mindnet/scripts/import_markdown.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Script: scripts/import_markdown.py — Markdown → Qdrant (Notes, Chunks, Edges)
Version: 3.5.0
Datum: 2025-09-09

Kurzbeschreibung
----------------
- Liest Markdown-Dateien ein, erzeugt Notes/Chunks/Edges **idempotent**.
- **Änderungserkennung** (nur Inhalte, keine FS-Zeitstempel) konfigurierbar:
  * ``--hash-mode``: body | frontmatter | body+frontmatter | full (Alias)
    - Env: ``MINDNET_HASH_MODE`` **oder** ``MINDNET_HASH_COMPARE`` (Body/Frontmatter/Full)
  * ``--hash-normalize``: canonical | none (Default: canonical)
  * ``--hash-source``: parsed (Default) | raw
    - "raw" hasht den **ungeparsten** Body (Frontmatter via Regex entfernt).
  * **NEU**: ``--compare-text`` (oder ENV ``MINDNET_COMPARE_TEXT=true``)
    - Zusätzlich zum Hash wird der *parsed* ``fulltext`` direkt verglichen.
    - Erkennt Änderungen auch dann, wenn eine Normalisierung Unterschiede glättet.
  * **NEU**: Signaturabgleich:
    - Falls sich die Signatur (z. B. body→full, parsed→raw, canonical→none) zwischen Alt/Neu unterscheidet,
      gilt die Note als **geändert** (Einmal-Update, um neue Signatur zu persistieren).

ENV / Qdrant
------------
- QDRANT_URL | QDRANT_HOST/QDRANT_PORT | QDRANT_API_KEY
- COLLECTION_PREFIX (Default: mindnet)
- VECTOR_DIM (Default: 384)
- MINDNET_NOTE_SCOPE_REFS: true|false (Default: false)
- MINDNET_COMPARE_TEXT: true|false (Default: false)

Aufruf-Beispiele
----------------
    # Standard (nur Body, kanonisiert, parsed-Quelle)
    python3 -m scripts.import_markdown --vault ./vault

    # Sensibel (jede Kleinigkeit): raw-Quelle + keine Normalisierung
    python3 -m scripts.import_markdown --vault ./vault --apply --hash-source raw --hash-normalize none

    # Full-Vergleich (Body+Frontmatter)
    MINDNET_HASH_COMPARE=Full python3 -m scripts.import_markdown --vault ./vault --apply

    # Zusätzlich Body-Text direkt vergleichen (maximale Sicherheit)
    python3 -m scripts.import_markdown --vault ./vault --apply --compare-text
"""
from __future__ import annotations
import argparse
import difflib
import json
import os
import sys
from typing import Dict, List, Optional, Tuple

from dotenv import load_dotenv
from qdrant_client.http import models as rest

from app.core.parser import (
    read_markdown,
    normalize_frontmatter,
    validate_required_frontmatter,
)
from app.core.note_payload import make_note_payload
from app.core.chunker import assemble_chunks
from app.core.chunk_payload import make_chunk_payloads
from app.core.edges import build_edges_for_note
from app.core.qdrant import (
    QdrantConfig,
    get_client,
    ensure_collections,
    ensure_payload_indexes,
)
from app.core.qdrant_points import (
    points_for_chunks,
    points_for_note,
    points_for_edges,
    upsert_batch,
)

try:
    from app.core.embed import embed_texts  # optional
except Exception:
    embed_texts = None


# ---------------------------------------------------------------------
# Helpers
# ---------------------------------------------------------------------

def iter_md(root: str) -> List[str]:
    out: List[str] = []
    for dirpath, _, filenames in os.walk(root):
        for fn in filenames:
            if not fn.lower().endswith(".md"):
                continue
            p = os.path.join(dirpath, fn)
            pn = p.replace("\\", "/")
            if any(ex in pn for ex in ["/.obsidian/", "/_backup_frontmatter/", "/_imported/"]):
                continue
            out.append(p)
    return sorted(out)

def collections(prefix: str) -> Tuple[str, str, str]:
    return f"{prefix}_notes", f"{prefix}_chunks", f"{prefix}_edges"

def fetch_existing_note_payload(client, prefix: str, note_id: str) -> Optional[Dict]:
    notes_col, _, _ = collections(prefix)
    f = rest.Filter(must=[rest.FieldCondition(key="note_id", match=rest.MatchValue(value=note_id))])
    points, _ = client.scroll(
        collection_name=notes_col,
        scroll_filter=f,
        with_payload=True,
        with_vectors=False,
        limit=1,
    )
    if not points:
        return None
    return points[0].payload or {}

def purge_note_artifacts(client, prefix: str, note_id: str) -> None:
    _, chunks_col, edges_col = collections(prefix)
    f_chunks = rest.Filter(must=[rest.FieldCondition(key="note_id", match=rest.MatchValue(value=note_id))])
    client.delete(collection_name=chunks_col, points_selector=f_chunks, wait=True)
    f_edges = rest.Filter(must=[rest.FieldCondition(key="note_id", match=rest.MatchValue(value=note_id))])
    client.delete(collection_name=edges_col, points_selector=f_edges, wait=True)

def _normalize_rel_path(abs_path: str, vault_root: str) -> str:
    try:
        rel = os.path.relpath(abs_path, vault_root)
    except Exception:
        rel = abs_path
    return rel.replace("\\", "/").lstrip("/")


# ---------------------------------------------------------------------
# Main
# ---------------------------------------------------------------------

def main() -> None:
    load_dotenv()
    ap = argparse.ArgumentParser()
    ap.add_argument("--vault", required=True, help="Pfad zum Obsidian-Vault (Root-Ordner)")
    ap.add_argument("--apply", action="store_true", help="Schreibt in Qdrant; ohne Flag nur Dry-Run")
    ap.add_argument("--purge-before-upsert", action="store_true",
                    help="Vor Upsert Chunks & Edges der GEÄNDERTEN Note löschen")
    ap.add_argument("--note-id", help="Nur eine bestimmte Note-ID verarbeiten")
    ap.add_argument("--embed-note", action="store_true", help="Optional: Note-Volltext einbetten")
    ap.add_argument("--force-replace", action="store_true",
                    help="Änderungserkennung ignorieren und immer upserten (+ optional Purge)")
    ap.add_argument("--hash-mode", choices=["body", "frontmatter", "body+frontmatter", "full"], default=None,
                    help="Vergleichsmodus: Body | Frontmatter | body+frontmatter (Alias: full)")
    ap.add_argument("--hash-normalize", choices=["canonical", "none"], default=None)
    ap.add_argument("--hash-source", choices=["parsed", "raw"], default=None,
                    help="Quelle für die Hash-Berechnung (Default: parsed)")
    ap.add_argument("--note-scope-refs", action="store_true",
                    help="(Optional) erzeugt zusätzlich references:note (Default: aus)")
    ap.add_argument("--debug-hash-diff", action="store_true",
                    help="Zeigt bei Bedarf einen kurzen Diff zwischen altem und neuem Body")
    ap.add_argument("--compare-text", action="store_true",
                    help="Parsed fulltext zusätzlich direkt vergleichen (über Hash hinaus)")
    args = ap.parse_args()

    note_scope_refs_env = (os.environ.get("MINDNET_NOTE_SCOPE_REFS", "false").strip().lower() == "true")
    note_scope_refs = args.note_scope_refs or note_scope_refs_env

    compare_text = args.compare_text or (os.environ.get("MINDNET_COMPARE_TEXT", "false").strip().lower() == "true")

    cfg = QdrantConfig.from_env()
    client = get_client(cfg)
    ensure_collections(client, cfg.prefix, cfg.dim)
    ensure_payload_indexes(client, cfg.prefix)

    root = os.path.abspath(args.vault)
    files = iter_md(root)
    if not files:
        print("Keine Markdown-Dateien gefunden.", file=sys.stderr)
        sys.exit(2)

    processed = 0
    for path in files:
        parsed = read_markdown(path)
        fm = normalize_frontmatter(parsed.frontmatter)

        try:
            validate_required_frontmatter(fm)
        except Exception as e:
            print(json.dumps({"path": path, "error": f"Frontmatter invalid: {e}"}))
            continue

        if args.note_id and fm.get("id") != args.note_id:
            continue

        processed += 1

        # Note-Payload (inkl. Hash-Steuerung & Quelle)
        note_pl = make_note_payload(
            parsed,
            vault_root=root,
            hash_mode=args.hash_mode,
            hash_normalize=args.hash_normalize,
            hash_source=args.hash_source,
            file_path=path,
        )

        if "fulltext" not in (note_pl or {}):
            note_pl["fulltext"] = parsed.body or ""
        if note_pl.get("path"):
            note_pl["path"] = _normalize_rel_path(
                os.path.join(root, note_pl["path"]) if not os.path.isabs(note_pl["path"]) else note_pl["path"], root
            )
        else:
            note_pl["path"] = _normalize_rel_path(parsed.path, root)

        note_id = note_pl["note_id"]

        # Change-Detection (nur Inhalte, keine FS-Timestamps)
        old_payload = None if args.force_replace else fetch_existing_note_payload(client, cfg.prefix, note_id)
        old_hash = None if not old_payload else old_payload.get("hash_fulltext")
        old_sig = (old_payload or {}).get("hash_signature")
        new_hash = note_pl.get("hash_fulltext")
        new_sig = note_pl.get("hash_signature")

        # 1) Signaturwechsel → als geändert behandeln (Einmal-Update)
        sig_changed = bool(old_sig) and bool(new_sig) and (old_sig.split(":")[:3] != new_sig.split(":")[:3])

        # 2) Hash-Vergleich
        hash_changed = (old_hash != new_hash)

        # 3) Optional: Parsed-Text direkt vergleichen (zusätzlich)
        text_changed = False
        if compare_text:
            old_text = (old_payload or {}).get("fulltext") or ""
            new_text = note_pl.get("fulltext") or ""
            text_changed = (old_text != new_text)

        changed = args.force_replace or sig_changed or hash_changed or text_changed

        # Optionales Debugging: kompakten Diff anzeigen
        if args.debug_hash_diff:
            old_text = (old_payload or {}).get("fulltext") or ""
            new_text = note_pl.get("fulltext") or ""
            # Hinweis, wenn Hash gleich aber Text verschieden (oder Signaturwechsel)
            if (not hash_changed) and (old_text != new_text or sig_changed):
                print(json.dumps({
                    "debug": "hash_equal_but_text_or_signature_differs",
                    "note_id": note_id,
                    "old_sig": old_sig,
                    "new_sig": new_sig,
                    "hash_mode": args.hash_mode or os.environ.get("MINDNET_HASH_MODE") or os.environ.get("MINDNET_HASH_COMPARE", "body"),
                    "hash_normalize": args.hash_normalize or os.environ.get("MINDNET_HASH_NORMALIZE", "canonical"),
                    "hash_source": args.hash_source or os.environ.get("MINDNET_HASH_SOURCE", "parsed"),
                }, ensure_ascii=False))
            if old_text and new_text:
                ud = list(difflib.unified_diff(
                    old_text.splitlines(), new_text.splitlines(),
                    fromfile="qdrant_fulltext(old)", tofile="vault_body(new)",
                    n=3
                ))
                if ud:
                    preview = "\n".join(ud[:50])
                    print(json.dumps({"note_id": note_id, "diff_preview": preview}, ensure_ascii=False))

        # Chunks + Embeddings (Nullvektor-Fallback)
        chunks = assemble_chunks(fm["id"], parsed.body, fm.get("type", "concept"))
        chunk_pls = make_chunk_payloads(fm, note_pl["path"], chunks)
        if embed_texts:
            vecs = embed_texts([getattr(c, "text", "") for c in chunks])  # type: ignore[attr-defined]
        else:
            vecs = [[0.0] * cfg.dim for _ in chunks]

        # Edges (neues Schema, mit note_id als Owner)
        note_refs = note_pl.get("references") or []
        edges = build_edges_for_note(
            note_id,
            chunk_pls,
            note_refs,
            include_note_scope_refs=note_scope_refs,
        )

        # Zusammenfassung pro Datei
        summary = {
            "note_id": note_id,
            "title": fm.get("title"),
            "chunks": len(chunk_pls),
            "edges": len(edges),
            "changed": changed,
            "decision": ("apply" if args.apply and changed else
                         "apply-skip-unchanged" if args.apply and not changed else
                         "dry-run"),
            "path": note_pl["path"],
            "hash_mode": args.hash_mode or os.environ.get("MINDNET_HASH_MODE") or os.environ.get("MINDNET_HASH_COMPARE", "body"),
            "hash_normalize": args.hash_normalize or os.environ.get("MINDNET_HASH_NORMALIZE", "canonical"),
            "hash_source": args.hash_source or os.environ.get("MINDNET_HASH_SOURCE", "parsed"),
            "hash_signature": note_pl.get("hash_signature"),
            "sig_changed": sig_changed,
            "hash_changed": hash_changed,
            "text_changed": text_changed,
        }
        print(json.dumps(summary, ensure_ascii=False))

        if not args.apply:
            continue

        if changed and args.purge_before_upsert:
            purge_note_artifacts(client, cfg.prefix, note_id)

        # Upserts
        notes_name, note_pts = points_for_note(cfg.prefix, note_pl, None, cfg.dim)
        upsert_batch(client, notes_name, note_pts)

        chunks_name, chunk_pts = points_for_chunks(cfg.prefix, chunk_pls, vecs)
        upsert_batch(client, chunks_name, chunk_pts)

        edges_name, edge_pts = points_for_edges(cfg.prefix, edges)
        upsert_batch(client, edges_name, edge_pts)

    print(f"Done. Processed notes: {processed}")


if __name__ == "__main__":
    main()