#!/usr/bin/env python3
"""Read-only sweep of this worktree and focused TMPDIR, including deleted cwd."""
import datetime
import json
import os
from pathlib import Path
HERE=Path(__file__).resolve().parent
REPO=HERE.parents[3]
RUN=HERE/'run3-focused'
def main():
    tmp=Path((RUN/'TMPDIR').read_text().strip())
    started=datetime.datetime.fromisoformat((RUN/'start.txt').read_text().strip().replace('Z','+00:00')).timestamp()
    boot=int(next(s.split()[1] for s in Path('/proc/stat').read_text().splitlines() if s.startswith('btime ')))
    ancestors=set();pid=os.getpid()
    while pid and pid not in ancestors:
        ancestors.add(pid)
        try:pid=int(next(s.split(':')[1] for s in Path(f'/proc/{pid}/status').read_text().splitlines() if s.startswith('PPid:')))
        except (OSError,StopIteration):break
    known={2880281,2880282,2880283,2880284,2880445,2880487,2880537,2883251,2880439,2469277,2469278,2469279,2469292,2469478,2469485,2469518,2469601,2470486}
    remaining=[];excluded=[]
    for proc in Path('/proc').iterdir():
        if not proc.name.isdigit() or int(proc.name) in ancestors:continue
        try:
            rawcwd=os.readlink(proc/'cwd');cwd=Path(rawcwd.removesuffix(' (deleted)'))
            if not(cwd==REPO or REPO in cwd.parents or cwd==tmp or tmp in cwd.parents):continue
            stat=(proc/'stat').read_text().rsplit(') ',1)[1].split()
            stamp=boot+int(stat[19])/os.sysconf('SC_CLK_TCK')
            item=dict(pid=int(proc.name),comm=(proc/'comm').read_text().strip(),cwd=rawcwd,started_utc=datetime.datetime.fromtimestamp(stamp,datetime.timezone.utc).isoformat())
            if int(proc.name) in known and stamp<started:excluded.append(item)
            else:remaining.append(item)
        except (OSError,ValueError,IndexError):continue
    result=dict(timestamp=datetime.datetime.now(datetime.timezone.utc).isoformat(),count=len(remaining),remaining=remaining,excluded_preexisting_infrastructure=excluded,note='Read-only /proc cwd sweep, including deleted cwd under this run TMPDIR. Own ancestry excluded. Only known MCP/npm/code-mode infrastructure PIDs with pre-run start ticks are excluded. No signalling.')
    (RUN/'sweep.json').write_text(json.dumps(result,indent=2)+'\n');print(json.dumps(result,indent=2))
if __name__=='__main__':main()