#!/usr/bin/env python3
"""PUBLIC self-check (arm A4 and controller): verifies deliverable presence and well-formedness only. It knows nothing about the
correct answers. Usage: python tools/selfcheck.py [--workspace .] [--json]"""
from __future__ import annotations
import argparse, json, os, sys, zipfile

def check(workspace: str) -> dict:
    task = json.load(open(os.path.join(workspace, "task.json"), encoding="utf-8"))
    out = os.path.join(workspace, "out"); problems = []; checks = []
    for d in task.get("required_deliverables", []):
        p = os.path.join(out, d["path"])
        if d["path"].endswith("/") or d.get("kind") in ("dir", "directory"):
            ok = os.path.isdir(p) and any(os.path.isfile(os.path.join(dp, f)) for dp, _, fs in os.walk(p) for f in fs)
            checks.append({"path": d["path"], "exists": ok})
            if not ok: problems.append(f"missing or empty directory deliverable: out/{d['path']}")
            continue
        ok = os.path.isfile(p) and os.path.getsize(p) > 0
        checks.append({"path": d["path"], "exists": ok})
        if not ok: problems.append(f"missing or empty deliverable: out/{d['path']}"); continue
        kind = d.get("kind") or os.path.splitext(p)[1].lstrip(".")
        if kind in ("xlsx", "docx", "pptx"):
            try:
                with zipfile.ZipFile(p) as z: z.testzip()
            except Exception as e:  # noqa: BLE001
                problems.append(f"out/{d['path']} is not a valid {kind} (zip) file: {e}"); continue
            if kind == "xlsx" and d.get("must_have_formulas"):
                try:
                    import openpyxl
                    wb = openpyxl.load_workbook(p); n = sum(1 for ws in wb.worksheets for row in ws.iter_rows() for c in row if isinstance(c.value, str) and c.value.startswith("="))
                    if n == 0: problems.append(f"out/{d['path']} contains no formulas (the brief requires a working spreadsheet)")
                    errs = 0
                    try:
                        wb2 = openpyxl.load_workbook(p, data_only=True)
                        errs = sum(1 for ws in wb2.worksheets for row in ws.iter_rows() for c in row if isinstance(c.value, str) and c.value.startswith("#"))
                    except Exception: pass  # noqa: BLE001
                    if errs: problems.append(f"out/{d['path']} has {errs} cells with error values (#REF!, #DIV/0!, ...)")
                except ImportError: pass
            if kind == "docx":
                try:
                    import docx
                    n = sum(len(par.text.split()) for par in docx.Document(p).paragraphs)
                    if n < 50: problems.append(f"out/{d['path']} has only {n} words")
                except ImportError: pass
        if kind == "json":
            try: json.load(open(p, encoding="utf-8"))
            except Exception as e: problems.append(f"out/{d['path']} is not valid JSON: {e}")  # noqa: BLE001
    cp = os.path.join(out, "COMPLETION.json")
    if os.path.isfile(cp):
        try:
            c = json.load(open(cp, encoding="utf-8"))
            for k in ("status", "summary", "deliverables"):
                if k not in c: problems.append(f"COMPLETION.json lacks '{k}'")
            if c.get("status") not in ("completed", "needs_clarification", "abstained"): problems.append("COMPLETION.json status must be completed | needs_clarification | abstained")
        except Exception as e: problems.append(f"COMPLETION.json invalid: {e}")  # noqa: BLE001
    else: problems.append("missing out/COMPLETION.json")
    return {"ok": not problems, "problems": problems, "checks": checks}

if __name__ == "__main__":
    ap = argparse.ArgumentParser(); ap.add_argument("--workspace", default="."); ap.add_argument("--json", action="store_true")
    a = ap.parse_args(); r = check(a.workspace)
    if a.json: print(json.dumps(r, indent=1))
    else:
        print("SELF-CHECK:", "OK" if r["ok"] else f"{len(r['problems'])} problem(s)")
        for p in r["problems"]: print(" -", p)
    sys.exit(0 if r["ok"] else 1)