Github encoding breaks pdf embedded fonts

Question

I am using doxygen with MiKTex and ghostscript to create documentation PDFs. These PDF's git pushed to my github repo. However If they are subsequently pulled back down again (e.g. on a different PC) they fail to open correctly as Adobe warns that it could not properly extract the embedded fonts.

I have found that this may be down to the GitHub transfer as the problem does not occur if I simply transfer the PDF's in question via USB key etc. Info from the Adobe forums seems to indicate that this could be down to the PDFs being wrongly encoded as ASCII when it should be binary.

How can I solve this so that the PDF's open correctly when pulled from the GitHub repo? Currently my IDE (eclipse) sets the encoding of the PDF files as UTF-8, should this be changed?

Answer 1

After some testing with the folks at github, it seems that this issue is down to Eclipse changing the encoding of the PDF sometime during the commit or push for some reason.

Using a new repo with GitHub for Windows or Git Bash did not present such problems.