ArxivParser CLI and API -> uses arXiv search API v0.5.6 to generate CSV of requested articles.

I have an API version of this parser running on my personal python box (mau.pythonanywhere.com). You can toy around with the API using this shared REPL I made.

You can edit the shared REPL. REPL will redirect you to a personal copy of the REPL code. If it hangs, just give it a reload.

The REPL prompt will generate a link you can click that will download the CSV of the given search entry. It will be generically titled, download.csv.

usage: cli.py [-h] [--amount AMOUNT] search_term [start_date] [end_date]

arXivParser CLI

positional arguments:
  search_term      The search term you're using to find articles.
  start_date       The start of the date range used to search for articles.
                   Default is today. Expected format is (YYYY-MM-DD).
  end_date         The end of the date range used to search for articles.
                   Default is a week ago. Expected format is (YYYY-MM-DD).

optional arguments:
  -h, --help       show this help message and exit
  --amount AMOUNT  Maximum number of articles to search for. Due to arXiv
                   limitations, this must be 25, 50, 100, or 200.

python cli.py "Contextual Bandits" 2013-06-02 2016-06-02 --amount 25

Will search for the term "Contextual Bandits" from the date 2013-06-02 until the date 2016-06-02. It will report at most 25 results.

python cli.py "Neural Architecture Search" 2018-01-01 2020-05-02

Will search for the term "Neural Architecture Search" from the date 2018-01-01 until the date 2020-05-02. It will report at most 50 results, the default parameter for amount.

python cli.py "Optimal Transport" --amount 25

Will search for the term "Optimal Transport" from the day a week ago until today. It will report at most 25 results. But also valid is: python cli.py "Optimal Transport" which will return 50 results by default.

python cli.py "Neural Architecture Search" 2018-01-01 2020-05-02 23

Will return an error, since 23 is not a valid number of entries to search for according to arXiv search v0.5.6.

Running python run_tests.py will run a comprehensive suite of tests for the Parser class.