Repository navigation
The direct ingest of ArXiv records was producing malformatted records for the case where the HTML entities > and < were embedded in LaTeX; the parser was converting these entities back to ascii, which is not the desired option. This release adds a post-parsing process for the title and abstract fields that detects these cases, and if so, uses cgi.escape (python2.7) and re-encoding with xmlcharrefreplace to convert them back to HTML entities.