The Astrophysics Data System (ADS) is the main archive of papers published in astrophysics (and many other disciplines, because it also archives many other arXiv disciplines also), providing many useful metadata on the publications.

Fortunately ADS also has a query API that is accessible via download programs like cURL: https://github.com/adsabs/adsabs-dev-api

Users just need to register on ADS first, and get a unique "key" (which looks like an SHA-1 checksum) that is assigned to them. Afterwards, they can directly use a 'curl' command to search within ADS and download their necessary paper (or metadata). There are several examples in the '.ipynb' files of the repository above that we can use to define a high-level interface for the query.

For the key, we can add an option to Query like '--adskey' to let the user specify their unique key which we will use to construct the download command. This option (and its value) can be stored in any of Gnuastro's directory, user or system configuration files. Therefore, users don't have to specify it on every call.

We can add a new set of paper-database specific options for querying paper databases like ADS (fore example instead of the existing '--database' we can have '--paper').

Here are some possible usages of adding ADS support:

With the command below we can get a list of papers by a certain first-author to get their precise ADS identifier.

astquery ads --author="^AUTHOR NAME"

With the following command, the raw title, author list and abstract of a desired paper (based on its ADS ID) can be downloaded and displayed on the terminal as if it was a plain-text file on their computer.

astquery ads --paper=2020arXiv200603018A --abstract

With the following command we can directly download the BibTeX information for this paper and write/append it into a 'paper.bib' with the other papers that a user wants to cite in their paper.

astquery ads --paper=2020arXiv200603018A --bibtex >> paper.bib

ADS apparently gives its outputs in JSON format, so it is good to add JSON support in Gnuastro to easily parse the outputs. I just defined task #15895 for this job.

This task was originally suggested by Javier Moldon and the links above were found with the help of Sergi Blanco-Cuaresma (from ADS).