[
Date Prev][
Date Next][
Thread Prev][
Thread Next][
Date Index][
Thread Index]
[
List Home]
|
RE: [cdt-dev] CDT DOM question
|
Well, in our naivety, we assumed that we could properly parse C++. It took a
lot of work from two very smart developers (not me) but we think we're well
on our way. The biggest challenges were ambiguities in C++ that require some
amount of semantic analysis to figure out the syntactic structure. Also C++
templates are quite the beast to produce efficient structure in the AST and
efficient semantic analysis algorithms.
Completeness was our objective. But then, the DOM parser is pretty new for
CDT 3.0 and hasn't been exercised as much as a commercial compiler would
have been. We'd love to hear feedback (and bugs) on it.
Doug Schaefer
Senior Software Developer
QNX Software Systems
> -----Original Message-----
> From: cdt-dev-bounces@xxxxxxxxxxx [mailto:cdt-dev-bounces@xxxxxxxxxxx] On
> Behalf Of Craig Rasmussen
> Sent: Friday, September 02, 2005 10:18 AM
> To: CDT General developers list.
> Subject: [cdt-dev] CDT DOM question
>
> I have a question from colleague regarding the completeness
> of the CDT DOM C++ parser. How completely does it parse the
> C++ standard and what is lost in the AST?
>
> The question comes from the common wisdom that EDG based
> front ends are the only compilers that can "really" parse C++.
>
> The questions comes up in the context of performance-related
> refactorings that might want fuse (or unroll) loops and need a
> lot of program context to be able to do this correctly.
>
> Ciao,
> Craig
>
> _______________________________________________
> cdt-dev mailing list
> cdt-dev@xxxxxxxxxxx
> https://dev.eclipse.org/mailman/listinfo/cdt-dev
