It isn't so easy to implement a lexer for VB(A/SCRIPT) and/or PHP for several reasons (search the C::B forum for "PHP lexer"). I have attached my last attempt for both. Both are not working 100% but maybe a good starting point. You should read the C::B WiKi on how to write a lexer and the wxScintilla documentation if you want to improve this work. But maybe that's already enough for you... ;-)
This is similar to what I was tinkering with, and had similar results. It seems that what *should* work as a subset of the HTML doctype (index 4) has the underlining, no code folding, and no highlighting problem, while other doctypes get partial highlighting, presumably because most everything gets interpreted as a tag parameter rather than full code, but then bomb out on the HTML. (I've had similar near-successes with document type index values of 8 and 15 as well, and I assume its because they all use the same engine for parsing VB-derived code.)
I just downloaded and installed the SciTE editor as well, and noticed that support for classic ASP seems to have a much cleaner implementation. Perhaps I'm just accentuating the fact that I'm a complete neophyte in the realm of these projects, but my understanding is that scintilla is the same editor package used in the CB code. If that's true, couldn't this all be patched together somehow without much effort?
Regardless, thanks for your post. Your attempts will definitely come in useful. I'll just have to continue to hack away at this stuff when real work doesn't get in the way.