On 2007/12/05 09:37, J.C. Roberts wrote:
>
> Unfortunately, using `make show-required-by` on the right-most doesn't
> really show all of the dependencies because it skips REGRESS_DEPENDS.
>
> In case of a regression on a regression, I take a more aggressive
> approach to finding all the dependencies, including REGRESS_DEPENDS.
>
> $ cd /usr/ports
> $ find . -type f -name 'Makefile' -exec sh -c 'if [ X`grep -l \
> > p5-whatever "{}"` == X"{}" ]; then echo {}; fi' \;
ouch :-) If you need to ask this sort of question of the ports tree,
take a look in the direction of sqlports. I get the impression this is
seldom used, but it's a really nice tool. Takes a while to build the
db (and a good way to notice Cs in your ports tree :) but you can
just grab the tgz from a package snapshot.
$ time echo ".headers on
select * from depends where dependspath = 'devel/p5-Test-Warn';" \
| sqlite3 /usr/local/share/sqlports
FULLPKGPATH|FULLDEPENDS|DEPENDSPATH|TYPE
net/p5-Net-Radius|::devel/p5-Test-Warn|devel/p5-Test-Warn|Regress
www/p5-WWW-Mechanize|::devel/p5-Test-Warn|devel/p5-Test-Warn|Regress
0m0.11s real 0m0.03s user 0m0.07s system
I find it's helpful to use a query browser GUI while working out
how to ask questions; sqliteman works well for me (then, half the time,
I end up going back to the CLI and feeding the output into a shell
script.. not exactly "the SQL way" but it gets the job done).
> The above gives me everything touched by a port, including the 'stated'
> regression test dependencies. Unfortunately, it doesn't give me the
> unstated (in the port Makefile) dependencies.
I think the only way to pick those up is to walk through each port
doing "pkg_delete /var/db/pkg/*; make regress", I'll try and do some
of this when I have reduced my to-do list a bit!
> It would be nice if I could automate the tree walking for regression
> failures, but some stuff in the tree never passes regression tests, for
> example, p5-Net-SSLeay just hangs.
I just had a look at this, it forks and execs sslecho.pl, but this
never exits, so the test doesn't complete. This is broken, it should
at least be REGRESS_IS_INTERACTIVE if not NO_REGRESS until someone
works out how to fix it (I tried adding another kill but that didn't
do the trick, haven't thought of anything else to try yet).
