"""Independent public-CSV arithmetic audit; no access to the private source database."""
import csv
import json
import sys
from collections import Counter
from pathlib import Path

root = Path(sys.argv[1]) if len(sys.argv) > 1 else Path(__file__).resolve().parent
summary = json.loads((root / "research.json").read_text())
with (root / "observations.csv").open(newline="") as f:
    observations = list(csv.DictReader(f))
with (root / "daily-account-categories.csv").open(newline="") as f:
    groups = list(csv.DictReader(f))
assert len(observations) == summary["recorded_decisions"]
assert [int(row["observation_id"]) for row in observations] == list(range(1, len(observations)+1))
counts = Counter(row["category"] for row in observations)
assert dict(sorted(counts.items())) == summary["categories"]
daily = Counter((row["utc_date"], row["account"], row["category"]) for row in observations)
assert len(daily) == len(groups)
for row in groups:
    key = row["utc_date"], row["account"], row["category"]
    assert daily[key] == int(row["recorded_decisions"]), key
for account in summary["accounts"]:
    actual = Counter(row["category"] for row in observations if row["account"] == account["account"])
    assert account["recorded_decisions"] == sum(actual.values())
    assert account["categories"] == dict(sorted(actual.items()))
print(f"PASS: {len(observations)} public observations and {len(groups)} grouped rows reconcile exactly; private-source completeness remains unaudited")