pull-requests

#!/usr/bin/env python

import json
import urllib2
import sys
import traceback
from dateutil.parser import parse as parsedate
import re

GITHUB_API = 'https://api.github.com'

def openurl(url, argv):
    headers = {}
    if argv:
        headers['Authorization'] = 'token ' + argv[0]
    req = urllib2.Request(url, headers=headers)
    return urllib2.urlopen(req)

def main(argv):
    config = json.load(file('pull-requests.config'))

    # only examine pull requests that came after the class_start
    start = parsedate(config['class_start'])

    # index students by github id
    students = {}
    for student in config['students']:
        students[student['ghid']] = student

    print '| %6s | %-22s | %-22s | %7s | %-50s |' \
        % ('status', 'submitter', 'author', 'commits', 'link')
    print '| ------ | ---------------------- | ---------------------- ' + \
        '| ------- | -------------------------------------------------- |'

    # loop over each project
    for project in config['projects']:
        org = project['org']
        repo = project['repo']

        # walk through all the pull requests for the given project.
        # this may require multiple API calls to step through multiple
        # pages of pull requests, so using a list of URLs which may grow
        links = [ '%s/repos/%s/%s/pulls?state=all' % (GITHUB_API, org, repo) ]
        for url in links:
            resp = openurl(url, argv)
            done = False

            # loop over each pull request in the project
            for pr in json.load(resp):
                # stop the loop when we reach a PR that came before class_start
                if parsedate(pr['created_at']) < start:
                    done = True
                    break

                # ignore PRs submitted by non-students
                submitter = pr['user']['login']
                if submitter not in students:
                    continue

                # get the state as 'open', 'closed', or 'merged'
                state = pr['state']
                if pr.get('merged_at'):
                    state = 'merged'

                # count the commits for each student who contributed to the PR
                authors = {}
                for commit in json.load(openurl(pr['commits_url'], argv)):
                    author = commit['author'] and commit['author']['login'] \
                        or submitter
                    if author in students:
                        authors[author] = authors.get(author, 0) + 1

                # report
                for author, commits in authors.items():
                    print '| %6s | %-22s | %-22s | %7d | %-50s |' \
                            % (state, submitter, author, commits,
                               pr['html_url'])

            if done:
                break

            # if there is a 'next' page, add it to the queue
            link = resp.info().getheader('Link')
            if link:
                match = re.match(r'^.*<([^>]+)>; rel="next"', link)
                if match:
                    links.append(match.group(1))

    return 0

try:
    status = main(sys.argv[1:])
except:
    traceback.print_exc(file=sys.stdout)
    status = 1
exit(status)