[Development] look-behind assertions in syntax HL?

[Thiago Macieira]

Em seg 17 fev 2014, às 11:06:52, Matthew Woehlke escreveu:
> On 2014-02-16 12:02, Thiago Macieira wrote:
> > Em dom 16 fev 2014, às 15:09:49, Giuseppe D'Angelo escreveu:
> >> I guess that for Kate's purposes a small wrapper class around QRegExp
> >> + QRegularExpression would suffice for supporting both syntaxes. For
> >> the future, we could instead think of adding wildcard and fixed-string
> >> pattern types to QRegularExpression.
> > 
> > Fixed strings are supported already: QRegularExpression::escape.
> 
> Does this "know" already to do a simply string compare to test the
> resulting regex? (And/or does compiling of the regex's already handle
> that case for any regex that is effectively a fixed string?)
> 
> For me that has been the main reason to use a fixed-string pattern type.

We pass it to the PCRE engine. How it compiles and manages the pattern is its 
own business.

> > For wildcards, we should simply add another static that converts ? to .
> > and * to .*. Do we want to support more interesting globbing things like
> > []?
> Are these supported mainly because the pattern may be coming from user
> input, or because internally the match implementation can be more
> efficient? If the latter, I'm not sure it is worth even having a "glob"
> pattern type. If the former, then supporting as many shell-isms as
> possible is probably useful.

I don't think we want to have a globbing class in Qt and I don't want to write 
it in QString either.

-- 
Thiago Macieira - thiago.macieira (AT) intel.com
  Software Architect - Intel Open Source Technology Center