[RkWard-devel] R Syntax Highlighting

[Arne Henningsen]

Hi Thomas, *!

On Tuesday 31 August 2004 14:24, Thomas Friedrichsmeier wrote:
> Hi,
>
> > I installed rkward and I think it looks very promising! I noticed that
> > rkward uses an extended version of Egon Willighagen's syntax highlighting
> > definitions. Since Egon does not maintain this file any more, I took over
> > his job and also extended the R syntax highlighting definitions. They can
> > be downloaded from my homepage (scroll to the very bottom):
> > http://www.uni-kiel.de/agrarpol/ahenningsen/index-e.html
> > Does rkward require special syntax highlighting definitions or can we
> > merge the two branches and just maintain only one file?
>
> When I added kate/syntax-highlighting, I found the syntax-highlighting
> shipped with KDE to be rather limited, and I wasn't aware of your version,
> so that's why I extended it myself. In fact, now you can download a new
> kate-hl definition from http://kate.kde.org/hl/, which contains some of
> those extensions/changes. I also submitted the hl-definition RKWard is
> using currently to kwrite-devel at kde.org, but apparently that slipped past
> attention and I did not bother to nag.

I did the same with my hl-definitions - with the same result!

> Of course it would be preferrable to have a single merged hl-definition.
> There's no real reason to maintain separate versions. For now, I'll
> continue to ship the hl-definition, since I feel it's better than the
> version KDE/kate comes with, but that's just a temporary solution.
> As to the hl-definition itself: It would be cool, if you could maintain the
> hl-definition. I'm happy about every bit of work I don't have to do myself.

OK, I'll try to merge these two files and send you the result. Then I'll be 
happy if you all could suggest improvements. After this, we can try again to 
submit the file to the kwrite-devel list.

> Technically, you'll note that the most visible difference between our
> definitions is, that I did not bother to add a special handling for named
> arguments. I felt that since it's not easily possible to identify
> symbol-names outside of "foo (...)", it was not worth the effort to do so
> inside function calls, either. Also, it would be rather difficult to
> correctly handle (unusual) constructs like:
>
> list (x=(a=c (1, 2, 3)), y=c(2, 3, 4))
>
> (i.e. assignments using "=" inside a function-call, which are not named
> arguments). Other than that, there are some differences in detail. It
> should be easily possible to merge those.
> Would you like to work on merging the two versions?

Yep, but it might take a while, because I am very busy at the moment (e.g. 
preparing a new R package). 

Arne

> Thomas
>
>
> -------------------------------------------------------
> This SF.Net email is sponsored by BEA Weblogic Workshop
> FREE Java Enterprise J2EE developer tools!
> Get your free copy of BEA WebLogic Workshop 8.1 today.
> http://ads.osdn.com/?ad_id=5047&alloc_id=10808&op=click
> _______________________________________________
> RKWard-devel mailing list
> RKWard-devel at lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/rkward-devel