invenio released versions sanity check

invenio-releases.py

# Iterate through all invenio repositoried and find which ones changes since
# the since_date. Also find out which repositories have their last commits
# which are not releases

# NOTE: pip install PyGithub

import os
import json
from datetime import datetime
from github import Github

g = Github(os.environ["GITHUB_TOKEN"])
since_date = datetime(2019, 11, 10)

# Repositories which have a commit since the since_date param
dirty_repos = []

# Repositories whose last commit is not a release
not_released = []

# This is a list of not used repositories (i.e. archived) which cause problems
skip_repositories = ['coordination']


def run():
  repos = g.get_organization("inveniosoftware").get_repos()
  print("{:<42} {}".format("Repository", "Last Commit"))
  for repo in repos:
    if repo.raw_data['name'] in skip_repositories:
      continue
    commits = repo.get_commits(since=since_date) or []
    if commits and commits.totalCount > 0:
      dirty_repos.append(repo.name)
      last_msg = commits[0].raw_data['commit']['message'].replace('\n', ' ')
      print("{:<42} {}".format(repo.name, last_msg))
      if not last_msg.lower().startswith('release'):
        not_released.append(repo.name)

  with open('not_released.json', 'w') as f:
    json.dump(not_released, f)


if __name__ == '__main__':
  try:
    run()
  except Exception as e:
    print(e)
    pass