daily arxiv on the remarkable

arxivtorm.py

#!/home/aaron/miniconda3/bin/python

# How to:
# 1. Get python running on your computer
# 2. Install arxiv: $ pip install arxiv
# 3. Get pandoc running!
# 4. Install rmapi
# 5. change the RMAPI variable to point to the rmapi executable (line 18)
# 6. change line 1 to point to your python interpreter ($ which python), make the script executable, etc.
# 7. change the QUERY variable to query arxiv to whatever you like (line 19)
# 8. Setup a cron job!

import arxiv
from datetime import datetime, timedelta
import pytz
import os

RMAPI = '/home/aaron/bin/rmapi'  # CHANGE to the install directory of rmapi
QUERY = 'cat:astro-ph.EP'

today = datetime.now()

replace_dict = {r'\[': r'$',
                r'\(': r'$',
                r'\]':r'$',
                r'\)':r'$',
                # '$ ':'$',
                # ' $':'$',
                '$\n':'$',
                r'\unit':'',
                }


def get_results():
    delta_yest = 1 if today.isoweekday() != 1 else 3  # account for weekends!
    yesterday = today - timedelta(days=delta_yest)
    yesterday = yesterday.replace(minute=0, hour=18, second=0, microsecond=0)
    delta_q = 1 if yesterday.isoweekday() != 1 else 3  # account for weekends!
    q_day = (yesterday - timedelta(days=delta_q))

    timezone = pytz.timezone("GMT")
    q_day = timezone.localize(q_day)
    yesterday = timezone.localize(yesterday)

    search = arxiv.Search(
        query = QUERY,
        max_results = 100,
        sort_by = arxiv.SortCriterion.SubmittedDate
    )
    filtered = [result for result in search.results() if result.published > q_day and result.published < yesterday]
    return filtered, yesterday, q_day


def upload(filtered, yesterday, q_day, delete=True):
    with open('arxiv.md','w') as f:
        f.write(f"\n\n=================================================\n\n")
        f.write(f"# Submissions to: {QUERY}\n\n")
        f.write(f"## received from {q_day.strftime('%d.%m.%Y, %H:%M')} to {yesterday.strftime('%d.%m.%Y, %H:%M')} GMT\n")
        f.write(f"\n\n=================================================\n\n")

        for res in filtered:
            summary = res.summary
            for r_key, r_val in replace_dict.items():
                summary = summary.replace(r_key, r_val)

            f.write("\n\n----------------------------------------------\n\n")
            f.write(r"### {}".format(res.title) + "\n\n")
            f.write(r"*{}*".format(', '.join([str(a) for a in res.authors]))+"\n\n")
            f.write(r"{}".format(summary)+"\n\n")
            f.write("#### comments:\n > "+r"{}".format(res.comment) + "\n\n\n")

    r = os.system('pandoc arxiv.md -o arxiv.pdf --pdf-engine=xelatex')
    if r != 0:
        print('something went wrong, when creating the pdf, using gfm instead')
        os.system('pandoc arxiv.md -o arxiv.pdf --pdf-engine=xelatex -f gfm')
    if delete:
        os.system('rm arxiv.md')
    os.system(f'{RMAPI} rm arxiv')
    os.system(f'{RMAPI} put arxiv.pdf')
    if delete:
        os.system(f'rm arxiv.pdf')


if __name__ == "__main__":
    if today.isoweekday() in range(1,6):
        filtered, yesterday, q_day = get_results()
        upload(filtered, yesterday, q_day, delete=False)
        print(f"received from {q_day.strftime('%d.%m.%Y, %H:%M')} to {yesterday.strftime('%d.%m.%Y, %H:%M')} GMT\n")
    else:
        print('Weekend!')