On Fri, May 13, 2011 at 2:40 PM, Tim Mooney <Tim.Mooney@xxxxxxxx> wrote:
... snip ...
Thanks for all the replies I received. In the end, I felt the best way
to address this issue was with altering the Perl source rather than
to 'lie to RPM', or to post-process the requirements generator
output.
In the end, I simply extracted that one line from within its long 'hereDoc'
and placed it inside of a simple print statement. It now prevents
the requirements generator from thinking it is a real requirement.
Ie.
print "use Net::SNMP::AgentX qw(:types :pdus :errors :options);\n";
In regard to: how to prevent rpmbuild from incorrectly discovering a Perl...:Perl is incredibly challenging to parse correctly, which is why the
What I have in one of my Perl programs something like:
print <<EOF;
use Net::SNMP::AgentX qw(:types :pdus :errors :options);
EOF
So my code 'emits' a line of text that contains a 'use' statement,
but it does not actually ever execute that 'use' statement,
therefore _my_ package is NOT dependent on that module.
dependency generator sometimes gets it wrong.
The easiest thing to try would be to break your very small here-doc into
pieces that the dependency detector won't notice. Something like
... snip ...
Thanks for all the replies I received. In the end, I felt the best way
to address this issue was with altering the Perl source rather than
to 'lie to RPM', or to post-process the requirements generator
output.
In the end, I simply extracted that one line from within its long 'hereDoc'
and placed it inside of a simple print statement. It now prevents
the requirements generator from thinking it is a real requirement.
Ie.
print "use Net::SNMP::AgentX qw(:types :pdus :errors :options);\n";
_______________________________________________ Rpm-list mailing list Rpm-list@xxxxxxxxxxxxx http://lists.rpm.org/mailman/listinfo/rpm-list
