Dag Wieers wrote: > On Mon, 18 May 2009, Christoph Maser wrote: > >> Am Sonntag, den 17.05.2009, 14:50 +0200 schrieb John Thomas: >> >>> You are correct sorry. I had to install >>> perl-Class-C3 >>> perl-Algorithm-C3 >>> perl-Sub-Name >>> perl-Devel-GlobalDestruction >>> then it worked fine. >>> >>> I wonder if those should be additional dependencies. >> >> Dag, shall i add those or should we rather try to find out what is wrong >> with %__perl_requires? >> >> Here ist what is found in my x86_64 mock environment: >> >> DEBUG: Requires(rpmlib): rpmlib(CompressedFileNames) <= 3.0.4-1 >> rpmlib(PayloadFilesHavePrefix) <= 4.0-1 rpmlib(VersionedDependencies) <= >> 3.0.3-1 >> DEBUG: Requires: perl >= 0:5.008001 perl >= 3:5.8.1 >> perl(Data::AMF::Formatter) perl(Data::AMF::Formatter::AMF0) >> perl(Data::AMF::Header) perl(Data::AMF::IO) perl(Data::AMF::Message) >> perl(Data::AMF::Parser) perl(Data::AMF::Parser::AMF0) perl(Moose) >> perl(Scalar::Util) perl(bytes) perl(constant) perl(strict) >> perl(warnings) >> >> >> Not what it should be > > Indeed it would be great if we could improve the %__perl_requires script > upstream, but since it is unlikely to be backported to RHEL3, RHEL4 and > RHEL5 we should probably add the missing dependencies too. > > The script is known to produce both false negatives as well as false > positives. And it is hard to filter out dependencies if there are false > positives :-(
The perl.req scripted used by rpm does not even try to parse "use base" calls. At one time I put some code in place on a development box that tried to do this without changing the whole script around. It was not able to handle multi line use base calls and was only able to pull the first module if multiple inheritance was used. The perldeps.pl script which is not used by default seems to handle use base properly but has issues in other ways. I have recently noticed this: http://search.cpan.org/~atourbin/rpm-build-perl-0.70/ I have not tried it but it appears to have a proper replacement for perl.req and since a linux distro actively maintains it I assume it is actively used. -- David Steinbrunner _______________________________________________ users mailing list [email protected] http://lists.rpmforge.net/mailman/listinfo/users
