TL;DR

I shared PDF-Collage on Codeberg

After playing a bit with pdfunnel, I figured that having a module would be extremely useful for embedding that kind of capability into wider programs.

Hence, I started working on PDF-Collage, which aims at supporting a wide and extensible set of possible ways of providing “input” templates.

At the very basic level, one template is just a sequence of commands, each represented by a hash reference with an op key to tell what the command is about, as well as its parameters. This is copied almost verbatim from pdfunnel, because it works.

The big itch to scratch was how to gather the different pieces, e.g. pages from a template PDF, because at this point we have to manage how we make sure that the commands in some JSON file remain “close” to the PDF template.

I started playing with the idea of embedding the commands in the PDF, as metadata, and it was working. Anyway it does seem a bit brittle, because I have no way to tell a regular PDF from one ready for being used with pdfunnel.

This is basically the reason why I started working on Data::Resolver, so I could abstract this aspect from the actual PDF manipulation.

I still have to write tests and documentation… so this is more or less alpha. I hope to write more in the coming days.

Stay safe and… cheers!