github.py

import base64
import logging
from typing import Optional

import httpx
from fastapi import HTTPException

from config import GITHUB_API, GITHUB_TOKEN

logger = logging.getLogger(__name__)


def parse_github_url(url: str) -> tuple[str, str]:
    """Extract (owner, repo) from a GitHub URL."""
    import re
    m = re.match(r"https?://github\.com/([^/]+)/([^/]+)", url)
    assert m is not None  # guaranteed — URL was already validated by Pydantic
    owner, repo = m.group(1), m.group(2)
    return owner, repo.removesuffix(".git")


def github_headers() -> dict:
    """Auth header if a token is set — raises rate limit from 60 to 5000 req/hr."""
    h = {"Accept": "application/vnd.github+json", "X-GitHub-Api-Version": "2022-11-28"}
    if GITHUB_TOKEN:
        h["Authorization"] = f"Bearer {GITHUB_TOKEN}"
    return h


async def get_repo_tree(client: httpx.AsyncClient, owner: str, repo: str) -> list[dict]:
    """
    Fetch the complete recursive file tree from GitHub.
    Returns a flat list of blob objects: [{path, size, sha, ...}, ...]
    One API call — GitHub handles the recursion.
    """
    url = f"{GITHUB_API}/repos/{owner}/{repo}/git/trees/HEAD"
    resp = await client.get(url, headers=github_headers(), params={"recursive": "1"})

    if resp.status_code == 404:
        raise HTTPException(404, {"status": "error", "message": "Repository not found or is private."})
    if resp.status_code == 403:
        raise HTTPException(429, {"status": "error", "message": "GitHub rate limit exceeded. Set GITHUB_TOKEN to increase it."})
    if resp.status_code != 200:
        raise HTTPException(502, {"status": "error", "message": f"GitHub API error: {resp.status_code}"})

    data = resp.json()
    if data.get("truncated"):
        logger.warning(f"GitHub tree truncated for {owner}/{repo} — repo is very large.")

    return [item for item in data.get("tree", []) if item["type"] == "blob"]


async def fetch_file(client: httpx.AsyncClient, owner: str, repo: str, path: str) -> Optional[str]:
    """
    Fetch a single file's content. GitHub returns it base64-encoded.
    Returns decoded text, or None if the file is binary or unreachable.
    """
    url = f"{GITHUB_API}/repos/{owner}/{repo}/contents/{path}"
    try:
        resp = await client.get(url, headers=github_headers())
        if resp.status_code != 200:
            return None

        data = resp.json()
        if data.get("encoding") != "base64" or not data.get("content"):
            return None

        raw = base64.b64decode(data["content"])

        # Text files never contain null bytes — fast binary detection
        if b"\x00" in raw[:512]:
            return None

        return raw.decode("utf-8", errors="replace")
    except Exception as e:
        logger.debug(f"Could not fetch {path}: {e}")
        return None