from collections import Counter
import datetime
import os
import pprint

import requests

github_token = os.getenv('GITHUB_TOKEN')
headers=dict(Authorization='token {}'.format(github_token))
all_pr_url = 'https://api.github.com/repos/theorchard/{repo}/pulls?sort=updated&direction=desc&state=closed&per_page=100'
review_state = ('APPROVED', 'DISMISSED', 'CHANGES_REQUESTED')


def count_output(count, header=None):
	sorted_count = count.most_common()
	if header:
		print(header)
		print('-------------')
	for item, num in sorted_count:
		print('{},{}'.format(item, num))
	print("\n")


repos = [
	'graphql-product',  # code owners added 03/06/2020
	'graphql-knowledge',  # code owners added 11/12/2020 (only /src/schema/ reviews)
	'graphql-participant',  # code owners added 11/03/2020 (only /src/schema/ reviews)
	'graphql-podcast',  # code owners added 11/20/2020 (only schema reviews)
	'graphql-user',   # code owners added 12/01/2020 (only schema reviews)
	'graphql-abacus',   # Missing
	'graphql-publishing',  # code owners added 11/09/2020 (only schema reviews)
	'graphql-sr-delivery']   # need to add


repos = ['terraform-infra'] 

# repos = ['graphql-sr-delivery']

pr_to_investigate = list()

all_authors = list()
all_pr_states = list()
all_approved_reviewers = list()
all_commented_reviewers = list()
all_reviewers = list()
min_merge = None
max_merge = None

for repo in repos:
	print('Analyzing {}'.format(repo))
	url = all_pr_url.format(repo=repo)
	response = requests.get(url, headers=headers)

	prs = response.json()
	for pr in prs:
		if pr['merged_at']:
			merged_at_date = datetime.datetime.strptime(pr['merged_at'], "%Y-%m-%dT%H:%M:%SZ")
			if min_merge is None or min_merge > merged_at_date:
				min_merge = merged_at_date

			if max_merge is None or max_merge < merged_at_date:
				max_merge =merged_at_date

			all_authors.append(pr['user']['login'])
			pr_reviews_response = requests.get(pr['url'] + '/reviews', headers=headers)
			pr_reviews = pr_reviews_response.json()

			reviewers = list()
			approved_reviewers = list()
			commented_reviewers = list()
			for pr_review in pr_reviews:
				all_pr_states.append(pr_review['state'])
				if pr_review['state'] == 'APPROVED':
					approved_reviewers.append(pr_review['user']['login'])
					all_approved_reviewers.append(pr_review['user']['login'])
				if pr_review['state'] == 'COMMENTED':
					commented_reviewers.append(pr_review['user']['login'])
					all_commented_reviewers.append(pr_review['user']['login'])
				if pr_review['state'] in review_state:
					reviewers.append(pr_review['user']['login'])
					all_reviewers.append(pr_review['user']['login'])


			pr_to_investigate.append(
				dict(
					pr_no=pr['number'],
					url=pr['url'],
					author=pr['user']['login'],
					state=pr['state'],
					created_at=pr['created_at'],
					merged_at=pr['merged_at'],
					approved_reviewers=list(dict.fromkeys(approved_reviewers)),
					commented_reviewers=list(dict.fromkeys(commented_reviewers))))


authors_counts = Counter(all_authors)
state_counts = Counter(all_pr_states)
approved_counts = Counter(all_approved_reviewers)
comment_counts = Counter(all_commented_reviewers)
reviewers_counts = Counter(all_reviewers)

print('Total {} PRs: {} ({}-{})\n'.format(repo, len(pr_to_investigate), min_merge.strftime("%Y-%m-%d"), max_merge.strftime("%Y-%m-%d")))
count_output(authors_counts, 'Authors')
count_output(state_counts, 'Review States')
count_output(reviewers_counts, 'PR Reviews')
count_output(comment_counts, 'PR Comments')