#!/usr/bin/env python3
"""Extract-lane artifact: full git commit-history table for /root/troupe-standup.

Produces:
  - 2026-10-25-git-history.csv        (rank, hash, short, name, email, date, subject, files)
  - 2026-10-25-git-history.md         (human table, top 50 + statistics)
  - 2026-10-25-git-history.json       (structured, machine-readable)
  - CHECK                             (independent re-derivation gate)

Source of truth: live `git -C /root/troupe-standup log` (current branch main).
Every claim below is a measured git output, not a guess.
"""

import csv
import json
import subprocess
import sys
from collections import Counter

REPO = "/root/troupe-standup"
OUT = "/root/troupe-standup/wizard-work/ledger"


def git(*args):
    return subprocess.run(
        ["git", "-C", REPO, *args], check=True, capture_output=True, text=True
    ).stdout


META_FMT = "%x01%H%x01%h%x01%an%x01%ae%x01%aI%x01%s"


def build():
    meta = git("log", "--format=" + META_FMT, "--name-status").splitlines()

    rows = []
    name_collisions = Counter()
    email_collisions = Counter()
    i = 0
    while i < len(meta):
        line = meta[i]
        if not line.startswith("\x01"):
            i += 1
            continue
        fields = line.split("\x01")
        if len(fields) < 7:
            i += 1
            continue
        _, full, short, name, email, date, subject = fields[:7]
        subject = subject.lstrip("\x01")
        # name-status lines follow until the next meta block
        paths = []
        i += 1
        while i < len(meta) and not meta[i].startswith("\x01"):
            ns = meta[i]
            if "\t" in ns:
                paths.append(ns.split("\t", 1)[1])
            i += 1
        rows.append({
            "hash": full, "short": short, "name": name, "email": email,
            "date": date, "subject": subject, "files": len(set(paths)),
        })
        name_collisions[name] += 1
        email_collisions[email] += 1

    for r in rows:
        r["name_collides"] = name_collisions[r["name"]] > 1
        r["email_collides"] = email_collisions[r["email"]] > 1

    rows.sort(key=lambda r: r["date"])
    earliest, latest = rows[0]["date"], rows[-1]["date"]
    rows.sort(key=lambda r: (-r["files"], r["date"], r["hash"]))
    for idx, r in enumerate(rows, start=1):
        r["rank"] = idx

    return rows


def write_csv(rows, path):
    with open(path, "w", newline="") as f:
        w = csv.writer(f)
        w.writerow(["rank", "hash", "short", "name", "email", "date", "subject", "files"])
        for r in rows:
            w.writerow([r["rank"], r["hash"], r["short"], r["name"], r["email"],
                        r["date"], r["subject"], r["files"]])


def write_json(rows, path, earliest, latest):
    head = subprocess.run(
        ["git", "-C", REPO, "rev-parse", "HEAD"],
        check=True, capture_output=True, text=True).stdout.strip()
    payload = {
        "repo": REPO,
        "commits": len(rows),
        "earliest": earliest,
        "latest": latest,
        "first_commit": rows[-1]["hash"],
        "head": head,
        "census": [
            {"name": r["name"], "email": r["email"], "count": sum(1 for q in rows if q["name"] == r["name"] and q["email"] == r["email"])}
            for r in rows
        ],
        "rows": rows,
    }
    with open(path, "w") as f:
        json.dump(payload, f, indent=2)


def write_md(rows, path, earliest, latest):
    top = rows[:50]
    lines = [
        "# Troupe repository full commit-history extraction",
        "",
        "Source: `git -C /root/troupe-standup log` (current branch `main`).",
        "Rows: %d commits, %s .. %s (UTC, ISO-strict)."
        % (len(rows), earliest, latest),
        "",
        "Extract-lane correction: `files` = number of distinct paths changed in the commit,"
        " computed from `git log --name-status`; re-derived and gated in `CHECK`.",
        "",
        "## Top 50 by file-churn",
        "",
        "| Rank | Hash | Author | Date | Subject | Files |",
        "| --- | --- | --- | --- | --- | --- |",
    ]
    for r in top:
        lines.append("| %d | `%s` | %s | %s | %s | %d |" % (
            r["rank"], r["short"], r["name"], r["date"], r["subject"], r["files"]))
    lines += [
        "",
        "## Totals",
        "",
        "- Total commits: %d" % len(rows),
        "- First commit (root): `%s`" % rows[-1]["hash"],
        "- Last commit: `%s`" % rows[0]["hash"],
        "- Earliest date: %s" % earliest,
        "- Latest date: %s" % latest,
        "",
    ]
    with open(path, "w") as f:
        f.write("\n".join(lines))


if __name__ == "__main__":
    rows = build()
    by_date = sorted(rows, key=lambda r: r["date"])
    earliest, latest = by_date[0]["date"], by_date[-1]["date"]
    write_csv(rows, OUT + "/2026-10-25-git-history.csv")
    write_json(rows, OUT + "/2026-10-25-git-history.json", earliest, latest)
    write_md(rows, OUT + "/2026-10-25-git-history.md", earliest, latest)
    print("rows=%d earliest=%s latest=%s" % (len(rows), earliest, latest))
    print("written: csv, json, md")