How do I add the full dublin core metadata set to a document and save it in PDFA?

[Guest]

We acquired a trial version of Acrobat DC to test whethet it would be useable for records management and archiving.

According to this webpage : PDF/A Metadata XMP, RDF & Dublin Core | PDF Association , Acrobat supports the 15 basis Dublin Core elements.

But when we tested the metadata appending process, only a very limited number of dc metadata was made available.

I then defined an XMP file with all dublin core properties, but when I append it to a document, it is not possible to assign individual values to the dc properties as they do not appear anywhere in a dialog box.

Moreover, when, after saving the document in question in PDF/A format, I look up its properties, all custom-defined metadata are not rendered so I assume they are lost.

How can I append my own additional, business-defined metadata scheme to a PDF/A document and have the possibility to edit its values in a dialog box as I can with the predefined xmp metadata?

If additional metadata cannot be included in the PDF/A file,  the whole concept of using pdf/A for longterm preservation of information is rendered void.

[David W. Goodrich]

Metadata is a specialized issue within PDF whose potential value generally goes unappreciated.  As you have already discovered, Acrobat itself just scratches the surface, allowing easy access to just a few of the Dublin Core categories.  There are a number of vendors offering software specifically designed for editing PDF metadata, and perhaps one of their offerings will do what you want.

It is important to note in your case that PDF/A imposes particular limitations of metadata.  Back in 2012, Dave Merchant wrote in these forums that "PDF/A only permits a subset of the XMP schema, if the document has anything else in it the standards check will fail."  At that time he expressed hope the situation would improve, but I don't know that it has.  Obviously, you will want to confirm what categories of metadata the most recent PDF/A permits, and whether the software you contemplate using to certify your PDFFs can handle that version.  You may find it simpler to fit your customizations within what PDF/A permits.

Note, too, that even though Acrobat DC may not provide dialog boxes for altering XMP metadata you may still need it to check that metadata you add really exists in the file: software for modifying PDFs is not always as careful as Adobe products, and may lose items that are important to you.  Acrobat, however, can at least show you custom metadata in a PDF even when it cannot modify it directly.  For example, see the third screen grab in a long entry I wrote last summer over on the InDesign forum, which shows a bunch of bibliographical information encoded according to the PRISM standard, including section, startingPage, and endingPage.

As I say, Acrobat itself does not provide direct access to those categories, but there is an indirect, down-and-dirty way to change, say, the value for startingPage: have Acrobat export the metadata to a *.xmp file, modify that value in a text editor, save, and then use Acrobat to import the modified *.xmp file.  As the 2012 thread implies, if you can figure out the structural requirements you can extend this to embed a custom namespace with custom categories -- and see if they validate as PDF/A.

For PRISM metadata I don't need to do go through the export/modify/import *.xmp routine because, as the 2nd screen grab shows, I have a custom panel for InDesign, my workhorse program; the panel "exposes" these categories and data I add there is automatically inserted when I export to PDF.  (I'm no programmer, so it took me a while to modify the generic panel that Adobe bundles with InDesign -- and Acrobat's user interface works differently.)

None of this is simple or direct.  What InDesign calls Keywords correspond to Dubline Core's "Subject", a so-called "bag container".

Good luck!

David