FabledGrimoire/scripts/import_fabledrulebook.py

"""One-shot import of FabledRulebook .md files into the Grimoire DB.

Usage:
    python -m scripts.import_fabledrulebook \\
        --source-dir /path/to/FabledRulebook \\
        --user-id 1 \\
        [--rulebook-title "FabledSword family"] \\
        [--subscribe-projects 3,4,5] \\
        [--dry-run] \\
        [--force]

Calls the services layer (services/rulebooks.py) — does not write raw SQL.
This means any structural mismatch surfaces as a clean service-level
ValueError. The script is not part of Alembic; the operator runs it once
after migration 0055 lands.
"""
from __future__ import annotations

import argparse
import asyncio
import logging
import re
from pathlib import Path


RULE_HEADING_RE = re.compile(r"^##\s+(.+?)\s*$", re.MULTILINE)
WHY_RE = re.compile(r"\*\*\s*Why:?\s*\*\*", re.IGNORECASE)
HOW_RE = re.compile(r"\*\*\s*How to apply:?\s*\*\*", re.IGNORECASE)
H1_RE = re.compile(r"^#\s+.+$", re.MULTILINE)


logger = logging.getLogger("import_fabledrulebook")


def parse_rule_body(title: str, body: str) -> dict:
    """Extract statement, why, how_to_apply from a rule's H2-section body."""
    body = body.strip()
    statement = body
    why = ""
    how_to_apply = ""

    why_match = WHY_RE.search(body)
    how_match = HOW_RE.search(body)

    if why_match:
        statement = body[: why_match.start()].strip()
        if how_match and how_match.start() > why_match.end():
            why = body[why_match.end() : how_match.start()].strip()
            how_to_apply = body[how_match.end() :].strip()
        else:
            why = body[why_match.end() :].strip()
    elif how_match:
        statement = body[: how_match.start()].strip()
        how_to_apply = body[how_match.end() :].strip()

    return {
        "title": title,
        "statement": statement,
        "why": why,
        "how_to_apply": how_to_apply,
    }


def parse_topic_file(path: Path) -> tuple[str, str, list[dict]]:
    """Returns (topic_title, topic_description, [rules])."""
    text = path.read_text(encoding="utf-8")
    topic_title = path.stem  # e.g. "git-workflow"

    parts = RULE_HEADING_RE.split(text)
    preamble = parts[0] if parts else ""
    # Strip H1 from preamble to get just the description.
    description = H1_RE.sub("", preamble, count=1).strip()

    rules: list[dict] = []
    for i in range(1, len(parts), 2):
        rule_title = parts[i].strip()
        body = parts[i + 1] if i + 1 < len(parts) else ""
        rules.append(parse_rule_body(rule_title, body))

    return topic_title, description, rules


async def run(args) -> None:
    source = Path(args.source_dir)
    if not source.exists():
        raise SystemExit(f"Source dir not found: {source}")

    md_files = sorted(p for p in source.glob("*.md") if p.name.lower() != "readme.md")
    if not md_files:
        raise SystemExit(f"No topic .md files found in {source}")

    readme = source / "README.md"
    description = readme.read_text(encoding="utf-8") if readme.exists() else ""

    if args.dry_run:
        print(f"DRY RUN — would import {len(md_files)} topic file(s):")
        for f in md_files:
            topic_title, _, rules = parse_topic_file(f)
            print(f"  TOPIC: {topic_title} ({len(rules)} rules)")
            for r in rules:
                preview = r["statement"][:80] + ("…" if len(r["statement"]) > 80 else "")
                empty_flag = "  [EMPTY STATEMENT — will skip]" if not r["statement"].strip() else ""
                print(f"    - {r['title']}: {preview}{empty_flag}")
        return

    from scribe.services import rulebooks as rb_service

    existing = await rb_service.find_rulebook_by_title(args.user_id, args.rulebook_title)
    if existing is not None and not args.force:
        raise SystemExit(
            f"Rulebook '{args.rulebook_title}' already exists for user "
            f"{args.user_id}; pass --force to create another."
        )

    rb = await rb_service.create_rulebook(
        user_id=args.user_id,
        title=args.rulebook_title,
        description=description,
    )
    logger.info("Created rulebook %d ('%s')", rb.id, rb.title)

    rules_created = 0
    rules_skipped = 0
    for i, f in enumerate(md_files):
        topic_title, topic_desc, rules = parse_topic_file(f)
        topic = await rb_service.create_topic(
            rulebook_id=rb.id,
            user_id=args.user_id,
            title=topic_title,
            description=topic_desc,
            order_index=i,
        )
        logger.info("  Created topic %d ('%s', %d rules)", topic.id, topic.title, len(rules))
        for j, rule in enumerate(rules):
            if not rule["statement"].strip():
                logger.warning("    SKIP empty-statement rule: %s / %s", topic_title, rule["title"])
                rules_skipped += 1
                continue
            await rb_service.create_rule(
                topic_id=topic.id,
                user_id=args.user_id,
                title=rule["title"],
                statement=rule["statement"],
                why=rule["why"],
                how_to_apply=rule["how_to_apply"],
                order_index=j,
            )
            rules_created += 1

    for pid in (args.subscribe_projects or []):
        await rb_service.subscribe_project(
            project_id=pid, rulebook_id=rb.id, user_id=args.user_id,
        )
        logger.info("Subscribed project %d to rulebook %d", pid, rb.id)

    print(
        f"\nImported rulebook '{rb.title}' (id={rb.id}): "
        f"{len(md_files)} topics, {rules_created} rules created"
        + (f", {rules_skipped} skipped" if rules_skipped else "")
    )


def _comma_ints(s: str) -> list[int]:
    return [int(x) for x in s.split(",") if x.strip()]


def main() -> None:
    parser = argparse.ArgumentParser(description=__doc__)
    parser.add_argument("--source-dir", required=True,
                        help="Path to the FabledRulebook directory")
    parser.add_argument("--user-id", type=int, required=True,
                        help="Grimoire user id who will own the rulebook")
    parser.add_argument("--rulebook-title", default="FabledSword family")
    parser.add_argument("--subscribe-projects", type=_comma_ints,
                        help="Comma-separated project ids to auto-subscribe")
    parser.add_argument("--dry-run", action="store_true")
    parser.add_argument("--force", action="store_true",
                        help="Allow creating a rulebook even if title exists")
    args = parser.parse_args()
    logging.basicConfig(level=logging.INFO, format="%(message)s")
    asyncio.run(run(args))


if __name__ == "__main__":
    main()