Bug #71928 for DBIx-Class: t/53lean_startup.t: Can't locate Tie/ExtraHash.pm

I've found the root cause of this, and I'm very curious why noone else is seeing this error yet. Obviously, nothing in DBIx-Class refers to Tie::ExtraHash. That dependency actually comes from namespace::clean, which is used by DBIx::Class::Carp. In namespace::clean, the "right thing" is done to use Tie::ExtraHash as a base class: use Tie::Hash; use base 'Tie::ExtraHash'; This works because base->import does the following: eval "require $base"; die $@ if $@ && $@ !~ /^Can't locate .*? at \(eval/; However, in t/53lean_startup.t, because the require method is wrapped in the first BEGIN block, the error that's thrown is NOT "at (eval)" it is "at t/53lean_startup.t", the regexp fails to match, and base->import re-throws the error. I've reduced the problem to a very simple test case, which fails for every perl I've tried, on every platform: BEGIN { *CORE::GLOBAL::require = sub { CORE::require($_[0]); }; } use Tie::Hash; use base 'Tie::ExtraHash'; Run this, and you get: pmoore@Renegade$ perl ~/base.pl Can't locate Tie/ExtraHash.pm in @INC (@INC contains: /Users/pmoore/lib/perl5 /opt/local/lib/perl5/site_perl/5.12.3/darwin-multi-2level /opt/lo\ cal/lib/perl5/site_perl/5.12.3 /opt/local/lib/perl5/vendor_perl/5.12.3/darwin-multi-2level /opt/local/lib/perl5/vendor_perl/5.12.3 /opt/local/l\ ib/perl5/5.12.3/darwin-multi-2level /opt/local/lib/perl5/5.12.3 /opt/local/lib/perl5/site_perl /opt/local/lib/perl5/vendor_perl .) at /Users/pm\ oore/base.pl line 5. ...propagated at /opt/local/lib/perl5/5.12.3/base.pm line 94. BEGIN failed--compilation aborted at /Users/pmoore/base.pl line 10. That appears to be the root cause of this problem. The wrapped require method results in confusing base.pm, because the error doesn't appear to come from the eval of that string, ot appears to come from some other file, and the code dies. If you download and install: http://backpan.perl.org/authors/id/R/RI/RIBASUSHI/namespace-clean-0.20_01.tar.gz then that test passes fine. This is because in 0.20_01, Tie::ExtraHash is refered to via: ./lib/namespace/clean.pm: tie( %^H, 'Tie::ExtraHash', namespace::clean::_ScopeGuard- Show quoted text
But in 0.21, this was changed to: ./lib/namespace/clean.pm: use base 'Tie::ExtraHash'; I have tried hacking in several workarounds, but in every case for which the hack "fixes" my test program above, it breaks something else in t/53lean_startup.t. For example, I can make the test script work if I do this: BEGIN { *CORE::GLOBAL::require = sub { eval { CORE::require($_[0]); }; if ( $@ ) { $@ =~ s{at\s+\S+}{at (eval 1)}gmsx; die $@; } }; } IOW, hack $@ so that it looks like it came from eval. However, if I add this to t/53lean_projects.t, I get an even stranger set of errors. [efsops@shou18l560-02 DBIx-Class-0.08195]$ perl -Mblib t/53lean_startup.t DBIx::Class::Schema::throw_exception(): Can't locate object method "table" via package "DBICTest::BaseResult" at t/lib/DBICTest/BaseResult.pm line 11. Compilation failed in require at (eval 1) line 10. ...propagated at /efs/dist/perl5/core/5.14.2- build002/.exec/powerpc32.aix.6/lib/perl5/base.pm line 94. BEGIN failed--compilation aborted at t/lib/DBICTest/Schema/Artist.pm line 4. Compilation failed in require at (eval 1) line 10. at /efs/dist/perl5/Class-C3- Componentised/1.001000/.exec/powerpc32.aix.6/5.14/lib/perl5/Class/C3/Componentised. pm line 155 Compilation failed in require at (eval 1) line 10. BEGIN failed--compilation aborted at t/lib/DBICTest.pm line 7. Compilation failed in require at (eval 1) line 10. BEGIN failed--compilation aborted at t/53lean_startup.t line 99. Clearly, that workaround is NOT general enough, although since require normally raises an exception, it's hard to understand why it doesn't. In any case, I don't really understand how to patch DBIx-Class to fix, or really just workaround, this issue. For now, I've downgraded namespace::clean to 0.20_01, but one of these days, I'll want/need to upgrade, so this will eventually need to be fixed.

On Wed, Oct 26, 2011 at 01:08:07PM -0400, Phillip Moore via RT wrote: Show quoted text
Fuck you base.pm. Ok 2 things here: 1) There was some work on this test done in the DBIC repo. Please checkout current master[1] and see if the issue has been fixed (the require() override is much much cleverer now) 2) Since someone *else* may fuck this up in weird and unknown ways, I am thinking to ship a version of n::c that sets @ISA instead of mucking with use base (given how n::c is a pretty central thing these days). I do not know if this is warranted yet, will ponder with interested parties. Thank you so very much for tracking this down, I am just as dumbfounded as you on why nobody else experiences this... Cheers

On Wed Oct 26 13:08:05 2011, WPMOORE wrote: Show quoted text
Right, we eventually got to the bottom of this, but forgot to close the ticket.The problem was indeed in DBIC itself, because it did not do the require override properly. I nicked the correct way from Zefram's Lexical::SealRequireHints, and even abstracted it into a nicer interface. However I did not put it on CPAN as we couldn't agree with Zefram what is the sane way of going forward. I am putting all the relevant information here, hoping that maybe you will get this to CPAN-dom (it is a very very tricky thing to do right, and it ought to be properly encapsulated). This works: BEGIN { unshift @INC, 't/lib'; require DBICTest::Util::OverrideRequire; DBICTest::Util::OverrideRequire::override_global_require( sub { $_[0]->() }); } use Tie::Hash; use base 'Tie::ExtraHash'; And this is how it is implemented: https://metacpan.org/source/ARODLAND/DBIx-Class-0.08196/t/lib/DBICTest/Util/OverrideRequire.pm