It’s been 1 years and a half that scintilla isn’t updated. There are some much interested feature or bug fix in recent releases
here is a completely experimental nightly build with Scintilla 3.7.0 embedded.
Scintilla has been updated, notepad++ modifications has been applied, but we don’t use the new API scintilla.
Anyone interested to test ? It’s on the build system website ( appVeyor). links :
informations for developers :
I used scintilla github mirror (https://github.com/mirror/scintilla)
I forked it (https://github.com/cmeriaux/scintilla).
I created branches to apply notepad++ modifications on each release (3-5-6 3-5-7 3-6-0 … 3-7-0 ) (patch applied from original repository, then patch has been cherry-picked over branches)
I declared scintilla as a git submodule ( https://github.com/cmeriaux/notepad-plus-plus/tree/scintilla )
I created branches with new releases of scintilla :
I’m giving a +1 for effort here. I don’t know really what advantages some of these newer versions of Scintilla provide, but surely it is not good for the version in Notepad++ to fall too far behind? :)
By the way, do you use unmodified sources of Scintilla (as is), or do you integrate some patches from Notepad++'s codebase?
Ah, you mentioned you applied the Notepad++'s patches.
So, I see no reason to not use this version of the SciLexer.dll! Instead, I see reasons to use it!
For example, it works faster when doing multi-editing in a big (80 MB) Unicode file.
Though - no. I’ve found a serious reason to not use this custom SciLexer.dll in its current form.
When searching with Regular Expressions case-insensitively, it does not find non-Latin characters (such as Cyrillic) in another (upper/lower) case at all! It does not work neither for “Find All in Current Document” nor for “Find in Files”. (Without Regular Expressions it’s OK, case-insensitive searching works for Cyrillic characters.)
The Notepad++'s SciLexer.dll handles case-insensitive with Regular Expressions correctly.
I think your “Regular Expressions” issue is due to the nightly build compilation process that doesn’t use “boost regular expression”. DonHO didn’t want that because it’s time consuming and sometimes fails the build.