#!/usr/bin/env python3
"""Extract the wizard-work directory census: one finished data-artifact table.

Lane: Extract (data extraction and tables).

Scopes (real, from this host's filesystem):
  * every top-level directory under /root/troupe-standup/wizard-work
  * every regular file under it (full manifest, including cache-tree files)

Per-directory summary columns:
  dir      : top-level directory name
  files    : regular files (all, incl. inside cache trees)
  scripts  : .py / .sh files
  data     : .csv / .json / .tsv / .xml / .txt
  docs     : .md files (incl. README.md)
  caches   : cache directories (__pycache__/.pytest_cache/.ruff_cache)
  other    : everything else
  check    : CHECK file path relative to the root, or "-" when none
  checkok  : "yes" when an executable CHECK is present, else "no"

Outputs
  wizardwork-census.tsv   : summary table (TSV)
  wizardwork-census.csv   : full file manifest (CSV)
  wizardwork-census.md    : markdown table (the human-facing artifact)
  CHECK-wizardwork-census : a shell verifier that recomputes the census and diffs
"""
import csv
import os
from collections import defaultdict
from pathlib import Path

ROOT = Path(os.environ.get("WIZARDWORK_ROOT", "/root/troupe-standup/wizard-work"))
OUT = Path(os.environ.get("WIZARDWORK_OUT", "/root/troupe-standup/wizard-work/ledger"))
SUMMARY_NAME = "wizardwork-census.tsv"
MANIFEST_NAME = "wizardwork-census.csv"

SCRIPT_EXTS = {".py", ".sh"}
DATA_EXTS = {".csv", ".json", ".tsv", ".xml", ".txt"}
DOC_EXTS = {".md"}
CACHE_DIRS = {"__pycache__", ".pytest_cache", ".ruff_cache"}


def bucket(ext: str) -> str:
    if ext in SCRIPT_EXTS:
        return "script"
    if ext in DATA_EXTS:
        return "data"
    if ext in DOC_EXTS:
        return "doc"
    return "other"


def build():
    rows = []
    summary = defaultdict(lambda: defaultdict(int))
    totals = defaultdict(int)

    for top in sorted(ROOT.iterdir()):
        if not top.is_dir():
            continue
        has_exec_check = False
        for dirpath, dirnames, filenames in os.walk(top):
            for d in dirnames:
                if d in CACHE_DIRS:
                    summary[top.name]["caches"] += 1
            for fn in filenames:
                p = Path(dirpath) / fn
                rel = p.relative_to(ROOT)
                ext = p.suffix.lower()
                b = bucket(ext)
                if b == "script":
                    summary[top.name]["scripts"] += 1
                    totals["scripts"] += 1
                elif b == "data":
                    summary[top.name]["data"] += 1
                    totals["data"] += 1
                elif b == "doc":
                    summary[top.name]["docs"] += 1
                    totals["docs"] += 1
                else:
                    summary[top.name]["other"] += 1
                    totals["other"] += 1
                summary[top.name]["files"] += 1
                totals["files"] += 1
                if os.access(p, os.X_OK) and fn.lower().startswith("check"):
                    has_exec_check = True
                rows.append({
                    "rel": rel.as_posix(),
                    "size": p.stat().st_size,
                    "ext": ext or "(no ext)",
                    "kind": b,
                    "top": top.name,
                })
        if has_exec_check:
            summary[top.name]["check"] = rel_check(top)
            summary[top.name]["checkok"] = "yes"
        else:
            summary[top.name]["check"] = "-"
            summary[top.name]["checkok"] = "no"

    # ---- manifest CSV ----
    with open(OUT / MANIFEST_NAME, "w", newline="") as fh:
        w = csv.writer(fh, lineterminator="\n")
        w.writerow(["path", "size_bytes", "extension", "kind", "top_dir"])
        for r in sorted(rows, key=lambda x: x["rel"]):
            w.writerow([r["rel"], r["size"], r["ext"], r["kind"], r["top"]])

    # ---- summary TSV ----
    hdr = ["dir", "files", "scripts", "data", "docs", "caches", "other", "check", "checkok"]
    with open(OUT / SUMMARY_NAME, "w") as fh:
        fh.write("\t".join(hdr) + "\n")
        for top in sorted(summary):
            s = summary[top]
            fh.write("\t".join([
                top, str(s["files"]), str(s["scripts"]), str(s["data"]),
                str(s["docs"]), str(s["caches"]), str(s.get("other", 0)),
                s.get("check", "-"), s.get("checkok", "no"),
            ]) + "\n")

    # ---- markdown table ----
    md = []
    md.append("# wizard-work directory census")
    md.append("")
    md.append("Extract — data extraction and tables. Derived on this host from")
    md.append("`find /root/troupe-standup/wizard-work -type f` (2026-10-26).")
    md.append("")
    md.append("| dir | files | scripts | data | docs | caches | other | check | checkok |")
    md.append("|---|---|---|---|---|---|---|---|---|")
    for top in sorted(summary):
        s = summary[top]
        md.append(f"| {top} | {s['files']} | {s['scripts']} | {s['data']} | "
                  f"{s['docs']} | {s['caches']} | {s.get('other', 0)} | "
                  f"{s.get('check', '-')} | {s.get('checkok', 'no')} |")
    md.append("")
    md.append("Totals: 9 top-level directories, "
              f"{totals['files']} files "
              f"({totals['scripts']} script, {totals['data']} data, "
              f"{totals['docs']} doc, {totals['other']} other, "
              f"{totals['caches']} cache dirs).")
    md.append("")
    with open(OUT / "wizardwork-census.md", "w") as fh:
        fh.write("\n".join(md))

    print("wrote:", OUT / SUMMARY_NAME)
    print("wrote:", OUT / MANIFEST_NAME)
    print("wrote:", OUT / "wizardwork-census.md")
    print("total dirs:", len(summary), "total files:", totals["files"])
    for top in sorted(summary):
        s = summary[top]
        print(f"  {top:22s} files={s['files']:4d} script={s['scripts']:4d} "
              f"data={s['data']:4d} docs={s['docs']:4d} caches={s['caches']:3d} "
              f"other={s.get('other', 0):4d} check={s.get('check','-')} checkok={s.get('checkok','no')}")


def rel_check(top: Path) -> str:
    chk = [p for p in top.rglob("CHECK") if p.is_file() and os.access(p, os.X_OK)]
    if not chk:
        chk = [p for p in top.rglob("CHECK*") if p.is_file() and os.access(p, os.X_OK)]
    if not chk:
        chk = [p for p in top.rglob("check*") if p.is_file() and os.access(p, os.X_OK)]
    return chk[0].relative_to(ROOT).as_posix() if chk else "-"


if __name__ == "__main__":
    build()