Description
I tried to follow workaround suggested in PDFBOX-3138 to be able to display czech national characters in PDF form field. I created sample form in OpenOffice 4.1.3 (file sampleform.pdf) and I created code that adds Type 0 font to existing PDF font and then it replaces current font used in form field.
I've got 3 problems:
1. After using PDFont font = PDType0Font.load(doc, new File("fonts/TIMESS__.ttf"));
the font is not loaded into PDF document resources. It is loaded there once it is used in content stream. OK, I created new dummy page and created dummy text. Finally I can remove the dummy page and the font is still loaded in PDF document resources.
2. Czech characters are not displayed, you can see just strange square chars. I investigated once I use the czech chars somewhere in content stream in my dummy text then all is displayed correctly (file sampleform_filled.pdf).
Ok, I did the workaround and I have now file sampleform_filled_workaround.pdf
3. Now when in Acrobat Reader I click in field with correctly displayed text I can see error "Cannot extract the embedded font 'AAAPNQ+TimesSansSerif'. Some characters may not display or print correctly." This can be workarounded if I flatten the form using acroForm.flatten(); ( file sampleform_filled_workaround_flatten.pdf)
I'm not sure about root causes for all 3 issues but the behaviour is strange so these issues should be investigated. Attaching java project and PDFs