PDF using the latest standard 2.0 are not supported

[FrankHossfeld]

I'll use OpenPDF to create a PDF from scratch, add the PDFA/1 stuff by adding the relevant tags and use the mustungproject to create the eRechnung part. The PDF is a valid PDFA/1 (checked with several tools) and everything works fine.

The latest version of OpenPDF now support the new (from 2017 ... ;-)) PDF standard 2.0. In opposite to the standards before, a PDF starts now with %PDF-2 and no longer with %PDF-1. Running new ZUGFeRDExporterFromA1().load()) now fails. The validation is saying: Header Syntax error, First line must match %PDF-1.\d. Which is correct, bc it is now %PDF-2.

It looks like the lib used to validate PDFs does not support the latest PDF standard.

I'll checked the genereated PDF aganist several chekers. They all confirm, that the document is valid and the conformance level applied is PDF/A-1b.

[jstaerk]

Can you please submit your test file?
I've been looking for one for years and I'd actually love to submit it to https://github.com/ZUGFeRD/corpus/ as well because so far nobody in the standardization committee (cc3) could get hold of a sample yet.

[FrankHossfeld]

sure, should it be an invoice or just only a PDF version 2.0?

[jstaerk]

Both would be great if you could. Thanks!

[jstaerk]

updated my libreoffice found out it already supports pdf 2

MustangGnuaccountingBeispielRE-20190610_507.pdf

[jstaerk]

created and pushed a branch, added a test, verapdf validates the input but not the output

[jstaerk]

Tilman Hausherr not only fixed https://issues.apache.org/jira/browse/PDFBOX-6088, he also provided a work around: (thank you very much). I now added a ZUGFeRDExporterFromA4 and now I feel I am stuck by not being able to set the conformance level to f https://issues.apache.org/jira/browse/PDFBOX-6090 (verapdf "All of the embedded files shall be compliant with ISO 19005-1, 19005-2 or 19005-4").

However, the following verapdf errors on the generated (run ZF2EdgeTest:testPDFA4Export) file (library/target/testout-ZF2PDFA4.pdf) could be workable in parallel:

• Specification: ISO 19005-4:2020, Clause: 6.1.3, Test number: 4 The Info key shall not be present in the trailer dictionary of PDF/A-4 conforming files unless there exists a PieceInfo entry in the document catalog dictionary Failed 1 occurrences Hide CosDocument containsPieceInfo == true || containsInfo == false root The Info key present in the trailer dictionary, but PieceInfo entry does not present in the document catalog dictionary
• Specification: ISO 19005-4:2020, Clause: 6.1.3, Test number: 5 If a document information dictionary is present, it shall only contain a ModDate entry Failed1 occurrences HideCosInfo size == 1 && ModDate != null root/trailer[0]/Info[0]Document information dictionary is present and contains CreationDate,Producer,ModDate,Creator key(s), but allowed only ModDate
• Specification: ISO 19005-4:2020, Clause: 6.7.3, Test number: 5 The value of "pdfaid:rev" shall be the four digit year Failed 1 occurrences HidePDFAIdentification /^\d{4}$/.test(rev) root/document[0]/metadata[0](172 0 obj PDMetadata)/XMPPackage[0]/PDFAIdentification[0] The value of "pdfaid:rev" (null) is not the four digit year

Does anybody want to contribute? Some of them might be easy and I don't have much time to work but I might have some time to explain.