scripts/experiments/run_analyzer_cpu_rebaseline.py

"""Driver of the CPU rebaseline run (25/08/2026) -- called by
``run_analyzer_cpu_rebaseline.bat``, which owns the environment.

It drives the SHIPPED pipeline, not a copy of it: the app is started in
"local" mode (``POOLCOACH_LOCAL_CV=1``), so the very functions of
``scripts/cv_worker.py`` run on a thread of this process, reached through
the public routes (``POST /api/analyzer/videos`` -> ``GET
/api/analyzer/shots``). No probe re-implements segmentation or analysis, so
the numbers cannot drift away from what the Space serves.

Waits on the RIGHT condition, never on a sleep: first the CV worker
heartbeat (the upload route fast-fails without it), then video status
``done`` AND every shot settled (``done`` or ``error``) -- a manifest on
disk appears long before the last shot is analyzed.
"""

from __future__ import annotations

import json
import os
import sys
import time
from pathlib import Path

APP_ROOT = Path(os.environ.get("POOLCOACH_APP_ROOT",
                               r"D:\poolcoach\poolcoach-app"))
VOD = Path(os.environ["POOLCOACH_REBASELINE_VOD"])

# The 4 corner pockets of run A2b (14/08) and of run C (24/08), in the order
# the front-end asks for them. Same VOD + same corners = the only two inputs
# the pipeline takes.
CORNERS = "[[527.5,869.7],[1380.1,864.0],[1232.3,278.5],[673.0,272.8]]"

WORKER_BOOT_TIMEOUT_S = 900.0    # YOLO + ShotNet + the pooltool Numba JIT
ANALYSIS_TIMEOUT_S = 4 * 3600.0  # a 5.3 minute rack on 2 CPU cores
POLL_S = 10.0

for _p in (APP_ROOT / "src", APP_ROOT):
    if str(_p) not in sys.path:
        sys.path.insert(0, str(_p))


def main() -> int:
    from fastapi.testclient import TestClient

    from app.main import app

    out_dir = Path(os.environ["POOLCOACH_ANALYZER_DIR"])
    print(f"[rebaseline] app root  : {APP_ROOT}")
    print(f"[rebaseline] vod       : {VOD} "
          f"({VOD.stat().st_size / 1e6:.1f} MB)")
    print(f"[rebaseline] result dir: {out_dir}", flush=True)

    t_boot = time.time()
    with TestClient(app) as client:
        while True:
            health = client.get("/api/health").json()
            if health.get("cv_worker_alive"):
                break
            if time.time() - t_boot > WORKER_BOOT_TIMEOUT_S:
                print(f"[rebaseline] FAILED: no CV worker heartbeat after "
                      f"{WORKER_BOOT_TIMEOUT_S:.0f}s -- health={health}")
                return 2
            time.sleep(2.0)
        print(f"[rebaseline] CV worker alive after "
              f"{time.time() - t_boot:.1f}s (mode={health.get('mode')})",
              flush=True)

        with open(VOD, "rb") as fh:
            resp = client.post(
                "/api/analyzer/videos",
                files={"video": (VOD.name, fh, "video/mp4")},
                data={"corners": CORNERS})
        if resp.status_code != 200:
            print(f"[rebaseline] FAILED: upload returned "
                  f"{resp.status_code}: {resp.text[:400]}")
            return 3
        video_id = resp.json()["id"]
        print(f"[rebaseline] video_id  : {video_id}", flush=True)

        t_run = time.time()
        last = ""
        while True:
            body = client.get("/api/analyzer/shots",
                              params={"video": video_id}).json()
            video = body.get("video", {})
            shots = body.get("shots", [])
            settled = [s for s in shots
                       if s.get("status") in ("done", "error")]
            line = (f"{video.get('status')} "
                    f"{len(settled)}/{len(shots)} settled "
                    f"(done={body.get('n_done')})")
            if line != last:
                print(f"[rebaseline] +{time.time() - t_run:7.1f}s  {line}",
                      flush=True)
                last = line
            if video.get("status") == "error":
                print(f"[rebaseline] FAILED: video status error: "
                      f"{video.get('message')}")
                return 4
            if (video.get("status") == "done" and shots
                    and len(settled) == len(shots)):
                break
            if time.time() - t_run > ANALYSIS_TIMEOUT_S:
                print(f"[rebaseline] FAILED: still not finished after "
                      f"{ANALYSIS_TIMEOUT_S:.0f}s")
                return 5
            time.sleep(POLL_S)

        elapsed = time.time() - t_run
        print(f"[rebaseline] DONE in {elapsed:.1f}s -- "
              f"{len(shots)} shots, {body.get('n_done')} analyzed",
              flush=True)
        summary = {"video_id": video_id,
                   "vod": str(VOD),
                   "corners": json.loads(CORNERS),
                   "elapsed_s": round(elapsed, 1),
                   "n_shots": len(shots),
                   "n_done": body.get("n_done"),
                   "result_dir": str(out_dir / video_id)}
        (out_dir / "rebaseline_run.json").write_text(
            json.dumps(summary, indent=1), encoding="utf-8")
    return 0


if __name__ == "__main__":
    raise SystemExit(main())