Re: Semantic error recovery

Steve Boswell (whatis@gni.ai.mit.edu) wrote:
: Are there compilers that recover from semantic errors without a cascade of
: meaningless error messages? How do they do it?


mauney@adm.csc.ncsu.edu (Jon Mauney) writes:
|> The fact is that most commercial compilers have appallingly bad error
|> handling, despite the fact that reasonable techniques for handling both
|> syntax and semantic errors have been available for what seems like eons.
|> ...
|> Apparently, compiler customers do not spend enough time ridiculing vendors
|> for their egregious front-ends.


While this last certainly is true to a certain extent, it must be conceded
that implementation of something like an FMQ approach can have a
significant effect on the performance (translation: speed) of a front end.
I think that most of us who have written or supported commericial
compilers can testify that compiler customers are in fact quite adept at
ridiculing what they regard as poor performance -- and that frequently
they prefer a faster compiler to one that provides better error
recovery/repair. If you want to sell compilers, the trick is to strike
the proper balance, and I would suggest that the "appallingly bad error
handling" of "most commercial compilers" is partly a matter of perspecitve
and largely a response to market demand rather than a consequence of
ignorance or lack of effort.
--
Gary H. Merrill [Principal Systems Developer, Compiler and Tools Division]
SAS Institute Inc. / SAS Campus Dr. / Cary, NC 27513 / (919) 677-8000
sasghm@theseus.unx.sas.com ... !mcnc!sas!sasghm
[I have to say that I'd be happy with a compiler that died after one or
two error messages so long as it ran really fast, like under a second per
compilation. -John]
--

Return to the comp.compilers page.
Search the comp.compilers archives again.