Are there other Forth testing tools out there?

I've tried using the Hayes tester a couple of times. I like the
syntax
and simplicity, but its failure messages are just useless. If I have
to
go back and insert debug statements and re-run the tests to figure out
what actually went wrong, why bother using it at all? I might as well
just write ad-hoc test code and shove it in a file.

I want something that (at least) displays the actual and expected
stack
contents, and preferably something that can be extended to deal with
strings or other data. I'd love to be able to do:

T{ 1 2 3 => 1 5 }T
T{ s" abcdef" 2 /string => s" def" }( string )T

And see:

T{ 1 2 3 => 1 5 }T
Returned: <3> 1 2 3
Expected: <2> 1 5

Josh Grams qualdan.com> writes:

On 2008-02-24 14:10:47 -0700, anton@mips.complang.tuwien.ac.at (Anton
Ertl) said:

In article <154c0481-de36-4f57-a9c2-80f2d4dec4b1@c33g2000hsd.googlegroups.com>,
Josh Grams qualdan.com> wrote:

The merit of tsuite.frt is that it is comprehensive.

One of the last errors in the Renesas Forth was in <.
Gosh, forget to push the result on the stack. What would you want
more than the information that < doesn't work?

"Albert van der Horst" wrote in message news:jwrwq5.g14@spenarnc.xs4all.nl...

On Feb 24, 9:39 pm, Albert van der Horst
wrote:

Well, if I'm the one *writing* the Forth (which is the situation I'm
talking about), I want to know exactly HOW it didn't work. Did it
forget to push the result? Did it push a wrong result? Did it push
an
extra result? Even if I'm not the author, I still want good failure
messages so I can file a decent bug report.

On Feb 25, 10:40 am, Josh Grams qualdan.com> wrote:

Go Josh!

Best, Charles

I find the Hayes tester useful when developing an application, the
tests are written in parallel with the code and at the end you have
regression tests available for when the code is modified.

Gerry

Ed wrote:

Are you considering only testing of Forth systems? What about using the test
suite to test applications? For instance, the extended version of the Hayes
tester, incorporating extensions by David Williams and Anton Ertl, have proved
valuable in finding problems with the Forth Scientific Library modules. These
problems went unnoticed for many years. The findings are documented in an
earlier thread on automated FSL tests.

Also, David Williams used the Hayes test suite to perform some stringent
regression tests on his ppc port of kForth. These tests uncovered several
problems with the x86 version, which were subsequently fixed.

Krishna

OK, I banged out a first draft...it basically works, and I've
tested the cases that I can think of. But it is a first draft,
and I've been staring at this code for too long at a stretch, so
I'd be surprised if I've caught everything.

http://qualdan.com/forth/tester-2008-02-26.tar.bz2

Any suggestions for how to improve this are welcome (I think).
Though...I did just ask for (and received) code and then went and
reinvented the wheel anyway, so... :)

I do have one question. It's easy to test without floats:

: environment? ( c-addr u -- false | i*x true ) 2drop false ;

But how do I test the combined float/data stack case? Anybody
have a Forth that does things that way? (x86 or PPC Linux,
preferably).

--Josh