Paper import from Neurips/Nips proceedings

[PantelisElinas]

Currently, when importing a paper from an external URL, e.g., arXiv.org, the import functionality runs on the server.

Retrieving the paper meta-data from an external source, however, can take a significant amount of time depending on the responsiveness of the paper server. This unnecessarily blocks Gnosis' web server from fulfilling other requests.

A solution to this is to move this operation to the client.

Implement the import functionality to the client via Javascript for importing papers from neurips.cc and nips.cc online proceedings. Note that the proceedings are accessible via the sub-domains papers.neurips.cc and papers.nips.cc. We should support import from both domains.

Note: For older papers, the abstract may be missing. See for example the following,

http://papers.neurips.cc/paper/109-applications-of-error-back-propagation-to-phonetic-classification

In such cases, we should identify the error and not import the paper informing the user.

Done Checklist

• [ ] Implemented paper import in JS
• [ ] Connected the JS code to the import button on the website
• [ ] Unit tests

[PantelisElinas]

I'm closing this because it has been tackled in #39