There is a nice way to work out where the errors and warnings in eval code are coming from.

If you take a look in perlsyn#Plain-old-comments-(Not!), you'll see a feature that allows you to reset Perls idea of the current file and current line.

If you can use this to embed such a comment into each piece of evalable code, you can have perl tell you both the piece of code and the line number within it that at which the warning or error was encountered.

A crude demo...

P:\test>type test.pl8
#! perl -slw
use strict;

my $code = q[
#line 4000 "evalcode.pl"
print "hello";
print undef;
];

eval $code;

P:\test>test
hello
Use of uninitialized value in print at evalcode.pl line 4001.
[download]

By carefully choosing the names you give to your eval blocks (perhaps including the source file name/line number) you should be able to easily translate between the resultant error message and your code.

It would be really nice if it was possible to retroactively add this to eval'd code in CORE and CPAN modules. It sure would save a lot of searching when the messages arise from outside of your own code. Perhaps some sort of recommendation for a standardised filename/eval block/line number to module/source line mapping could be devised?