Yeah, it should be straightforward to at least substantially improve things. My suggestions would be either: (a) refine existing approach, but by getting back the actual DiagnosticMappingInfo object, and still reverse engineering the effect of flags.
(b) change getDiagnosticLevel to return a 'why' code that would go into the actual diagnostic object. This avoids the gross reverse engineering. - Daniel On Wed, Sep 28, 2011 at 10:44 PM, Chandler Carruth <[email protected]> wrote: > On Wed, Sep 28, 2011 at 6:01 PM, Daniel Dunbar <[email protected]> wrote: >> >> The TextDiagnosticPrinter code is still fragile as it is just "reverse >> engineering" what the diagnostic engine is doing. Not my current priority to >> fix though. > > Is this straightforward to fix? I'm hacking a lot on the text diagnostic > printer, working to make a generic interface for textually formatting > diagnostics that can be shared amongst various clients, so I'd be happy to > take a look while there. _______________________________________________ cfe-commits mailing list [email protected] http://lists.cs.uiuc.edu/mailman/listinfo/cfe-commits
