How export metadata to Adobe XMP format

LiborA

I would like to add correct metadata to my PDF files. Is there a way to export metadata to the Adobe XPM format?
I would like to see new functionality of Zotero: add metadata directly to the PDF attachments similarly as it is available in the Jabref?

adamsmith

I think this is generally a good idea, but given the lack of standards for metadata in XMP, it's probably not super high priority (and it's not currently possible from Zotero)

LiborA

XMP use Dublin core, Jabref use your own format. I think Dublin core is a good basis.

LiborA

I've tried modifying the Unqualified Dublin Core exporter, but I do not know how to achieve the output for authors in the form:
<dc:creator>
<rdf:Seq>
<rdf:li>First Author</rdf:li>
<rdf:li>Second Author</rdf:li>
</rdf:Seq>
</dc:creator>
instead of current output:
<dc:creator>First Author</dc:creator>
<dc:creator>Second Author</dc:creator>

puvalp

This should be high priority, I still have to use Jabref when I want to tag PDF documents with anything beyond the regular tags, can Zotero write the basic standard pdf metadata such as; Title, Author, Subject and Keywords? Even this would finally free me from Jabref.

jabref can read what it has written to the pdf it is an open method that other software can implement as well, that is what is important, even exiftool can read Jabref XMP data. Music managers can write metadata to music files it is about time document and reference managers start taking this seriously.

adamsmith

Why should this be high priority? XMP data is seeing absolutely minimal usage, including by publishers of academic articles. Zotero also provides higher-quality metadata than the DC it could include in XMP tags (which distinguishes it from mp3s).
I think it'd be cool to build this into Zotero, but the idea that XMP support is crucial in any way is belied also by the relative lack of requests for it.

puvalp

Please note that I am not just talking about reading and writing XMP, but even the ability to write basic tags. I think it is essential, I am not in a position at all times to use Zotero for research, when I index my library with something like Recoll it does not look at the Zotero database for information it parses the metadata of the pdf documents, it is not a nice thing to have a search query return a pdf with an ugly cryptic name for its title, usually some part of the filename. I do not name pdf files after the title either, neither does Zotero.


Elsevier have started using XMP, The developer of Calibre had thought that XMP read write was not worth implementing but finally did it because he thought that since a big player like Elsevier is embedding XMP in articles he should give Calibre the ability to read it, he went on to say since the ability to read XMP would result in some filing bug reports about modifying data and not being able to update the associated files, he had to implement the write capability as well.

The Jabref developers are no time wasters to implement this either, neither the exiftool developer, I wonder why these people would spend the time to do this if it was in such low demand, there is even a Latex package to embed XMP in the final pdf. I am not saying that this should be at the top of the priority list, I am saying it should be somewhere on it, when Torvalds was doing his work there was no great demand for another operating system, sometimes people do not know how nice and far reaching something is until they start using it.

The world does not work on just demand, or how many people are using this or that, it also thrives on innovation and insight, Zotero is a lovely innovative piece of software I am just suggesting something along this line that I know will address real needs.

dstillman

I do not name pdf files after the title either, neither does Zotero.

Zotero does by default.

puvalp

Is this something new? Some time ago I poked around and saw some numbered folders and files, I thought Zotfile extension was needed to do any kind of renaming. I am very sorry for misleading information, should have checked again.

dstillman

Zotero has always renamed files you save from the web. We extended that to files you add manually earlier this year, when we turned on PDF recognition by default.

Stored files use random strings for folders, but the files within are named after the metadata, and you can set up a smart folder in your OS to search the PDFs within if you don't want to search from Zotero.

puvalp

dstillman, thanks for the info.