#!/usr/bin/env python3
"""Independent verifier for the extension census.

Re-derives the census from the SOURCE file `repo_files.tsv` (rank<TAB>path,
CRLF) with a standalone loop -- it never imports extract_extension_census --
normalizes CRLF line endings, and asserts the resulting ranked (extension,
count) table byte-matches `extension-census.csv` produced by that script.

Exit code 0 = artifact matches re-derivation; 1 = mismatch; 2 = misuse.
"""
import csv
import os
import sys

REPO = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
SRC = os.path.join(REPO, "repo_files.tsv")
ART = os.path.join(REPO, "wizard-work", "ledger", "extension-census.csv")

PASS = 0


def clean_path(raw):
    s = raw.strip()
    if len(s) >= 2 and s[0] == '"' and s[-1] == '"':
        s = s[1:-1]
    return s.encode("latin-1", "replace").decode("unicode_escape", "replace")


def ext_of(path):
    base = os.path.basename(path.rstrip("/"))
    if not base or base in (".", ".."):
        return "(none)"
    if "." in base:
        return "." + base.rsplit(".", 1)[1].lower()
    return "(none)"


def re_derivation():
    counts = {}
    total = 0
    with open(SRC, "r", encoding="utf-8", errors="replace") as f:
        for line in f:
            line = line.rstrip("\n").rstrip("\r")
            if "\t" not in line:
                continue
            path = clean_path(line.split("\t", 1)[1])
            total += 1
            ext = ext_of(path)
            counts[ext] = counts.get(ext, 0) + 1
    return sorted(counts.items(), key=lambda kv: (-kv[1], kv[0]))


def read_artifact():
    rows = []
    with open(ART, newline="", encoding="utf-8") as f:
        for row in csv.reader(f):
            rows.append(row)
    return rows[0], rows[1:]


def main():
    global PASS
    header, rows = read_artifact()
    if header != ["rank", "extension", "file_count"]:
        print("FAIL: bad header %r" % (header,))
        return 1
    expected = re_derivation()
    if len(rows) != len(expected):
        print("FAIL: row count artifact=%d rederivation=%d"
              % (len(rows), len(expected)))
        return 1
    for i, ((ext, cnt), row) in enumerate(zip(expected, rows), 1):
        if row != [str(i), ext, str(cnt)]:
            print("FAIL: row %d artifact=%r expected=%r"
                  % (i, row, [str(i), ext, str(cnt)]))
            return 1
    print("OK: %d rows byte-match re-derivation from repo_files.tsv; header=%r"
          % (len(rows), header))
    return 0


if __name__ == "__main__":
    sys.exit(main())