Re: [cdt-dev] LLVM Project Blog: A path forward for an LLVM toolchain on

[Doug Schaefer <dschaefer@xxxxxxx>]

That's one possible solution, although the repo you mention is 2 years without a change. But I have a feeling an external database accessed via Java either through shared mapping of the database file or some really fast IPC would be the way to go.

My biggest worry with using Clang is that we don't end up back where we were with the full indexer with index times measured in hours. That's what gave us such a bad reputation in the early days.

Doug

From: Roland Schulz <roland@xxxxxxx>
Reply-To: "CDT General developers list." <cdt-dev@xxxxxxxxxxx>
Date: Wednesday, 11 September, 2013 3:55 PM
To: "CDT General developers list." <cdt-dev@xxxxxxxxxxx>
Subject: Re: [cdt-dev] LLVM Project Blog: A path forward for an LLVM toolchain on Windows

On Wed, Sep 11, 2013 at 12:49 PM, Nathan Ridge <zeratul976@xxxxxxxxxxx> wrote:

> There is no doubt LLVM is becoming a huge force. Though I'm more interested in it from the parser/indexer side. I wonder how we'd plug it in..

Clang, on the other hand, comes very close to implementing the standard to a tee, and it would be really nice to be able to reuse the excellent work that clang developers have been doing (and will continue to do) in CDT. I'm not sure whether/how that's possible.

Maybe it would be possible to run clang or a clang based tool as a separate process from CDT. And read the externally generated database from CDT/java. This could be an optional 2nd parser which would only be used if clang is available. Such a clang based indexer generating a database is available at: https://github.com/exclipy/clang_indexer

Roland

 

 

Regards,
Nate

--
ORNL/UT Center for Molecular Biophysics cmb.ornl.gov
865-241-1537, ORNL PO BOX 2008 MS6309