EvoScientist/scripts/release.py

"""Local release script: unified version bump + build + Gitea release.

Usage:
    python scripts/release.py patch [--webui-repo PATH] [--dry-run] [--resume] [--skip-tests] [--breaking-db]
"""

from __future__ import annotations

import argparse
import hashlib
import json
import os
import re
import subprocess
import sys
import urllib.request
from pathlib import Path

STEPS = ["precheck", "test", "bump", "build", "checksums", "tag", "release", "verify"]
STATE_FILE = ".release-state.json"


def compute_next_version(current: str, bump: str) -> str:
    if re.fullmatch(r"\d+\.\d+\.\d+", bump):
        return bump
    parts = [int(p) for p in current.split(".")][:3]
    while len(parts) < 3:
        parts.append(0)
    if bump == "patch":
        parts[2] += 1
    elif bump == "minor":
        parts[1] += 1
        parts[2] = 0
    elif bump == "major":
        parts[0] += 1
        parts[1] = 0
        parts[2] = 0
    else:
        raise ValueError(f"unknown bump: {bump}")
    return ".".join(str(p) for p in parts)


def extract_changelog(text: str, version: str) -> str | None:
    pattern = re.compile(
        rf"^## \[?{re.escape(version)}\]?[^\n]*\n(.*?)(?=^## |\Z)",
        re.MULTILINE | re.DOTALL,
    )
    match = pattern.search(text)
    if not match:
        return None
    body = match.group(1).strip()
    return body or None


def write_checksums(artifacts: list[Path], dest: Path) -> None:
    lines = []
    for artifact in artifacts:
        digest = hashlib.sha256(artifact.read_bytes()).hexdigest()
        lines.append(f"{digest}  {artifact.name}")
    dest.write_text("\n".join(lines) + "\n", encoding="utf-8")


class ReleaseState:
    def __init__(self, path: Path):
        self.path = path
        self.done: list[str] = []
        if path.exists():
            try:
                self.done = json.loads(path.read_text(encoding="utf-8")).get("done", [])
            except (json.JSONDecodeError, OSError):
                self.done = []

    def is_done(self, step: str) -> bool:
        return step in self.done

    def mark_done(self, step: str) -> None:
        if step not in self.done:
            self.done.append(step)
        self.path.write_text(json.dumps({"done": self.done}, indent=2), encoding="utf-8")


def _current_backend_version(pyproject: Path) -> str:
    match = re.search(r'^version = "([^"]+)"', pyproject.read_text(encoding="utf-8"), re.MULTILINE)
    if not match:
        raise RuntimeError(f"version not found in {pyproject}")
    return match.group(1)


def _gitea_api(base: str, token: str, method: str, path: str, payload: dict | None = None) -> dict:
    body = json.dumps(payload).encode() if payload is not None else None
    req = urllib.request.Request(
        f"{base}/api/v1{path}",
        data=body,
        method=method,
        headers={
            "Authorization": f"token {token}",
            "Content-Type": "application/json",
            "User-Agent": "EvoScientist release-script",
        },
    )
    with urllib.request.urlopen(req, timeout=60) as resp:
        return json.loads(resp.read())


def _upload_asset(base: str, token: str, repo: str, release_id: int, path: Path) -> None:
    boundary = "----evosci-release"
    data = path.read_bytes()
    body = (
        f"--{boundary}\r\n"
        f'Content-Disposition: form-data; name="attachment"; filename="{path.name}"\r\n'
        f"Content-Type: application/octet-stream\r\n\r\n"
    ).encode() + data + f"\r\n--{boundary}--\r\n".encode()
    req = urllib.request.Request(
        f"{base}/api/v1/repos/{repo}/releases/{release_id}/assets?name={path.name}",
        data=body,
        method="POST",
        headers={
            "Authorization": f"token {token}",
            "Content-Type": f"multipart/form-data; boundary={boundary}",
        },
    )
    with urllib.request.urlopen(req, timeout=300):
        pass


def main(argv=None, *, run=subprocess.run, repo_root: Path | None = None) -> int:
    parser = argparse.ArgumentParser(prog="release.py")
    parser.add_argument("bump")
    parser.add_argument("--webui-repo", default="../EvoScientist-WebUI")
    parser.add_argument("--dry-run", action="store_true")
    parser.add_argument("--resume", action="store_true")
    parser.add_argument("--skip-tests", action="store_true")
    parser.add_argument("--breaking-db", action="store_true")
    args = parser.parse_args(argv)

    backend = (repo_root or Path(__file__).resolve().parent.parent).resolve()
    webui = (
        Path(args.webui_repo)
        if Path(args.webui_repo).is_absolute()
        else (backend / args.webui_repo).resolve()
    )
    pyproject = backend / "pyproject.toml"
    current = _current_backend_version(pyproject)
    target = compute_next_version(current, args.bump)

    base = os.environ.get("EVOSCIENTIST_UPDATE_BASE_URL", "https://git.foksai.com").rstrip("/")
    repo = os.environ.get("EVOSCIENTIST_UPDATE_REPO", "ouyangbo/EvoScientist")
    token = os.environ.get("EVOSCIENTIST_RELEASE_TOKEN", "")
    if not args.dry_run and not token:
        print("EVOSCIENTIST_RELEASE_TOKEN is required (or use --dry-run)", file=sys.stderr)
        return 2

    state_path = backend / STATE_FILE
    if args.resume:
        state = ReleaseState(state_path)
    else:
        state_path.unlink(missing_ok=True)
        state = ReleaseState(Path(os.devnull)) if args.dry_run else ReleaseState(state_path)

    print(f"release plan: {current} -> {target} (backend={backend}, webui={webui})")

    def step(name, fn):
        if state.is_done(name):
            print(f"[skip] {name}")
            return None
        print(f"[run ] {name}")
        result = fn()
        if not args.dry_run:
            state.mark_done(name)
        return result

    def precheck():
        for r in (backend, webui):
            out = run(["git", "-C", str(r), "status", "--porcelain"], capture_output=True, text=True)
            if out.stdout.strip():
                raise RuntimeError(f"{r} has uncommitted changes")
            branch = run(
                ["git", "-C", str(r), "branch", "--show-current"],
                capture_output=True,
                text=True,
            )
            if branch.stdout.strip() != "main":
                raise RuntimeError(f"{r} is not on main")

    def tests():
        if args.skip_tests:
            return
        run([sys.executable, "-m", "pytest", "-x", "-q"], cwd=backend).check_returncode()
        run(["npx", "vitest", "run"], cwd=webui).check_returncode()

    def bump():
        if args.dry_run:
            print(f"  would set version {target} in pyproject.toml and webui package.json")
            return
        text = pyproject.read_text(encoding="utf-8")
        pyproject.write_text(
            text.replace(f'version = "{current}"', f'version = "{target}"', 1),
            encoding="utf-8",
        )
        run(["npm", "version", target, "--no-git-tag-version"], cwd=webui).check_returncode()

    dist = backend / "dist" / f"release-{target}"

    def build():
        if args.dry_run:
            print("  would run: uv build; npm run build && npm pack")
            return
        dist.mkdir(parents=True, exist_ok=True)
        run(["uv", "build", "--out-dir", str(dist)], cwd=backend).check_returncode()
        run(["npm", "run", "build"], cwd=webui).check_returncode()
        run(["npm", "pack", "--pack-destination", str(dist)], cwd=webui).check_returncode()

    def checksums():
        if args.dry_run:
            return
        artifacts = [
            p
            for p in dist.iterdir()
            if p.suffix in {".whl", ".tgz", ".gz"} and p.name != "checksums.txt"
        ]
        write_checksums(artifacts, dist / "checksums.txt")

    def tag():
        msg = f"chore: release v{target}"
        for r in (backend, webui):
            if args.dry_run:
                print(f"  would commit+tag v{target} in {r}")
                continue
            run(["git", "-C", str(r), "add", "-A"]).check_returncode()
            run(["git", "-C", str(r), "commit", "-m", msg]).check_returncode()
            run(["git", "-C", str(r), "tag", f"v{target}"]).check_returncode()
            run(["git", "-C", str(r), "push", "origin", "main", f"v{target}"]).check_returncode()

    def create_release():
        if args.dry_run:
            print(f"  would create release v{target} on {base}/{repo}")
            return
        notes = []
        for changelog in (backend / "CHANGELOG.md", webui / "CHANGELOG.md"):
            if changelog.exists():
                section = extract_changelog(changelog.read_text(encoding="utf-8"), target)
                if section:
                    notes.append(f"### {changelog.parent.name}\n\n{section}")
        body = "\n\n".join(notes) or f"Release v{target}"
        if args.breaking_db:
            body = "BREAKING-DB\n\n" + body
        rel = _gitea_api(
            base,
            token,
            "POST",
            f"/repos/{repo}/releases",
            {
                "tag_name": f"v{target}",
                "name": f"v{target}",
                "body": body,
                "prerelease": bool(args.breaking_db),
            },
        )
        release_id = rel["id"]
        for artifact in dist.iterdir():
            if artifact.name == "checksums.txt" or artifact.suffix in {".whl", ".tgz", ".gz"}:
                _upload_asset(base, token, repo, release_id, artifact)

    def verify():
        if args.dry_run:
            return
        sums = (dist / "checksums.txt").read_text(encoding="utf-8")
        for line in sums.splitlines():
            digest, name = line.split()
            url = f"{base}/{repo}/releases/download/v{target}/{name}"
            req = urllib.request.Request(url, headers={"Authorization": f"token {token}"})
            with urllib.request.urlopen(req, timeout=120) as resp:
                actual = hashlib.sha256(resp.read()).hexdigest()
            if actual != digest:
                raise RuntimeError(f"uploaded asset checksum mismatch: {name}")
        print("release verified")

    step("precheck", precheck)
    step("test", tests)
    step("bump", bump)
    step("build", build)
    step("checksums", checksums)
    step("tag", tag)
    step("release", create_release)
    step("verify", verify)
    print(f"done: v{target}")
    return 0


if __name__ == "__main__":
    sys.exit(main())