digital-psychology/tools/cece_daemon_deep.py

#!/usr/bin/env python3
"""Daemonize + run cece_capture_deep outside the parent process group.

Cursor/agent shells often SIGKILL the whole process group when a command ends;
nohup alone is not enough. Double-fork so the crawl survives.
"""
from __future__ import annotations

import atexit
import faulthandler
import os
import runpy
import sys
import time
from pathlib import Path

ROOT = Path("/Users/jack/Project/digital-psychology")
LOG = Path("/tmp/cece-capture-deep-full.log")
FAULT = Path("/tmp/cece-fault.log")
PIDFILE = Path("/tmp/cece-capture-deep.pid")


def log(msg: str) -> None:
    line = f"[{time.strftime('%F %T')}] {msg}\n"
    with LOG.open("a", encoding="utf-8") as f:
        f.write(line)


def daemonize() -> None:
    # first fork
    if os.fork() > 0:
        sys.exit(0)
    os.setsid()
    # second fork
    if os.fork() > 0:
        sys.exit(0)
    # detach stdio
    sys.stdin.close()
    sys.stdout.flush()
    sys.stderr.flush()
    out = open(LOG, "a", encoding="utf-8", buffering=1)
    sys.stdout = out  # type: ignore
    sys.stderr = out  # type: ignore


def main() -> None:
    os.chdir(ROOT)
    daemonize()
    PIDFILE.write_text(str(os.getpid()), encoding="utf-8")
    ff = FAULT.open("a", encoding="utf-8")
    faulthandler.enable(file=ff, all_threads=True)
    log(f"daemon start pid={os.getpid()} argv={sys.argv[1:]}")

    def _atexit() -> None:
        log(f"daemon atexit pid={os.getpid()}")

    atexit.register(_atexit)
    # remaining args after this script name
    args = sys.argv[1:]
    sys.argv = ["cece_capture_deep.py", *args]
    try:
        runpy.run_path(str(ROOT / "tools" / "cece_capture_deep.py"), run_name="__main__")
        log("daemon finished ok")
    except BaseException as ex:
        log(f"daemon error: {type(ex).__name__}: {ex}")
        import traceback

        traceback.print_exc()
        raise SystemExit(1)


if __name__ == "__main__":
    main()