[prev in list] [next in list] [prev in thread] [next in thread] 

List:       gcc-patches
Subject:    Re: Specialization in templates
From:       Mark Mitchell <mark () markmitchell ! com>
Date:       1998-10-31 17:30:12
[Download RAW message or body]

>>>>> "Alexandre" == Alexandre Oliva <oliva@dcc.unicamp.br> writes:

    Alexandre> Ok, I managed to introduce a modification in the
    Alexandre> grammar that catches this case, and prints an error
    Alexandre> message.  However, after that, the template processing
    Alexandre> prints another error message that I can't find a way to
    Alexandre> avoid :-(

I guess I would suggest we kind of stay out of the grammar at the
moment.  I really do have a partly done new recursive-descent parser,
which can do better error recover; it's waiting for the time and money
required to finish it.  I would expect that sometime within the next
few months, those resources will become available.

The bison grammar is quite fragile; sometimes even reasonable changes
cause things to break in odd ways.  I'm not commenting on your patch,
just on the general case. 

So, I have nothing to say about your patch; just a general
encouragement not to try to push the square LALR(1) into the round
hole of great parser for C++. :-)

-- 
Mark Mitchell 			mark@markmitchell.com
Mark Mitchell Consulting	http://www.markmitchell.com

[prev in list] [next in list] [prev in thread] [next in thread]