Fix bug with Hanja conversions for DBCS.Neil2-40/+30
From johnsonj.
2015-05-17Change log.Neil1-0/+8