services/hf_to_github.py

import sys
import os
import subprocess
import uuid
import requests
import json
from git import Repo
from huggingface_hub import snapshot_download

def get_github_user(token):
    try:
        response = requests.get(
            "https://api.github.com/user",
            headers={
                "Authorization": f"token {token}",
                "Accept": "application/vnd.github.v3+json"
            },
            timeout=10
        )
        if response.ok:
            return response.json().get("login")
    except Exception as e:
        print(f"Error fetching GitHub user: {e}")
    return None

def create_github_repo(token, name):
    try:
        # First check if it exists
        user = get_github_user(token)
        if not user:
            return False

        check_res = requests.get(
            f"https://api.github.com/repos/{user}/{name}",
            headers={
                "Authorization": f"token {token}",
                "Accept": "application/vnd.github.v3+json"
            },
            timeout=10
        )

        if check_res.status_code == 200:
            print(f"Repository {user}/{name} already exists.")
            return True

        # Create it
        print(f"Creating repository {user}/{name}...")
        response = requests.post(
            "https://api.github.com/user/repos",
            headers={
                "Authorization": f"token {token}",
                "Accept": "application/vnd.github.v3+json"
            },
            json={
                "name": name,
                "private": True,
                "description": f"Imported from Hugging Face Space"
            },
            timeout=10
        )
        return response.ok
    except Exception as e:
        print(f"Error creating GitHub repo: {e}")
        return False

def parse_hf_repo_id(hf_input):
    if not hf_input: return None, None
    if "huggingface.co/spaces/" in hf_input:
        parts = hf_input.split("huggingface.co/spaces/")[1].split("/")
        return parts[0], parts[1]
    if "/" in hf_input:
        parts = hf_input.split("/")
        return parts[0], parts[1]
    return None, None

def parse_github_repo_id(git_input):
    if not git_input: return None, None
    if "github.com/" in git_input:
        parts = git_input.split("github.com/")[1].split("/")
        return parts[0], parts[1].replace(".git", "")
    if "/" in git_input:
        parts = git_input.split("/")
        return parts[0], parts[1]
    return None, git_input

def get_hf_token():
    return os.environ.get("HF_TOKEN")

def get_github_token():
    return os.environ.get("GITHUB_TOKEN")

def handle_upload(hf_space_input, git_repo_input, branch):
    hf_profile, hf_space = parse_hf_repo_id(hf_space_input)
    if not hf_profile or not hf_space:
        return {"error": f"Invalid HF Space ID: '{hf_space_input}'"}

    hf_space_full = f"{hf_profile}/{hf_space}"

    git_token = get_github_token()
    if not git_token:
        return {"error": "GITHUB_TOKEN not found in environment secrets."}

    git_profile, git_repo = parse_github_repo_id(git_repo_input)

    # If no git_repo provided, use hf_space name
    if not git_repo:
        git_repo = hf_space

    # If no profile provided, fetch it from token
    if not git_profile:
        git_profile = get_github_user(git_token)
        if not git_profile:
            return {"error": "Could not determine GitHub user profile from token."}

    if not branch:
        branch = "main"

    hf_token = get_hf_token()
    if not hf_token:
        return {"error": "HF_TOKEN not found in environment secrets."}

    # Ensure Repo exists
    create_github_repo(git_token, git_repo)

    work_dir = "/tmp/jules_uploads"
    os.makedirs(work_dir, exist_ok=True)
    upload_dir = os.path.join(work_dir, f"upload_{uuid.uuid4().hex}")
    os.makedirs(upload_dir, exist_ok=True)

    try:
        # 1. Download HF Space
        print(f"Downloading {hf_space_full}...")
        snapshot_download(
            repo_id=hf_space_full,
            repo_type="space",
            local_dir=upload_dir,
            token=hf_token
        )

        # 2. Target URL
        target_url = f"https://{git_token}@github.com/{git_profile}/{git_repo}.git"

        # 3. Push to GitHub
        repo = Repo.init(upload_dir)

        # Setup branch
        try:
            repo.git.checkout("-b", branch)
        except:
            repo.git.checkout(branch)

        repo.git.add(A=True)
        if repo.git.status("--short"):
            repo.index.commit(f"Import from HF Space {hf_space_full}")

        if "origin" in [remote.name for remote in repo.remotes]:
            origin = repo.remote("origin")
            origin.set_url(target_url)
        else:
            origin = repo.create_remote("origin", target_url)

        print(f"Pushing to GitHub {git_profile}/{git_repo}...")
        repo.git.push("origin", branch, force=True)

        return {
            "status": "success",
            "message": f"{hf_space_full} uploaded to {git_profile}/{git_repo} ({branch})",
            "repo_id": f"{git_profile}/{git_repo}"
        }

    except Exception as e:
        return {"error": str(e)}

if __name__ == "__main__":
    if len(sys.argv) < 4:
        print(json.dumps({"error": "Missing arguments"}))
        sys.exit(1)

    hf_space = sys.argv[1]
    git_repo = sys.argv[2]
    branch = sys.argv[3]

    result = handle_upload(hf_space, git_repo, branch)
    print(json.dumps(result))