#!/usr/bin/env python3
"""Independent verifier for the replies volume census artifacts.

Recomputes ground truth from the LIVE <repo>/replies/ directory using shell
pipelines (grep/sed/cut/sort/uniq via subprocess), then asserts the committed
TSV artifacts and the markdown summary match that ground truth.

The extractor parses with a compiled regex over pathlib; this verifier parses
with shell text tools -- two independent code paths over the same evidence
directory, so agreement is a real cross-check.

Exit status:
    0  artifact matches ground truth
    1  mismatch (or artifact missing / unreadable)

Used as the CHECK line:  python3 wizard-work/ledger/verify_reply_volume_census.py
"""

import csv
import subprocess
import sys
from pathlib import Path

REPO_ROOT = Path(__file__).resolve().parents[2]
REPLIES = REPO_ROOT / "replies"
DATE_TSV = REPO_ROOT / "wizard-work" / "ledger" / "reply-volume-census.tsv"
RECIP_TSV = REPO_ROOT / "wizard-work" / "ledger" / "reply-volume-by-recipient.tsv"
MD = REPO_ROOT / "wizard-work" / "ledger" / "reply-volume-census.md"

DATE_PAT = r"^[0-9]{4}-[0-9]{2}-[0-9]{2}-to-"
FIND_BASENAMES = "find replies -maxdepth 1 -type f -name '*.md' -printf '%f\\n'"
SEP = "-------------------------------------------------------------------------------"


def sh(cmd):
    """Run a shell command and return stripped stdout."""
    return subprocess.run(
        cmd, shell=True, check=True, capture_output=True, text=True
    ).stdout


def ground_truth():
    """Independent ground truth derived from the live replies/ directory."""
    date_cmd = (FIND_BASENAMES + " | grep -E '" + DATE_PAT + "' | "
                "sed -E 's/-to-.*//' | sort | uniq -c")
    date_files = sh(date_cmd).strip().splitlines()
    date_counts = {}
    for line in date_files:
        cnt, date = line.split(None, 1)
        date_counts[date] = int(cnt)

    # Recipients: strip the date prefix, drop ".md", then drop a trailing "-NN".
    recip_cmd = (FIND_BASENAMES + " | grep -E '" + DATE_PAT + "' | "
                 "sed -E 's/^" + DATE_PAT + "//' | sed -E 's/\\.md$//' | "
                 "sed -E 's/-[0-9]+$//' | sort | uniq -c")
    recip_files = sh(recip_cmd).strip().splitlines()
    recip_counts = {}
    for line in recip_files:
        cnt, recip = line.split(None, 1)
        recip_counts[recip] = int(cnt)

    return date_counts, recip_counts, sum(date_counts.values())


def parse_tsv(path):
    rows = {}
    with path.open(newline="", encoding="utf-8") as fh:
        reader = csv.reader(fh, delimiter="\t")
        next(reader, None)  # skip header row
        for row in reader:
            if len(row) == 2:
                rows[row[0]] = int(row[1])
    return rows


def main():
    date_gt, recip_gt, total_gt = ground_truth()
    problems = []

    for label, tsv_path, gt in (("date", DATE_TSV, date_gt),
                                ("recipient", RECIP_TSV, recip_gt)):
        if not tsv_path.is_file():
            problems.append("missing " + str(tsv_path))
            continue
        rows = parse_tsv(tsv_path)
        rows_data = {k: v for k, v in rows.items() if k != "TOTAL"}
        if rows_data != gt:
            problems.append(
                "%s: data rows differ; got %d keys, ground truth %d keys"
                % (label, len(rows_data), len(gt))
            )
        if rows.get("TOTAL", -1) != sum(gt.values()):
            problems.append(
                "%s: TOTAL row %r != sum of rows %d"
                % (label, rows.get("TOTAL"), sum(gt.values()))
            )

    # Markdown summary must name the correct headline numbers.
    if MD.is_file():
        text = MD.read_text(encoding="utf-8")
        for token in ("430", "34", "17", "README.md"):
            if token not in text:
                problems.append("markdown missing assertion: %s" % token)
    else:
        problems.append("missing " + str(MD))

    if problems:
        print("CHECK FAIL")
        for p in problems:
            print("  -", p)
        sys.exit(1)

    print(
        "CHECK PASS: %d replies, %d dates, %d recipients "
        "(md_files=%d, canonical=%d)"
        % (total_gt, len(date_gt), len(recip_gt), md_total(), total_gt)
    )
    sys.exit(0)


def md_total():
    return len(sh("ls -1 replies/*.md").splitlines())


if __name__ == "__main__":
    main()