Closed chris-asl closed 7 years ago
@chris-asl It might be easier to use MIME types instead of parsing the output. That is what I did for refextract https://github.com/inspirehep/refextract/commit/10f0ccfb9ff40e9e20a1d1e01a5f1def3c2eff2b.
@michamos Thank you for your suggestion! :+1:
Will hopefully fix: https://sentry.cern.ch/inspire-sentry/inspire-labs/group/822229/
Ensure detect_images_and_tex method ignores hidden/metadata files. Also, refactor to avoid creating a new process for checking the magic string returned from 'file' command line utility, by adding a wrapper library for libmagic.
Closes (#12).
Signed-off-by: Chris Aslanoglou chris.aslanoglou@gmail.com
Sentry: https://sentry.cern.ch/inspire-sentry/inspire-labs/group/821464/