We write a software that create PDF files. How we can check if the resulting pdf files are PDF/A compatible? Are there any test suite for it available?
-
2Some context for people unfamiliar with PDF/A: http://en.wikipedia.org/wiki/PDF/A – Chris Dolan Feb 21 '09 at 06:29
-
Related: http://stackoverflow.com/questions/464539/convert-to-pdf-a-and-check-compliance-under-linux – Mechanical snail Aug 16 '12 at 01:24
-
http://www.access-for-all.ch/en/pdf-lab/pdf-accessibility-checker-pac/downloading-pac.html is a tool that validates lots of aspects and explains why things have failed. – Mike de Klerk Sep 11 '15 at 07:11
-
"Moved" to http://softwarerecs.stackexchange.com/questions/18161/is-there-free-or-open-source-software-for-checking-pdf-a-compliance – Nemo Oct 05 '15 at 16:15
-
2http://verapdf.org/ is in the process of creating one. – Thorbjørn Ravn Andersen Dec 08 '16 at 09:20
5 Answers
A list of PDF/A validators is on the pdfa.org web site here:
A free online PDF/A validator is available here:
A report on the accuracy of many of these PDF/A validators is available from PDFLib:
Se as well:

- 1,616
- 1
- 18
- 46

- 1,853
- 15
- 10
-
3
-
3Now known as PDFBox Preflight https://pdfbox.apache.org/downloads.html#recent – Tilman Hausherr Nov 17 '14 at 22:11
-
4Unfortunately the link at pdf.org is not available any more. At http://verapdf.org you will probably find the most "official" online tool for PDF validation. It is quite new and afaik in Beta. Until this changes I also stick with the preflight-validation of Adobe Acrobat – rominator007 Mar 01 '17 at 10:17
The 3-Heights™ PDF Validator Online Tool provides good feedback for different PDF/A conformance levels and versions.
- PDF/A1-a
- PDF/A2-a
- PDF/A2-b
- PDF/A1-b
- PDF/A2-u

- 4,501
- 6
- 50
- 90
-
1
-
2
-
Very nice. This gives me a list of "violations" of PDF/A for my file, such as _The value of the key Parent is 271 0 R but must be 208 0 R._ But I have no idea how to "fix" these. I tried saving in the free adobe acrobat but it doesn't help. – Michael Jul 23 '18 at 23:10
-
1Here is something that just worked for me on my mac laptop: I opened my (ordinary) PDF file in the mac "Preview" software. I "printed" it as a .ps file. I then re-opened the .ps file in Preview and saved as PDF. It now passes the pdf/a compliance. – Michael Jul 23 '18 at 23:40
-
1Thanks @Michael, that was helpful! For those on Linux, I tried something similar which also worked: I opened the file on [evince](https://wiki.gnome.org/Apps/Evince), and printed it to a PDF file. The resulting file passed PDF/A validation! Note that simply doing "File > Save as" was not sufficient; I had to use the Print menu for this to work. – waldyrious Aug 01 '19 at 21:45
pdf validation with OPEN validator:
DROID (Digital Record Object Identification) https://github.com/digital-preservation/droid/
JHOVE - JSTOR/Harvard Object Validation Environment https://jhove.openpreservation.org/
-
1DROID doesn't validate, it only identifies. JHOVE does validate PDFs, but not PDF/A. – Andy Jackson May 30 '13 at 14:54
-
2Jhove does validate PDF/A. Check there site: [link](http://jhove.sourceforge.net/pdf-hul.html).... – Ibn Saeed Oct 26 '14 at 08:17
-
1Ibn Saeed: citing from the linked doc: Note that the PDF module does not parse the contents on streams, so it cannot determine conformance to PDF/A to the degree required by the ISO standard. – mykhal Jun 29 '15 at 14:33
-
1@mykhal «That is not possible, since parts of the PDF/A rules are not machine-checkable AFAIK (like "metadata matches data" -- jhove is unable to understand the semantics of your document).» http://www.konradvoelkel.com/2013/03/scan-to-pdfa/#comment-2705 – Nemo Oct 05 '15 at 15:09
-
1Jhove's PDF/A module is quite buggy. Its output is nearly meaningless. – jbarlow Feb 20 '17 at 21:58
Do you have Adobe PDFL or Acrobat Professional? You can use preflight operation if you do.

- 108,024
- 16
- 131
- 187
If you download the latest version of Adobe Acrobat Reader, it will tell you if your pdf is PDF/A compliant. Just open the PDF file and a big blue marking should appear.
OpenOffice supports PDF/A. For some reason "PDF/A-1" is called
"SelectPdfVersion"internally in OpenOffice. Just add 1 to that value and your output should be PDF/A.
The different values can be
0 = PDFXNONE
1 = PDFX1A2001
2 = PDFX32002
3 = PDFA1A
4 = PDFA1B
You set
FilterDatato be a
HashMap('SelectPdfVersion',1) //1 for PDFX1A2001

- 23,901
- 30
- 103
- 143
-
6This is wrong. The Reader show only if the PDF/A signature is available. Not if it is a valid PDF/A file. – Horcrux7 Mar 18 '10 at 12:50
-
@Horcrux7 can you site your sources for this? How do you know this? I have used a few API's that generate PDF/A and non PDF/A compliant files. When I produce a PDF/A file, Adobe Reader 10 will launch with an informational bar at top:O "The file you have opened complies with the PDF/A standard and has been opened read-only to prevent modification". Can you share what you know about this?L – JohnZaj Jul 19 '12 at 03:35
-
1We have write a software that generate PDF and PDF/A files (see my original posting). That I know that also not PDF/A compliance files was show as PDF/A file in Adobe Reader. We need only write the PDF/A signature. – Horcrux7 Jul 25 '12 at 20:07