a couple of possible routes:

• use pygments + qsyntaxhighlighter -- pygments to lexically analyze the text, and set the colors on the various things ourselves.

• use our own syntax highlighting in qsyntaxhighligher

• use pygments to generate rich text formatted output in our qsyntaxhighlighter subclass

• using gnu source highlight via qt: http://srchiliteqt.sourceforge.net/

Useful links:
http://doc.trolltech.com/4.5/qsyntaxhighlighter.html
http://doc.trolltech.com/4.5/richtext-syntaxhighlighter.html
http://lateral.netmanagers.com.ar/static/highlighter.py

The first option sounds very nice and should be the most practical.

Some things that make things tricky --
remember that we're looking at diff hunks. That complicates things like:

 /* this is a comment
 + and this is a new line
   and this is an existing line
@@ -1,20 diff header +2,25 @@
 int main()
 {
 +     int i;
        char c;

As you can see, we only ever see the beginning of the comment inside of
the diff hunk. Syntax highlighting multiple lines means we have to remember
some state and some context, but the "we're looking at diffs here" complicates that since pygments really wants to work on the whole file at once.

I remember you mentioning that they support syntax highlighting diffs, but I have a feeling that it doesn't highlight the code inside the hunks. If pygments is smart enough to handle that then I'd be quite impressed.

we might have to do some preprocessing to only highlight the diff hunks and ignore the diff headers, but I'd say we shouldn't do more than just naively apply the syntax highlighting to the diff hunks, and let pygments do its thing (maybe we can tell it the language if we know that much). If it's smart enough to figure out a decent hilighting strategy, then more power to it.