PDF metadata---How and can XMP be done in free software?

James Howison james at freelancepropaganda.com
Wed Oct 22 18:08:05 EDT 2003


I working on a project to put citation metadata into academic pdfs (to 
help me manage the pdfs that I download).  The basic idea is, "Why 
can't I manage academic PDFs the way I manage MP3s?".

I'm blogging about this here: http://shangorilla.syr.edu/themp/

Relevance to this list is that we're both trying to get metadata into 
PDF files and that, eventually, I'd like to incorporate CC academic 
style licences as I imagine is intended to eventually happen here:

http://www.pubmedcentral.com/articlerender.fcgi?artid=212690

But for now I'd be happy just to get the citation inside the PDF file 
so that applications like http://bibdesk.sf.net can file downloaded 
papers away nicely.

There are two ways that I can think of to go about this and I have 
questions relevant to each that I hope people here can help me with:

1.  Kludge for proof of concept:  Add bibtex record to PDF document 
information field.  I can do this using the perl PDF::API2 (as 
described on the blog here: 
http://shangorilla.syr.edu/archives/themp/000206.html) but I can't find 
an open source C library capable of doing this (clibpdf and the free 
PDFlib both, I think, cannot edit existing files).

Anyone know a free c library that can read/write/edit the Document 
Information Dictionary for PDFs?  (or maybe it is really easy and I 
don't need a library ... pointers?)

2.  Use XMP after converting the bibtex (or other biblio format to rdf)
(maybe using something like http://www.cs.vu.nl/~mcaklein/bib2rdf/)

But I'm quite confused about how to add XMP packets to PDF files using 
a totally open source process (ie without access to the Acrobat API). 
Adobe's current answer seems to be, "It's really hard and we're not 
interested in supporting it (at least using their XMP toolkit)". I 
ranted about this on the Adobe boards and mirrored that post here:

http://shangorilla.syr.edu/archives/themp/000203.html

I may well have gone off half-cocked about this, so my question to you 
rdf and metadata gurii considering adding CC licences to PDFs,

How are you planning on getting CC licences into PDFs?
Are you planning on using XMP?
What open source libraries or tools are you planning on using to do it?

If people have comments on the main aim of the themp project (or a link 
to it already working ;) I'd be really interested in hearing from you 
either on the blog or by email.

Thanks in advance,
James





More information about the cc-metadata mailing list