(I had accidentally sent this response to the wrong thread.) First of all, I *really* appreciate all the time and help.
Particularly, thanks for pointing me to the Debugger. Foolishly, I’d just been using Acrobat to look at the properties, which led to some unpredictable results. I *think* that my solution requires setting both the Document Information and the Metadata, due to the behavior of Acrobat, Preview, and other PDF readers. For the sake of posterity (and feedback, if anyone has any further experience or insights), this is what I’m seeing. For an original pdf document with the metadata in the Document Information and XMP matching, if I only set the XMP data, Acrobat seems to prefer the Document Information when it displays properties. If I only set the Document Information, Acrobat prefers/displays the XMP metadata. Consequently, in both cases it appeared to me that my code was not overwriting the existing values, when it really was. Using the Debugger I was able to see the changes to the XMP metadata and/or the Document information. In summary, setting both the XMP metadata and the document information overwrites the existing data (and sets new metadata) without issue and creates predictable results in various pdf readers. Again, thanks for the help! -- Matthew Clemente From: Tilman Hausherr <[email protected]> <[email protected]> Reply: [email protected] <[email protected]> <[email protected]> Date: July 20, 2018 at 10:46:48 PM To: [email protected] <[email protected]> <[email protected]> Subject: Re: Overwriting Metadata Am 20.07.2018 um 23:37 schrieb Matthew Clemente: > Thanks Tilman. > > I set up my code to match yours (it was pretty similar), and I’m > getting the same result. I can’t overwrite existing fields > via XMPMetadata. > > For what it’s worth, I’m using version 2.0.11 of PDFBox and XMPBox; > not sure if that would make a difference. > > I’m assuming, with the approach you’re using, that you are able to > change the Author and Title? Your question is somewhat unclear... or I misunderstood it ... you wrote that you failed with both /Info and XMP /Metadata. With /Info (my small reply) I was able to change just the subject and keep the rest. Does this work or not? With the larger code I replaced the whole metadata and didn't try to replace just a single field. Possible explanation: you looked at the PDFs with Adobe Reader. IIRC that one displays what's in the XMP metadata first, i.e. if there is /Info and /Metadata What do you really want, replace an individual field or replace the whole metadata? To alter individual fields, this should work like this: XMPMetadata xmp = xmpParser.parse(meta.createInputStream()); DublinCoreSchema dc = xmp.getDublinCoreSchema(); if (dc != null) dc.setDescription("descr"); else /// do as before I took that code from the ExtractMetadata example from the source code download. (I didn't test. It's in the middle of the night and I couldn't sleep) Tilman > > -- > Matthew Clemente > > From: Tilman Hausherr <[email protected]> > <mailto:[email protected]> > Reply: [email protected] <mailto:[email protected]> > <[email protected]> <mailto:[email protected]> > Date: July 20, 2018 at 4:14:48 PM > To: [email protected] <mailto:[email protected]> > <[email protected]> <mailto:[email protected]> > Subject: Re: Overwriting Metadata > >> It works for me... here's my code: >> >> >> import java.io.ByteArrayOutputStream; >> import java.io.File; >> import java.io.IOException; >> import javax.xml.transform.TransformerException; >> import org.apache.pdfbox.pdmodel.PDDocument; >> import org.apache.pdfbox.pdmodel.common.PDMetadata; >> import org.apache.xmpbox.XMPMetadata; >> import org.apache.xmpbox.schema.DublinCoreSchema; >> import org.apache.xmpbox.xml.XmpSerializer; >> >> public class ChangeMeta >> { >> public static void main(String[] args) throws IOException, >> TransformerException >> { >> PDDocument doc = PDDocument.load(new File("testing.pdf")); >> XMPMetadata xmp = XMPMetadata.createXMPMetadata(); >> DublinCoreSchema dc = xmp.createAndAddDublinCoreSchema(); >> dc.setDescription("descr"); >> XmpSerializer serializer = new XmpSerializer(); >> ByteArrayOutputStream baos = new ByteArrayOutputStream(); >> serializer.serialize(xmp, baos, true); >> PDMetadata metadata = new PDMetadata(doc); >> metadata.importXMPMetadata(baos.toByteArray()); >> doc.getDocumentCatalog().setMetadata(metadata); >> doc.save(new File("testing-new.pdf")); >> } >> } >> >> And the proof that it worked: >> >> >> >> >> Tilman >> >> Am 20.07.2018 um 21:15 schrieb Matthew Clemente: >>> Forgive me if this question has an obvious answer; perhaps I’m not taking >>> the right approach to the problem. >>> >>> My goal is to save a version of a pdf, with modified metadata. In most >>> cases, I’ll be removing metadata (setting the author, title, description to >>> blank), though in some cases I’ll be adding information to those fields. >>> >>> I’ve tried both approaches from these StackOverflow answers: >>> https://stackoverflow.com/questions/40295264/how-to-add-metadata-to-pdf-document-using-pdfbox >>> >>> >>> That is, I’ve tried creating the metadata via XMPMetadata and using >>> importXMPMetadata(). I’ve also tried using the Document Information object >>> (inputDoc.getDocumentInformation().setCreator("Some meta”);). >>> >>> In both cases, if the field is empty in the original document I’ve loaded, >>> the new value is set without issue. However, if the metadata field already >>> contains a value, the new value is not applied. >>> >>> Is there a way for me to overwrite metadata, or am I approaching this all >>> wrong? >>> >>> Here’s a pdf I was using while testing (it has a title and author set, but >>> no subject):https://www.dropbox.com/s/olk2zhnh47ohtpk/testing.pdf?dl=0 >>> >>> Thanks, in advance, for any insight. >>> >>
