backup github repos to gitea

gh_mirrors_to_gitea.py

# requirements: httpx, python-dotenv
import os

import httpx
from dotenv import load_dotenv
from datetime import datetime as dt
import pytz
from dateutil import parser

load_dotenv()

gitea_api_key = os.getenv("GITEA_API_KEY")
gitea_server_url = os.getenv("GITEA_SERVER_URL")
github_token = os.getenv("GITHUB_TOKEN")
github_username = "tobi-de"
# repos_api_path = f"https://api.github.com/users/{github_username}/repos?type=owner&visibility=all"
repos_api_path = f"https://api.github.com/user/repos?type=owner"


def get_next_link(link):
    links = link.split(',')
    for l in links:
        if 'rel="next"' in l:
            return l.split(";")[0].replace("<", "").replace(">", "").strip()


def retrieve_all_gh_repos():
    repos = []
    with httpx.Client(headers={"Authorization": f"token {github_token}"}) as client:
        url = repos_api_path
        while True:
            resp = client.get(url)
            data = resp.json()
            repos.extend({"clone_addr": d["clone_url"], "description": d["description"], "private": d["private"],
                          "repo_name": d["name"], "pushed_at": d["pushed_at"]} for d in data)
            link = resp.headers.get("link")
            has_next = link and 'rel="next"' in link
            if not has_next:
                break
            url = get_next_link(link)
    return repos


def main():
    repos = retrieve_all_gh_repos()
    print(len(repos))
    current_datetime_utc = dt.now(pytz.timezone('UTC'))
    with httpx.Client(
            base_url=f"{gitea_server_url}/api/v1",
            headers={"Authorization": f"token {gitea_api_key}"},
    ) as client:
        for repo in repos:
            print(repo["repo_name"])
            last_pushed = parser.parse(repo["pushed_at"])
            mirror_interval = "24h" if (current_datetime_utc - last_pushed).days < 15 else f"{7 * 24}h"
            resp = client.post("repos/migrate", json={
                "auth_password": github_token,
                "auth_username": github_username,
                "clone_addr": repo["clone_addr"],
                "description": repo["description"],
                "private": repo["private"],
                "repo_name": repo["repo_name"],
                "service": "github",
                "issues": True,
                "labels": True,
                "mirror": True,
                "releases": True,
                "wiki": True,
                "pull_requests": True,
                "mirror_interval": mirror_interval
            })
            print(resp)
    print(len(repos))


if __name__ == "__main__":
    main()

# /// script
# requires-python = ">=3.11"
# dependencies = [
#   "httpx",
#   "python-dotenv",
# ]
# ///
import os

import httpx
from dotenv import load_dotenv
from datetime import datetime as dt

load_dotenv()

github_token = os.getenv("GITHUB_TOKEN")
github_username = "tobi-de"
# repos_api_path = f"https://api.github.com/users/{github_username}/repos?type=owner&visibility=all"
repos_api_path = f"https://api.github.com/user/repos?type=owner"


def get_next_link(link):
    links = link.split(',')
    for l in links:
        if 'rel="next"' in l:
            return l.split(";")[0].replace("<", "").replace(">", "").strip()


def retrieve_all_gh_repos():
    repos = []
    with httpx.Client(headers={"Authorization": f"token {github_token}"}) as client:
        url = repos_api_path
        while True:
            resp = client.get(url)
            data = resp.json()
            repos.extend({"clone_addr": d["clone_url"], "description": d["description"], "private": d["private"],
                          "repo_name": d["name"], "pushed_at": d["pushed_at"]} for d in data)
            link = resp.headers.get("link")
            has_next = link and 'rel="next"' in link
            if not has_next:
                break
            url = get_next_link(link)
    return repos

if __name__ == "__main__":
    import subprocess

    repos = retrieve_all_gh_repos()
    for repo in repos:
        print("Pulling ", repo["repo_name"])
        result = subprocess.run(["gh", "repo", "clone", repo["repo_name"]], check=False, stdout=subprocess.PIPE, stderr=subprocess.PIPE)
        if result.returncode != 0:
            print("Error : ", result.stderr.decode("utf-8"))
            continue

from pathlib import Path
import subprocess

def check_repo_status(repo_path):
    # Check if it's a git repository
    if not (repo_path / ".git").is_dir():
        return False
    
    # Check for uncommitted changes
    uncommitted_changes = subprocess.run(
        ['git', 'status', '--porcelain'],
        stdout=subprocess.PIPE,
        stderr=subprocess.PIPE,
        text=True,
        cwd=repo_path
    ).stdout.strip()
    
    # Fetch updates from the remote
    subprocess.run(['git', 'remote', 'update'], cwd=repo_path, stdout=subprocess.PIPE, stderr=subprocess.PIPE, text=True)
    
    # Check if branch is behind or diverged from the remote
    branch_status = subprocess.run(
        ['git', 'status', '-uno'],
        stdout=subprocess.PIPE,
        stderr=subprocess.PIPE,
        text=True,
        cwd=repo_path
    ).stdout
    
    not_up_to_date = "Your branch is behind" in branch_status or "have diverged" in branch_status

    return bool(uncommitted_changes) or not_up_to_date

def check_repos_in_directory(directory):
    # Loop through each folder in the directory
    for repo_path in directory.iterdir():
        if repo_path.is_dir() and check_repo_status(repo_path):
            print(f"Repository {repo_path.name} has uncommitted changes or is not up to date.")

# Set your directory path here
directory_path = Path("/path/to/your/repos")
check_repos_in_directory(directory_path)