src/researchlink/services/github_client.py

"""GitHub repository inspection utilities."""

from __future__ import annotations

import re
from pathlib import Path

import httpx

GITHUB_API = "https://api.github.com"
HEADERS = {
    "Accept": "application/vnd.github.v3+json",
    "User-Agent": "ResearchLink-AI/0.1",
}


def parse_github_url(url: str) -> tuple[str | None, str | None]:
    """Extract (owner, repo) from a GitHub URL."""
    match = re.search(r"github\.com[/:]([^/]+)/([^/\s.]+?)(?:\.git)?(?:/|$)", url)
    if match:
        return match.group(1), match.group(2)
    return None, None


def fetch_repo_metadata(owner: str, repo: str) -> dict:
    """Fetch basic repo metadata via GitHub API (unauthenticated, rate-limited)."""
    url = f"{GITHUB_API}/repos/{owner}/{repo}"
    try:
        resp = httpx.get(url, headers=HEADERS, timeout=15.0)
        if resp.status_code == 200:
            return resp.json()
        return {"error": f"HTTP {resp.status_code}"}
    except Exception as e:
        return {"error": str(e)}


def fetch_readme(owner: str, repo: str) -> str | None:
    """Fetch the raw README content via GitHub API."""
    url = f"{GITHUB_API}/repos/{owner}/{repo}/readme"
    try:
        resp = httpx.get(url, headers={**HEADERS, "Accept": "application/vnd.github.raw"}, timeout=15.0)
        if resp.status_code == 200:
            return resp.text
        return None
    except Exception:
        return None


def fetch_tree(owner: str, repo: str) -> list[dict]:
    """Fetch the top-level tree (non-recursive) via GitHub API."""
    url = f"{GITHUB_API}/repos/{owner}/{repo}/git/trees/HEAD"
    try:
        resp = httpx.get(url, headers=HEADERS, timeout=15.0)
        if resp.status_code == 200:
            data = resp.json()
            return data.get("tree", [])
        return []
    except Exception:
        return []


def detect_language_and_frameworks(tree_items: list[dict], readme: str | None) -> dict:
    """Heuristically detect language and frameworks from repo tree."""
    files = [item["path"] for item in tree_items if item.get("type") == "blob"]
    folders = [item["path"] for item in tree_items if item.get("type") == "tree"]

    languages: list[str] = []
    frameworks: list[str] = []
    setup_commands: list[str] = []
    has_tests = False
    test_paths: list[str] = []

    ext_map = {".py": "Python", ".js": "JavaScript", ".ts": "TypeScript", ".go": "Go",
               ".rs": "Rust", ".cpp": "C++", ".c": "C", ".java": "Java", ".r": "R"}

    for f in files:
        ext = Path(f).suffix.lower()
        if ext in ext_map and ext_map[ext] not in languages:
            languages.append(ext_map[ext])

    if "requirements.txt" in files:
        setup_commands.append("pip install -r requirements.txt")
    if "setup.py" in files or "pyproject.toml" in files:
        setup_commands.append("pip install -e .")
    if "environment.yml" in files:
        setup_commands.append("conda env create -f environment.yml")
    if "Makefile" in files:
        setup_commands.append("make install")

    for f in files + folders:
        if "test" in f.lower() or "spec" in f.lower():
            has_tests = True
            test_paths.append(f)

    text_to_check = " ".join(files + folders + ([readme] if readme else []))
    fw_hints = {
        "PyTorch": ["torch", "pytorch"],
        "TensorFlow": ["tensorflow", "tf"],
        "JAX": ["jax", "flax"],
        "Gymnasium": ["gymnasium", "gym"],
        "RLlib": ["rllib"],
        "Stable-Baselines3": ["stable_baselines", "sb3"],
        "FastAPI": ["fastapi"],
        "Flask": ["flask"],
        "NumPy": ["numpy"],
    }
    for fw, hints in fw_hints.items():
        if any(h in text_to_check.lower() for h in hints):
            frameworks.append(fw)

    return {
        "languages": languages,
        "frameworks": frameworks,
        "setup_commands": setup_commands,
        "has_tests": has_tests,
        "test_paths": test_paths[:10],
    }