I've updated pdfcpu to today's repository. Running WSL/Ubuntu

When I fill the attached form f8949lt.pdf using:
pdfcpu form fill f8949lt.pdf fields.json filled.pdf
The € character in the filled field causes the field's contents to display as "☒" in Acrobat. It is not a problem to fill other pdfs using this character, nor is it a problem to manually type the character into a field in the form. I can only guess what the issue might be. Probably the file is not expecting non-ANSI characters, and maybe does not have appearance streams generated for UTF-16 (font information indicates all are ANSI encoded). Ordinarily, I think pdfcpu generates appearance streams, but in this case perhaps it does not notice missing characters. Or possibly pdfcpu writes the UTF-16 value, but it maps to a non-printable character in the ANSI appearance stream - though I have no idea why this corrupts all of the characters in the field.

f8949lt.pdf

fields.json

filled.pdf