2023-09-13 18:15:40 +03:00

244 lines
6.5 KiB
Python

#!/usr/bin/env python
"""Collect data on the most-starred repos using GitHub's GraphQL API."""
import json
import os
import requests
import time
from datetime import datetime, timedelta
from requests.adapters import HTTPAdapter
from urllib3.util.retry import Retry
from config import Parser
token = Parser.GithubToken
def query(payload, variables=None):
session = requests.Session()
session.headers.update({'Authorization': f'bearer {token}'})
retry = Retry(connect=3, backoff_factor=0.5)
adapter = HTTPAdapter(max_retries=retry)
session.mount('http://', adapter)
session.mount('https://', adapter)
r = session.post(
'https://api.github.com/graphql',
json={"query": payload, "variables": variables or {}}
)
r.raise_for_status()
return r.json()
repo_query = '''
query popular_repos($start: String, $num: Int!){
rateLimit {
cost
remaining
resetAt
}
search(query: "is:public %s", type: REPOSITORY, first: $num, after: $start) {
repositoryCount
pageInfo {
hasNextPage
endCursor
}
edges {
node {
... on Repository {
nameWithOwner
createdAt
forkCount
isFork
updatedAt
primaryLanguage {
name
}
stargazers {
totalCount
}
watchers {
totalCount
}
}
}
}
}
}
'''
count_query = '''
query {
rateLimit {
cost
remaining
resetAt
}
search(query: "is:public %s", type: REPOSITORY, first: 1) {
repositoryCount
}
}
'''
def get_repos(q, cursor, num):
return query(repo_query % q, {'start': cursor, 'num': num})['data']
def get_count(q):
return query(count_query % q)['data']['search']['repositoryCount']
def scrape(q, out_file):
#path = f'responses/{out_file}'
#if os.path.exists(path):
# print('Skipping', path, 'already exists')
# return
all_repos = []
cursor = None
#print('Creating', path)
while True:
r = get_repos(q, cursor, 100)
search = r['search']
pi = search['pageInfo']
cursor = pi['endCursor']
has_next = pi['hasNextPage']
total = search['repositoryCount']
#if total > 2000:
# raise ValueError(f'Too many results for {q}: {total}')
all_repos += [e['node'] for e in search['edges']]
print(r['rateLimit'])
print(len(all_repos), ' / ', total, cursor)
if not has_next or r['rateLimit']['remaining'] < 10:
break
#print(all_repos)
from pymongo import MongoClient
client = MongoClient("mongodb://admin:admin@localhost:27017")
db = client['git']
collection = db['repos']
for repo in all_repos:
entity = {}
filter_dict = {}
for key, value in repo.items():
if key == "nameWithOwner":
filter_dict[key] = value
else:
entity[key] = value
collection.update_one(filter_dict, {"$set": entity}, upsert=True)
#with open(path, 'w') as out:
#print(out)
#json.dump(all_repos, out)
time.sleep(4)
def scrape_star_range(low, high):
"""Scrape a simple star range [low, high]."""
out_file = f'repos.stars={low}..{high}.json'
q = 'stars:%d..%d' % (low, high)
scrape(q, out_file)
def scrape_breaks():
breaks = json.load(open('breaks.json'))
for hi, lo in zip(breaks[:-1], breaks[1:]):
scrape_star_range(lo, hi - 1)
def scrape_star_dates():
for stars in range(123, 15, -1):
out_file = f'repos.star={stars}.-2015.json'
q = 'stars:%d created:<=2015' % stars
scrape(q, out_file)
out_file = f'repos.star={stars}.2016-.json'
q = 'stars:%d created:>=2016' % stars
scrape(q, out_file)
def query_for_star_years(stars, start, end):
q = 'stars:%s' % stars
if start == 2010 and end == 2023:
return q
elif start == 2010:
return f'{q} created:<={end}'
elif end == 2023:
return f'{q} created:>={start}'
else:
return f'{q} created:{start}..{end}'
def split_interval(a, b):
d = int((b - a) / 2)
return [(a, a + d), (a + d + 1, b)]
def split_by_year(stars, start, end):
if start == 2010 and end == 2023:
c = 1001 # we know this will fail.
elif start == end:
split_by_days(
stars,
datetime(start, 1, 1),
datetime(start, 12, 31)
)
return
else:
q = query_for_star_years(stars, start, end)
c = get_count(q)
if c <= 1000:
out_file = f'repos.star={stars}.{start}-{end}.json'
print(f'query: {q}')
scrape(q, out_file)
else:
if start == end:
raise ValueError(f'Can\'t split any more for {stars} / {start}')
print(f'{stars} {start}..{end} -> {c}, will split')
for a, b in split_interval(start, end):
split_by_year(stars, a, b)
def split_by_days(stars, day_start, day_end):
start_fmt = day_start.strftime('%Y-%m-%d')
end_fmt = day_end.strftime('%Y-%m-%d')
q = query_for_star_years(stars, start_fmt, end_fmt)
c = get_count(q)
if c <= 1000:
out_file = f'repos.star={stars}.{start_fmt}-{end_fmt}.json'
print(f'query: {q}')
scrape(q, out_file)
else:
days = (day_end - day_start).days
if days == 0:
raise ValueError(f'Can\'t split any more: {stars} / {day_start} .. {day_end}')
for a, b in split_interval(0, days):
dt_a = day_start + timedelta(days=a)
dt_b = day_start + timedelta(days=b)
split_by_days(stars, dt_a, dt_b)
def scrape_star_dates_split():
#for stars in range(83, 15, -1):
for stars in range(40, 15, -1):
split_by_year(stars, 2010, 2023)
def scrape_range_days():
# Scrape from a low star range up, splitting by creation date (which never changes).
# ranges = [(15, 20), (21, 25), (26, 30), (31, 35), (36, 40), (41, 45), (46, 49)]
#ranges = [(50, 60), (61, 70), (71, 80), (81, 90), (91, 100)]
#ranges = ranges + [(100, 119), (120, 139), (140, 159), (160, 179), (180, 200)]
#ranges = ranges + [(201, 225), (226, 250), (251, 300), (301, 400), (401, 500)]
#ranges = ranges + [(501, 700), (701, 1000), (1001, 1500), (1501, 5000), (5001, 1_000_000)]
ranges = [(1001, 1500), (1501, 5000), (5001, 1_000_000)]
for a, b in ranges:
stars = f'{a}..{b}'
split_by_days(stars, datetime(2007, 1, 1), datetime(2024, 2, 2))
#if __name__ == '__main__':
# scrape_breaks()
# scrape_star_dates()
# scrape_star_dates_split()
#scrape_range_days()