Decoder2704/python-chatbot-jay

1#!/usr/bin/env python32"""3Run the data/artifact build pipeline in order (subprocess, sequential).4 5Steps:6  1. build_answers_v2.py7  2. build_final_dataset.py8  3. build_retrieval_artifacts.py9"""10 11from __future__ import annotations12 13import subprocess14import sys15from pathlib import Path16 17_BACKEND_DIR = Path(__file__).resolve().parent.parent18if str(_BACKEND_DIR) not in sys.path:19    sys.path.insert(0, str(_BACKEND_DIR))20 21from app.paths import ARTIFACTS_DIR, DATA_DIR, SCRIPTS_DIR22 23_STEPS: list[tuple[str, Path]] = [24    ("build_answers_v2.py", SCRIPTS_DIR / "build_answers_v2.py"),25    ("build_final_dataset.py", SCRIPTS_DIR / "build_final_dataset.py"),26    ("build_retrieval_artifacts.py", SCRIPTS_DIR / "build_retrieval_artifacts.py"),27]28 29 30def _list_files(root: Path, label: str) -> None:31    print(f"\n--- Files under {label} ({root}) ---")32    if not root.exists():33        print("  (directory does not exist)")34        return35    if not root.is_dir():36        print("  (not a directory)")37        return38    files = sorted(p for p in root.rglob("*") if p.is_file())39    if not files:40        print("  (no files)")41        return42    for f in files:43        print(f"  {f}")44 45 46def main() -> int:47    print(f"Backend directory: {_BACKEND_DIR}")48    print("Running build steps sequentially (subprocess)...\n")49 50    for i, (name, script_path) in enumerate(_STEPS, start=1):51        if not script_path.is_file():52            print(f"[{i}/3] FAIL: script not found: {script_path}", file=sys.stderr)53            return 154 55        print(f"[{i}/3] Running {name} ...")56        print(f"        command: {sys.executable} {script_path}")57        result = subprocess.run(58            [sys.executable, str(script_path)],59            cwd=str(_BACKEND_DIR),60        )61        if result.returncode != 0:62            print(63                f"[{i}/3] FAIL: {name} exited with code {result.returncode}",64                file=sys.stderr,65            )66            return result.returncode67 68        print(f"[{i}/3] OK: {name}\n")69 70    print("All build steps completed successfully.\n")71    _list_files(DATA_DIR, "backend/data/")72    _list_files(ARTIFACTS_DIR, "backend/artifacts/")73    return 074 75 76if __name__ == "__main__":77    raise SystemExit(main())78