#!/usr/bin/env python3
"""
Rebuild index.json from SKILL.md frontmatter in this repository.

Usage:
    python3 scripts/rebuild_index.py
    python3 scripts/rebuild_index.py --check   # exit 1 if index.json is out of date

The index is the canonical machine-readable manifest of all skills in this
repository. Regenerate it whenever a SKILL.md is added, removed, or has its
frontmatter changed. Commit index.json together with the SKILL.md change.
"""
from __future__ import annotations

import argparse
import json
import re
import sys
from pathlib import Path

REPO_ROOT = Path(__file__).resolve().parent.parent
SKILLS_ROOT = REPO_ROOT / "skills"
INDEX_PATH = REPO_ROOT / "index.json"

FRONTMATTER_RE = re.compile(r"^---\n(.*?)\n---\n", re.DOTALL)
KEY_RE = re.compile(r"^([A-Za-z_][A-Za-z0-9_]*):\s*(.+?)(?=\n[A-Za-z_]|\Z)", re.MULTILINE | re.DOTALL)


def _grab(frontmatter: str, key: str) -> str:
    m = re.search(rf"^{key}:\s*(.+?)(?=\n[A-Za-z_]|\Z)", frontmatter, re.MULTILINE | re.DOTALL)
    if not m:
        return ""
    val = m.group(1).strip()
    # strip YAML block indicators and quotes
    val = val.strip(">").strip("|").strip()
    if (val.startswith('"') and val.endswith('"')) or (val.startswith("'") and val.endswith("'")):
        val = val[1:-1]
    return val.strip()


def _grab_tags(frontmatter: str) -> list[str]:
    m = re.search(r"tags:\s*\[(.*?)\]", frontmatter, re.DOTALL)
    if not m:
        return []
    return [t.strip().strip('"').strip("'") for t in m.group(1).split(",") if t.strip()]


def build_index() -> dict:
    skills: list[dict] = []
    for skill_md in sorted(SKILLS_ROOT.rglob("SKILL.md")):
        rel = skill_md.relative_to(SKILLS_ROOT)
        parts = rel.parts
        category = parts[0] if len(parts) > 1 else "root"
        name = parts[-2] if len(parts) > 1 else "root"

        text = skill_md.read_text(encoding="utf-8")
        fm_match = FRONTMATTER_RE.match(text)
        frontmatter = fm_match.group(1) if fm_match else ""

        desc = _grab(frontmatter, "description").replace("\n", " ").strip()
        version = _grab(frontmatter, "version")
        author = _grab(frontmatter, "author")
        tags = _grab_tags(frontmatter)

        all_files = sorted(
            str(p.relative_to(SKILLS_ROOT))
            for p in skill_md.parent.rglob("*")
            if p.is_file()
        )

        skills.append(
            {
                "category": category,
                "name": name,
                "path": str(rel),
                "description": desc,
                "version": version,
                "author": author,
                "tags": tags,
                "files": all_files,
            }
        )

    return {
        "repository": "Radix-Skills",
        "version": 1,
        "description": (
            "Delt AI-skill-bibliotek for Radix — bruges af alle medarbejderes "
            "AI-agenter (Hermes, Codex m.fl.) til dansk regnskab/ERP/data/devops."
        ),
        "categories": sorted({s["category"] for s in skills}),
        "skill_count": len(skills),
        "skills": skills,
    }


def main() -> int:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--check", action="store_true", help="Exit 1 if index.json is stale")
    args = parser.parse_args()

    new_index = build_index()
    new_json = json.dumps(new_index, ensure_ascii=False, indent=2) + "\n"

    if args.check:
        if not INDEX_PATH.exists():
            print("index.json does not exist", file=sys.stderr)
            return 1
        current = INDEX_PATH.read_text(encoding="utf-8")
        if current != new_json:
            print("index.json is out of date — run scripts/rebuild_index.py", file=sys.stderr)
            return 1
        return 0

    INDEX_PATH.write_text(new_json, encoding="utf-8")
    print(
        f"Wrote {INDEX_PATH.relative_to(REPO_ROOT)} with {new_index['skill_count']} skills "
        f"across {len(new_index['categories'])} categories: {', '.join(new_index['categories'])}"
    )
    return 0


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