On Wed, May 14, 2025, at 10:30 AM, Aaron Merey wrote:
> If it's not quite generic enough for test-subr.sh we can always
> include this in a stacktrace-subr.sh file, for instance.
Yeah, I'm thinking this is the more likely outcome.
>> +#
>> +# TODO(REVIEW): Better shell-isms for comparing file and regex?
>> +# \(\s\e\d\)\+\i\s\a\d\d\i\c\t\e\d\\t\o\b\a\ck\s\l\a\s\h\e\s
>
> I think the use of sed here is fine. If you wanted to improve
> readability you could perform the scrubbing across multiple sed
> commands each with a comment explaining the step.
Anything's good enough with enough comments!
There's also the possibility of awk being powerful enough to make
me want to use it. Do you know if awk is considered sufficiently
standard for the testsuite to depend on it?
> FYI here is my results diff from running this test:
> -snip-
Thanks, that's another item for the dataset I got from the trybot runs.
Unfortunately, I won't know for sure until I start to test things like bsd....
is there a list somewhere of unices we're supposed to support?
--
All the best,
Serhei
http://serhei.io
