Mailed with author of Locale::TextDoamin. I think the idea of Locale::TextDoamin::OO is different to Locale::TextDomain.
The ...::OO::Extract is build to extract multiple or all files into one po/pot file. The relation to source is stored as reference in po/pot file. So it is easy to find out where it is from. That means also that you have less but longer files, typical one for each language and sub-language mixed with all files content, pm, html, tt and so on.
Yes it would be possible to write a simple script to extract and write lots of pot files, one of each input file. But that was not the intention of first use in my big web project. There is no domain for each single source file. I run lots of project specific code during extraction, means where to extract and how to group to destination po files.
Please describe your project a little that I can understand your needs better. Is it web or is it command line or is it mixed.
