zulip/tools/fetch-contributor-data

#!/usr/bin/env python3
"""
Fetch contributors data from Github using their API, convert it to structured
JSON data for the /team page contributors section.
"""
import os
import sys

sys.path.insert(0, os.path.join(os.path.dirname(__file__), '..'))
from scripts.lib.setup_path import setup_path

setup_path()

import argparse
import logging
from datetime import date
from random import randrange
from time import sleep
from typing import Dict, List, Union

from typing_extensions import TypedDict

os.environ['DJANGO_SETTINGS_MODULE'] = 'zproject.settings'
import django

django.setup()

import json

import requests
from django.conf import settings

duplicate_commits_file = os.path.join(os.path.dirname(__file__), 'duplicate_commits.json')

parser = argparse.ArgumentParser()
parser.add_argument('--max-retries', type=int, default=10,
                    help='Number of times to retry fetching data from Github')
args = parser.parse_args()

class ContributorsJSON(TypedDict):
    date: str
    contrib: List[Dict[str, Union[int, str]]]

class Contributor(TypedDict):
    avatar_url: str
    contributions: int
    login: str

logger = logging.getLogger('zulip.fetch_contributors_json')

def fetch_contributors(repo_link: str, max_retries: int) -> List[Contributor]:
    contributors: List[Contributor] = []
    retry_attempts = 0
    page_index = 1

    while True:
        response: requests.Response = requests.get(f"{repo_link}?page={page_index}", verify=os.environ.get('CUSTOM_CA_CERTIFICATES'))
        if response.status_code == 200:
            data = response.json()
            if len(data) == 0:
                return contributors
            contributors.extend(data)
            retry_attempts = 0
            page_index += 1
        else:
            retry_attempts += 1
            if retry_attempts > args.max_retries:
                logger.warning("Failed retries fetching contributors data from Github.")
                sys.exit(1)

            sleep_time = randrange(0, min(64, 2**retry_attempts))
            sleep(sleep_time)

def write_to_disk(json_data: ContributorsJSON, out_file: str) -> None:
    with open(out_file, 'w') as f:
        json.dump(json_data, f, indent=2, sort_keys=True)
        f.write("\n")

def update_contributor_data_file() -> None:
    """
    Get contributors data from Github and insert them into a temporary
    dictionary. Retry fetching each repository if responded with non HTTP 200
    status.
    """

    # This dictionary should hold all repositories that should be included in
    # the total count, including those that should *not* have tabs on the team
    # page (e.g. if they are deprecated).
    repositories = {
        'server': 'https://api.github.com/repos/zulip/zulip/contributors',
        'desktop': 'https://api.github.com/repos/zulip/zulip-desktop/contributors',
        'mobile': 'https://api.github.com/repos/zulip/zulip-mobile/contributors',
        'python-zulip-api': 'https://api.github.com/repos/zulip/python-zulip-api/contributors',
        'zulip-js': 'https://api.github.com/repos/zulip/zulip-js/contributors',
        'zulipbot': 'https://api.github.com/repos/zulip/zulipbot/contributors',
        'terminal': 'https://api.github.com/repos/zulip/zulip-terminal/contributors',
        'zulip-ios-legacy': 'https://api.github.com/repos/zulip/zulip-ios-legacy/contributors',
        'zulip-android': 'https://api.github.com/repos/zulip/zulip-android/contributors',
    }

    data: ContributorsJSON = dict(date=str(date.today()), contrib=[])
    contribs_list: Dict[str, Dict[str, Union[str, int]]] = {}

    for name, link in repositories.items():
        contributors = fetch_contributors(link, args.max_retries)
        for contributor in contributors:
            username = contributor['login']
            if username in contribs_list:
                contribs_list[username][name] = contributor['contributions']
            else:
                contribs_list[username] = {
                    'avatar': contributor['avatar_url'],
                    'name': username,
                    name: contributor['contributions']
                }

    # remove duplicate contributions count
    # find commits at the time of split and subtract from zulip-server
    with open(duplicate_commits_file) as f:
        duplicate_commits = json.load(f)
        for committer in duplicate_commits:
            if committer in contribs_list and contribs_list[committer].get('server'):
                total_commits = contribs_list[committer]['server']
                assert isinstance(total_commits, int)
                duplicate_commits_count = duplicate_commits[committer]
                original_commits = total_commits - duplicate_commits_count
                contribs_list[committer]['server'] = original_commits

    data['contrib'] = list(contribs_list.values())
    write_to_disk(data, settings.CONTRIBUTOR_DATA_FILE_PATH)

if __name__ == "__main__":
    update_contributor_data_file()