Hi Rob: I intentionally made 'make test' do *only* the tests, since I could not figure out why the 'pure_all' target was causing a rebuild during 'make test' right after that target was built using 'make'.
The right fix would be to get 'pure_all' to not rebuild a second time during 'make test' when it just built during 'make'. That said, the proper recipe should always be: perl Makefile.PL make make test make install So my fix allows that to work. You just need the 'make' step. Regards, Doug [EMAIL PROTECTED] Software Engineer III UCAR - COSMIC, Tel. (303) 497-2611 On Thu, 23 Oct 2008, Sisyphus wrote: > > ----- Original Message ----- > From: "Chris Marshall" <[EMAIL PROTECTED]> > To: "perldl list" <[email protected]> > Sent: Thursday, October 23, 2008 12:12 PM > Subject: [Perldl] please test PDL-cvs for round 02 > > >> All- >> >> Fixes for the dumper.t and inlinepdlpp.t >> problems with PDL-2.4.3_01 have >> been pushed into the cvs tree (along with >> some other tweaks). >> >> Please test build on your system and let >> me know if it works for you. It builds >> fine on cygwin but I would like a check >> from another OS (or two) before >> uploading PDL-2.4.3_02 to PAUSE/CPAN. >> >> I'll go ahead and upload in about 9hrs >> or when I get 2 more OS check reports, >> whichever comes first. >> > > There's something wrong with Doug's MY::test rewrite in the latest CVS (for > me, anyway). > > We should be able to run: > perl Makefile.PL > make test > > But if I try running 'dmake test' or 'nmake test' immediately after running > 'perl Makefile.PL', it just goes straight into running the tests - the > nmake/dmake stage (that creates and populates the blib) is left out !! And > then, if I run something like 'perl t/inlinepdlpp.t', I get an error that > blib could not be found (even though there's no reason that the command I > ran needs to find blib). > > Does anybody else get the same ? > > I won't have a chance to take a decent look at this until Monday. > Chris, if you want to get a 02 out before that, then go for it - I don't > think the problem I'm seeing will have any bearing on CPAN-testers results. > We can sort out the details later. But it looks like I'll be rewriting > Doug's change as: > > sub MY::test { > if($^O =~ /MSWin32/i) { > package MY; # so that "SUPER" works right > my $inherited = shift->SUPER::test(@_); > my @lines = split "\n", $inherited; > my @new = (); > my $pass = 1; > foreach (@lines) { > if ($pass) { > push @new, $_; > $pass = 0 if /^test\b/; > } else { > if (/^\s*$/) { > push @new, ''; > $pass = 1; > } > } > } > return join("\n", @new,"\n"); > } > else { > # This subroutine re-writes the 'test' section > # to get rid of 'pure_all' as a dependency for > # certain test targets. This causes 'make test' to > # re-run the entire 'make' target instead of just testing > # PDL. > # > # It also gets rid of the 'subdirs-test' section, which > # attempts to run tests in individual sub-directories. > # This is also causing a significant rebuild after the main > # test is done. > # D. Hunt 10/20/2008 > # But Win32 really needs the original. > # Sisyphus 23/10/08 > package MY; # so that "SUPER" works right > my $inherited = shift->SUPER::test(@_); > $inherited =~ s/pure_all//g; > $inherited =~ s/subdirs\-test$//mg; > return $inherited; > } > } > > Either that, or settle for PDL being the only module on CPAN that requires > we run 'dmake'/'nmake' before 'dmake test'/'nmake test'. > > Cheers, > Rob > > > _______________________________________________ > Perldl mailing list > [email protected] > http://mailman.jach.hawaii.edu/mailman/listinfo/perldl > _______________________________________________ Perldl mailing list [email protected] http://mailman.jach.hawaii.edu/mailman/listinfo/perldl
