Customised, daily arxiv summary
The following python script takes as inputs
- a list of authors
- a list of keywords
and prints into the terminal a list of articles, that either contain the authorname, or whose title contains any of the keywords. The list also contains a clickable link that leads to the arxiv PDF file.
The script can also be used with command-line arguments.
import requests
from bs4 import BeautifulSoup
import argparse
def fetch_arxiv_math_na(keywords, authors, subject, max_results=1000):
"""
keywords and authors are strings, and not case sensitive.
max_results valid values: 25, 50, 100, 250, 500, 1000, 2000
subjects (examples):
- math.NA: Numerical Analysis
- math.AP: Analysis of PDEs
- math.DG: Differential Geometry
- math.OC: Optimization and Control
- physics.plasm-ph: Plasma Physics
return:
- prints the title and arXiv link of the papers that match the keywords or authors.
"""
url = "https://arxiv.org/list/"+str(subject)+"/recent?skip=0&show="+str(max_results)
# fetch page
response = requests.get(url)
response.raise_for_status()
soup = BeautifulSoup(response.text, "html.parser")
entries = soup.find_all("dt")
details = soup.find_all("dd")
# extract human readable info
for dt, dd in zip(entries, details):
# arXiv ID and link
link_tag = dt.find("a", title="Abstract") # search for HTML tag <a> with title "Abstract"
if not link_tag:
continue
arxiv_id = link_tag.text.strip()
abs_link = f"https://arxiv.org/abs/{arxiv_id}"
# title
title_line = dd.find("div", class_="list-title mathjax")
if title_line:
title = title_line.text.replace("Title:", "").strip()
else:
title = ""
# authors
author_line = dd.find("div", class_="list-authors")
if author_line:
author_text = author_line.text.replace("Authors:", "").strip()
else:
author_text = ""
# match title or author
title_lc = title.lower()
authors_lc = author_text.lower()
if any(k.lower() in title_lc for k in keywords) or any(a.lower() in authors_lc for a in authors):
print(f"{title}\
\n{author_text}\
\n{abs_link}\n")
if __name__ == "__main__":
# Example usage
keywords = ["finite element",
"fluid",
"navier"
]
authors = ["hiptmair",
"schöberl"]
fetch_arxiv_math_na(keywords, authors, subject="math.NA")
To use the tool with command line arguments, write the following into the main function:
parser = argparse.ArgumentParser(description="Fetch recent arXiv papers by keyword or author")
parser.add_argument("--keywords", nargs="*", default=[], help="List of keywords to search in titles")
parser.add_argument("--authors", nargs="*", default=[], help="List of author names to search")
args = parser.parse_args()
fetch_arxiv_math_na(args.keywords, args.authors, subject="math.NA")
The .py file can be downloaded here.