I think that  XMP and PDF/A  require that you "compile" all custom
ontologies into their own magical schema language, and embed them as a Bag
in the XMP packet.  Also, you can't use any ontologies that would require
you to use typed RDF literal (rdf:datatype is forbidden).

My take is:

   - Every ontology from which a class or  field is used requires an entry
    in the  pdfaExtension:schemas Bag.
      - Every class from that ontology which has an instance created used
      requires an entry in the pdfaSchema:valueType sequence.
         - There may be some cases where using  rdf:type is ok, but it
         seems a bit iffy.
      - Every property from that ontology which is   used requires an entry
      of in the pdfaSchema:property sequence
         - 'ObjectProperty's which will have blank node values use one of
         the custom valueType  (element name, not URI)
         - 'ObjectProperty's which will refer to named resources use a URI.
         - 'DataProperty's  that have unspecified or  string range can be
         defined as Text.
         - DataProperties of other types cannot be used. since XMP bans
         rdf:datatype.

I guess the only general solution may be to embed a  single Text type
property that included everything (ontologies + data) as  (XML escaped)
Turtle.

 If all Adobe wanted was XML, why did they bother using RDF?





On Tue, Jun 12, 2012 at 1:29 PM, Ben Companjen <b...@companjen.name> wrote:

> Hi Simon,
>
> Thanks for your reply.
> I have used xmpincl and recently stumbled upon pdfx [1] for producing
> PDF/A compliant files (which require some XMP).
> I started work on extending pdfx's functionality to support some more URIs
> and embedded reference lists [2], but reading through some of the XMP and
> PDF/A specifications I got a bit put off by the limitations imposed on the
> RDF that can be included as XMP in PDF/A and started wondering if it's
> actually possible.
>
> I wanted to say I had found the Citation Typing Ontology, but then saw it
> is part of SPAR.
>
> Ben
>
> [1] 
> http://www.ctan.org/tex-**archive/macros/latex/contrib/**pdfx/<http://www.ctan.org/tex-archive/macros/latex/contrib/pdfx/>
> [2] https://github.com/bencomp/**pdfx-ext<https://github.com/bencomp/pdfx-ext>
>
>
> On 12/6/'12 12:15 AM, Simon Spero wrote:
>
>> You can  include XMP packets in pdftex files using the \pdfcatalog
>> primitive ; there are a couple of macro packages that add a little bit of
>> a
>> wrapper:
>>
>> xmpincl basically just wraps and inserts the contents of an external
>> RDF+XML file as the value of  /Metadata .
>> http://www.ctan.org/tex-**archive/macros/latex/contrib/**xmpincl/<http://www.ctan.org/tex-archive/macros/latex/contrib/xmpincl/>
>>
>> There is no standard ontology for citation data; you might look at
>> BIBO http://purl.org/ontology/bibo/     or
>> SPAR - http://purl.org/spar/
>>
>> Simon
>> On Mon, Jun 11, 2012 at 5:07 AM, Ben Companjen<b...@companjen.name>
>>  wrote:
>>
>>  Hi all,
>>>
>>> I have been wondering whether it is possible to put a bibliography (list
>>> of references in a journal article, for example) in an XMP packet in a
>>> PDF
>>> document. Having read about efforts to get the references from a document
>>> by parsing text, I thought it might make sense (and be possible) to put a
>>> machine-readable version in the document.
>>>
>>> I haven't found anyone who had done this (yet), so I started wondering
>>> why. Is it too obvious to write about? Or is it impossible?
>>>
>>> Eventually I'd like to have it automatically created at the same time
>>> BibTeX creates the 'normal' bibliography for my pdfLaTeX document. And if
>>> the PDF/A standard allows it, have it in such a way that it complies to
>>> the
>>> long-term archival PDF standard. It appears, however, that PDF/A-1 allows
>>> the XMP description of the document itself to have only one author [1],
>>> which makes me doubt the use of PDF/A in general.
>>> If I manage to get it working, could it be a topic of interest for the
>>> Code4Lib Journal?
>>>
>>> Regards,
>>>
>>> Ben
>>>
>>> [1] 
>>> www.pdfa.org/wp-content/****uploads/2011/08/tn0003_**<http://www.pdfa.org/wp-content/**uploads/2011/08/tn0003_**>
>>> metadata_in_pdfa-1_2008-03-****182.pdf<http://www.pdfa.org/**
>>> wp-content/uploads/2011/08/**tn0003_metadata_in_pdfa-1_**2008-03-182.pdf<http://www.pdfa.org/wp-content/uploads/2011/08/tn0003_metadata_in_pdfa-1_2008-03-182.pdf>
>>> >
>>>
>>>

Reply via email to