mmarcu10.py

import json, re
import requests
from urlextract import URLExtract
import sys, gzip


utid = 'mmarcu10'
base= { 'model':'https://huggingface.co/', 'data': 'https://huggingface.co/datasets/', 'source': 'https://' }
post = '/raw/main/README.md'
postGH = 'blob/master/README.md' # or it could be 'blob/main/README.md'

extU = URLExtract()
DOIpattern = r'\b(10\.\d{4,9}\/[-._;()/:A-Z0-9]+)\b/i'
#r1\b(10[.][0-9]{4,}(?:[.][0-9]+)*/(?:(?!["&\'<>])[[:graph:]])+)\b'

def extractURLs (c):
 res = extU.find_urls (c)
 return res

def extractDOIs (c):
 res = re.findall (DOIpattern, c)
 return res

fo = gzip.open(f"output/{utid}2.json.gz", 'w')

def run (tp):
 post0 = post
 with open(f"input/{utid}_{tp}", 'r') as f:
  for line in f:
   line = line.strip ()
   if tp == 'source':
    (npapers,line) = line.split(';');
    post0 = postGH
   print(line)
   url = base[tp] + f"{line}{post0}"
   try:
      r = requests.get (url)
      content = r.text
      #github returns repos that do not exist, need to detect that here
      #github when you give master instead of main, that might cause issues as well
      if r.status_code != 200 and tp == 'source':
         url = base[tp] + f"{line}" + 'blob/main/README.md'
         r = requests.get (url)
         content = r.text
      urls = extractURLs(content)
      dois = extractDOIs(content)
      res = { 'ID': line, 'type': tp, 'url': url, 'content': content, 'links': urls, 'dois': dois }
      out = json.dumps(res, ensure_ascii=False)
      fo.write((out+"\n").encode())
   except requests.exceptions.RequestException as e:
      print(f"Request failed for '{url}': {e}")

run('model')
run('data')
run('source')