|Compiler stress tests? firstname.lastname@example.org (1997-01-02)|
|Re: Compiler stress tests? email@example.com (Duane Sand) (1997-01-03)|
|Re: Compiler stress tests? firstname.lastname@example.org (1997-01-03)|
|Re: Compiler stress tests? email@example.com (1997-01-04)|
|Re: Compiler stress tests? firstname.lastname@example.org (1997-01-12)|
|Re: Compiler stress tests? email@example.com (1997-01-16)|
|Re: Compiler stress tests? firstname.lastname@example.org (Stephen Baynes) (1997-01-17)|
|Re: Compiler stress tests? email@example.com (John Haxby) (1997-01-22)|
|From:||Duane Sand <firstname.lastname@example.org>|
|Date:||3 Jan 1997 23:05:51 -0500|
email@example.com (David Keppel) wrote:
>Somebody asked me about this and I don't know the answer: where do you
>find compiler "stress test" generators/suites/etc.? Commercial
>products are fine.
About 10-15 years ago, there was a nice article in _Software Practise
& Experience_ about stress testing a compiler via a generator.
It used random expansions of grammar rules to generate text that
was correct in a context-free sense, and added corruptions to
that token stream. I don't recall what (if anything) it did about
having identifier uses be (mostly) consistent with declarations.
One nice way to exercise a parser's syntactic error recovery scheme
is to feed it valid source code from a different programming language.
Return to the
Search the comp.compilers archives again.