"""Independent public diagnostic audit (Python 3.10+, standard library only).

Download this file plus funding.json/events.json/cross.json and the three
*-diagnostics.csv files into one directory. Run: python reproduce.py
Checks derived published diagnostics. Raw price/return/source validity requires
the original official archives or entitled Market Events releases.
"""
import csv
import json
from collections import Counter, defaultdict
from datetime import datetime
from pathlib import Path

P = Path(__file__).resolve().parent
load = lambda n: json.loads((P / n).read_text())
rows = lambda n: list(csv.DictReader((P / n).open(newline="")))
yes = lambda v: v == "True"
instant = lambda s: datetime.fromisoformat(s.replace("Z", "+00:00")).timestamp()

funding = load("funding.json")
fd = rows("funding-diagnostics.csv")
for expected in funding["sensitivity"]:
    rs = [r for r in fd if int(r["assumed_oi_publication_delay_seconds"]) == expected["assumed_oi_publication_delay_seconds"]]
    assert len(rs) == expected["settlements"]
    ages = [instant(r["settlement_utc"]) - instant(r["matched_oi_utc"]) if r["matched_oi_utc"] else None for r in rs]
    for r, age in zip(rs, ages):
        assert age is None or abs(age - float(r["oi_age_seconds"])) < 1e-5
        assert age is None or age + 1e-6 >= expected["assumed_oi_publication_delay_seconds"]
    assert sum(a == 0 for a in ages) == expected["exact_matches"]
    assert sum(a is not None and 0 <= a <= 300 for a in ages) == expected["within_300s"]
    assert sum(a is not None and 0 <= a <= 900 for a in ages) == expected["within_900s"]
for expected in funding["contracts"]:
    rs = [r for r in fd if r["symbol"] == expected["symbol"] and r["assumed_oi_publication_delay_seconds"] == "0"]
    assert len(rs) == expected["funding_rows"]
    assert sum(yes(r["exact_event_time_match"]) for r in rs) == expected["exact_matches"]
    assert sum(yes(r["within_300s"]) for r in rs) == expected["asof_300s"]

events = load("events.json")
ed = rows("event-diagnostics.csv")
assert len(ed) == events["rows"] == len({r["diagnostic_id"] for r in ed})
local = lambda r: r["feature_status"] == "locally_known" and yes(r["pit_flag"]) and yes(r["available_by_sample_time"])
eligible = lambda r: local(r) and yes(r["valid_active_buy"]) and yes(r["complete_values_24h"])
assert [len(ed), sum(local(r) for r in ed), sum(local(r) and yes(r["valid_active_buy"]) for r in ed), sum(eligible(r) for r in ed)] == [r["rows"] for r in events["funnel"]]
assert dict(Counter(r["sample_type"] for r in ed)) == events["samples"]
for h, expected in events["outcomes"].items():
    assert dict(Counter(r[f"status_{h}"] for r in ed)) == expected
for m in events["matrix"]:
    rs = [r for r in ed if r["symbol"] == m["symbol"] and r["event_kind"] == m["event_kind"]]
    assert len(rs) == m["rows"]
    assert sum(local(r) for r in rs) == m["locally_known"]
    assert sum(yes(r["complete_values_24h"]) for r in rs) == m["descriptive_24h"]
    assert sum(eligible(r) for r in rs) == m["strict_active_buy_24h"]

cross = load("cross.json")
cd = rows("cross-diagnostics.csv")
for expected in cross["sensitivity"]:
    groups = defaultdict(list)
    for r in cd:
        if r["symbol"] == expected["symbol"] and r["quote"] == "USDT":
            groups[r["grid_utc"]].append(r)
    count = 0
    for group in groups.values():
        fresh = []
        for r in group:
            age = instant(r["grid_utc"]) - instant(r["observed_at_utc"]) if r["observed_at_utc"] else None
            assert age is None or abs(age - float(r["local_age_seconds"])) < 1e-5
            fresh.append(age is not None and 0 <= age <= expected["tolerance_seconds"])
        count += len(group) == 3 and all(fresh)
    assert count == expected["all_three_locally_fresh"]
print(json.dumps({"status":"PASS", "funding_diagnostics":len(fd), "event_diagnostics":len(ed), "cross_diagnostics":len(cd), "event_matrix_cells":len(events["matrix"]), "limitation":"Independent reaggregation of public diagnostics; not independent raw-source certification."}))