Add a URL parser

A URL parser sounds rather general but maybe we could try some simple redirection logic from a given URL. E.g. arXiv URLs can easily be identified and then forwarded to the arXiv parser. DOI URLs are also recognizable. But general Journal landing pages may be tricky because they may not contain the DOI itself..

We could also think about direct URL support on (e.g.) the arXiv parser such that it can take either a pure arXiv ID or the full URL which could greatly increase the usability.