Scripts/collect_github_repos_commits_per_ci_files.py

import requests
import time
import csv
import os
import sys
import pandas as pd

from github_tokens import init_rest_client_auth

tokens: list[str] = []
authorization_headers: list[dict[str, str]] = []
token_index = 0
last_request_time = {}  # Tracks last request time per token


def _ensure_auth() -> None:
    global tokens, authorization_headers
    if not tokens:
        tokens, authorization_headers = init_rest_client_auth()


def makeRequest(api_url, params):
    global token_index
    _ensure_auth()
    num_tries = 0
    max_tries = 3

    while True:
        num_tries += 1
        if num_tries > max_tries:
            print(' -- Max retries reached. Exiting...', end='')
            break
        
        # Find the first available token
        available_token = None
        current_time = time.time()
        for i in range(len(authorization_headers)):
            if i not in last_request_time or (current_time - last_request_time[i] >= 1):
                available_token = i
                break
        
        if available_token is None:
            min_wait = min(last_request_time.values()) + 1 - current_time
            time.sleep(max(min_wait, 1))
            continue
        
        token_index = available_token
        headers = authorization_headers[token_index]
        last_request_time[token_index] = time.time()
        
        while True:
            try:
                response = requests.get(api_url, headers=headers, params=params, timeout=30)
                break
            except requests.exceptions.RequestException as e:
                print(f' -- Request error: {e}. Retrying in 60 seconds...', end='')
                time.sleep(60)
            except Exception as e:
                print(f' -- Another request error: {e}. Retrying in 60 seconds...', end='')
                time.sleep(60)

        if response.status_code == 200:
            return response.json()
        elif response.status_code == 403 and ('message' in response.json() and 'rate limit' in response.json()['message'].lower()):
            last_request_time[token_index] = time.time()
        elif response.status_code == 422 or response.status_code == 401:
            return None
        else:
            print(f' -- Request failed with status code {response.status_code}. Retrying... ' + api_url, end='')
            time.sleep(1)
            
def search_repo_commits_by_ci_file_path(repo_name, ci_file_path, csv_file, csv_writer):
    '''
    Perform a GitHub Search API query for commits per repository based on the path to a CI configuration file.
    Write results directly to a file.
    '''
    GITHUB_API_URL = f'https://api.github.com/repos/{repo_name}/commits'
    page = 1
    per_page = 100
    num_commits = 0
    while True:
        params = {
            'path': ci_file_path,
            'per_page': per_page,
            'page': page,
        }

        data = makeRequest(GITHUB_API_URL, params)

        if not data or len(data) == 0:
            #print(f'No commits found.')
            return num_commits

        #print(f'  Page {page} for query: {GITHUB_API_URL}?path={ci_file_path}&per_page={per_page}&page={page}')

        for commit in data:
            num_commits += 1
            commit_info = {
                'repo_name': repo_name,
                'configuration_file': ci_file_path,
                'commit_sha': commit['sha'],
                'parent_sha': ', '.join([parent['sha'] for parent in commit['parents']]) if commit['parents'] else '',
                'commit_message': commit['commit']['message'].lstrip().rstrip().replace('\n', '\\n'),
                'author_login': commit['author']['login'] if commit['author'] else '',
                'author_name': commit['commit']['author']['name'],
                'author_email': commit['commit']['author']['email'],
                'author_date': commit['commit']['author']['date'],
                'committer_login': commit['committer']['login'] if commit['committer'] else '',
                'committer_name': commit['commit']['committer']['name'],
                'committer_email': commit['commit']['committer']['email'],
                'committer_date': commit['commit']['committer']['date'],
                'comment_count': commit['commit']['comment_count']
            }
            csv_writer.writerow(commit_info.values())
            csv_file.flush()
        
        if len(data) < per_page or page == 10:
            break

        page += 1

    if page == 10 and len(data) == 100:
        print('[This request has more than 1000 results]')
    
    return num_commits

if __name__ == '__main__':
    last_stopped_repo = 'immu0001/Udacity-Data-Engineer-nanodegree' # Last one for all services
    
    input_path = f'../Data/all_DE_repositories.csv'
    commits_output_path = f'../Data/DE_commits_per_ci_files.csv'
    repos_ci_output_path = f'../Data/DE_ci_services.csv'

    ci_file_paths = {
        # Keep in sync with ci-services.tex (ICSME data paper Table).
        '.github/workflows': 'GitHub Actions',
        '.travis.yml': 'Travis CI',
        'circle.yml': 'CircleCI',
        '.circleci/config.yml': 'CircleCI',
        '.gitlab-ci.yml': 'GitLab',
        '.appveyor.yml': 'AppVeyor',
        'appveyor.yml': 'AppVeyor',
        'azure-pipelines.yml': 'Azure Pipelines',
        'bitbucket-pipelines.yml': 'Bitbucket',
        '.cirrus.yml': 'Cirrus',
        '.scrutinizer.yml': 'Scrutinizer CI',
        'codeship-services.yml': 'Codeship',
        '.semaphore/semaphore.yml': 'Semaphore',
        'wercker.yml': 'Wercker',
        'Jenkinsfile': 'Jenkins',
        'bitrise.yml': 'Bitrise',
        'bamboo.yml': 'Bamboo',
        '.gocd.yaml': 'GoCD',
        'codemagic.yaml': 'Codemagic',
    }

    first_commits_time_csv = True if not os.path.exists(commits_output_path) else False
    commits_csv_file = open(commits_output_path, 'a', newline='', encoding='utf-8')
    commits_csv_writer = csv.writer(commits_csv_file, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)
    if first_commits_time_csv:
        commits_csv_writer.writerow(['repo_name', 'language', 'configuration_file', 'commit_sha', 'parent_sha', 'commit_message', 'author_login', 'author_name', 'author_email', 'author_date', 'committer_login', 'committer_name', 'committer_email', 'committer_date', 'comment_count'])
        commits_csv_file.flush()

    first_time_repos_ci_mapping_csv = True if not os.path.exists(repos_ci_output_path) else False
    repos_ci_csv_file = open(repos_ci_output_path, 'a', newline='', encoding='utf-8')
    repos_ci_csv_writer = csv.writer(repos_ci_csv_file, delimiter=',', quotechar='"', quoting=csv.QUOTE_MINIMAL)
    if first_time_repos_ci_mapping_csv:
        repos_ci_csv_writer.writerow(['repo_name', 'language', 'ci_services_count', 'ci_services_list'] + list(ci_file_paths.keys()))
        repos_ci_csv_file.flush()

    chunksize = 1000
    row_id = -1
    search_started = False
    print('Waiting to reach the last repo we stopped at last run...')
    
    for repos_df in pd.read_csv(input_path, chunksize=chunksize):
        for _, repo in repos_df.iterrows():
            row_id += 1

            if last_stopped_repo == '' or repo['full_name'] == last_stopped_repo: # Last repo we stopped at last run
                search_started = True
            
            if not search_started:
                continue

            print(f'{row_id+1}: {repo['full_name']}')
            ci_services_commit_counts = []
            ci_services_list = []
            for ci_file_path, ci_service_name in ci_file_paths.items():
                print(f'    -> {ci_file_path} ', end='')
                num_commits = search_repo_commits_by_ci_file_path(repo['full_name'], ci_file_path, csv_file=commits_csv_file, csv_writer=commits_csv_writer)
                ci_services_commit_counts.append(num_commits if num_commits > 0 else '')
                ci_services_list += [ci_service_name] if num_commits > 0 and ci_service_name not in ci_services_list else []
                print(f'[{num_commits}]' if num_commits > 0 else '')
            
            if len(ci_services_list) > 0:
                repos_ci_csv_writer.writerow([repo['full_name'], repo['language'], len(ci_services_list), ci_services_list] + ci_services_commit_counts)
                repos_ci_csv_file.flush()