#!/usr/bin/env python3
"""Extract the Wizards Table roster from WIZARDS-TABLE.md into clean tables.

Lane: Extract (data extraction & tables).
Output:
  - /root/.hermes/wizard-outputs/ledger/wizard-roster.csv
  - /root/.hermes/wizard-outputs/ledger/roster-summary.json
  - /root/troupe-standup/wizard-work/ledger/wizard-roster.csv
  - /root/troupe-standup/wizard-work/ledger/roster-summary.json
"""
import csv
import json
import os
import re

SRC = "/root/troupe-standup/WIZARDS-TABLE.md"
OUT_DIR = "/root/.hermes/wizard-outputs/ledger"
REPO_DIR = "/root/troupe-standup/wizard-work/ledger"

os.makedirs(OUT_DIR, exist_ok=True)
os.makedirs(REPO_DIR, exist_ok=True)

with open(SRC, "r", encoding="utf-8") as f:
    lines = f.read().splitlines()

# Locate the roster header row.
header_idx = None
for i, ln in enumerate(lines):
    cells = [c.strip() for c in ln.split("|")]
    cells = [c for c in cells if c != ""]
    if cells == ["Wizard", "Model", "Lane", "Latest check-in", "Current unit"]:
        header_idx = i
        break

if header_idx is None:
    raise SystemExit("roster header not found")

# Collect data rows that follow the header + separator, until a non-table line.
rows = []
j = header_idx + 2  # skip header, skip separator
while j < len(lines):
    ln = lines[j].strip()
    if not ln:
        j += 1
        continue
    cells = [c.strip() for c in ln.split("|")]
    cells = [c for c in cells if c != ""]
    # separator line (all dashes) -> stop
    if all(re.fullmatch(r"-+", c) for c in cells) and len(cells) >= 2:
        break
    # stop at any heading or non-table line
    if ln.startswith("#"):
        break
    if len(cells) == 5:
        rows.append(cells)
    j += 1

expected_cols = ["Wizard", "Model", "Lane", "Latest check-in", "Current unit"]
assert len(rows) == 10, f"expected 10 roster rows, got {len(rows)}"
for r in rows:
    assert len(r) == 5, f"row has {len(r)} cells: {r}"
    assert r[0] == r[0].strip() and r[0], f"empty wizard name: {r}"

wizards = [
    {"Wizard": r[0], "Model": r[1], "Lane": r[2], "Latest check-in": r[3], "Current unit": r[4]}
    for r in rows
]

# ---- CSV ----
csv_path = os.path.join(REPO_DIR, "wizard-roster.csv")
with open(csv_path, "w", newline="", encoding="utf-8") as f:
    w = csv.DictWriter(f, fieldnames=expected_cols, lineterminator="\n")
    w.writeheader()
    w.writerows(wizards)

# ---- Summary JSON ----
staffed = [w["Wizard"] for w in wizards if "Staffed" in w["Current unit"] or "Staffed" in w["Lane"] or "staffed" in w["Current unit"].lower()]
staffed = [w["Wizard"] for w in wizards if "2026-10-15" in w["Latest check-in"]]
staffed += [w["Wizard"] for w in wizards if "Staffed" in w["Current unit"]]
staffed = sorted(set(staffed))

summary = {
    "source": SRC,
    "extract_at": "2026-10-21",
    "total_wizards": len(wizards),
    "wizards": wizards,
    "staffed_count": len(staffed),
    "staffed": staffed,
    "lanes": sorted(set(w["Lane"] for w in wizards)),
    "checkin_dates": sorted(set(w["Latest check-in"] for w in wizards if w["Latest check-in"] != "— (replies daily)")),
}

sum_path = os.path.join(REPO_DIR, "roster-summary.json")
with open(sum_path, "w", encoding="utf-8") as f:
    json.dump(summary, f, indent=2, ensure_ascii=False)

# Mirror into the hermès output dir.
csv_out = os.path.join(OUT_DIR, "wizard-roster.csv")
sum_out = os.path.join(OUT_DIR, "roster-summary.json")
for s, d in [(csv_path, csv_out), (sum_path, sum_out)]:
    with open(s, "rb") as f:
        data = f.read()
    with open(d, "wb") as f:
        f.write(data)

print("rows extracted:", len(rows))
print("staffed:", staffed)
print("lanes:", summary["lanes"])
print("wizard dates:", summary["checkin_dates"])
print("wrote:", csv_path, sum_path)