A reimplementation of the classic arXiv abstract (/abs) page as python/Flask application with minimal design changes. The application integrates with the existing database and network filesystem.
Features
The following list of features reproduces the features in the classic /abs implementation:
Author parsing and search link generation
Collapse display of large author lists
Support URL filtering on metadata fields
RDF tag generation
Resolve group/archive/category names where needed from taxonomy defined in arxiv-base
Extra services: generate download links in full-text links section of page
Extra services: license linking
Extra services: browse context links
Extra services: ancillary files
Extra services: trackbacks section
Extra services: references and citations section (INSPIRE, ADS)
Extra services: bookmarks section
Extra services: citation meta tags
Other
Cache control headers for abs page
URLs for resources that might be accessed from multiple repos of arxiv-ng
Replace clickthrough (/ct) route
Logging support
Changes from Classic
There are several minor changes from the classic /abs page display:
The submitter name shown on the "From:" line now reflects the submitter name from the version being viewed; classic only displayed the submitter name from the latest version
UTC is now used instead of GMT in the date line
The enigmatic, single-character format codes are now omitted from the date line; in classic these appeared after the source file size parenthetical