mindnet/app/core/ingestion/ingestion_processor.py

"""
FILE: app/core/ingestion/ingestion_processor.py
DESCRIPTION: Der zentrale IngestionService (Orchestrator). 
             WP-24c: Integration der Symmetrie-Logik (Automatische inverse Kanten).
             WP-25a: Integration der Mixture of Experts (MoE) Architektur.
             WP-15b: Two-Pass Workflow mit globalem Kontext-Cache.
             WP-20/22: Cloud-Resilienz und Content-Lifecycle integriert.
             AUDIT v3.3.6: Strikte Phasentrennung (Phase 2 global am Ende).
                          Fix für .trash-Folder und Pydantic 'None'-Crash.
                          Vollständige Wiederherstellung des Business-Loggings.
VERSION: 3.3.6 (WP-24c: Full Transparency Orchestration)
STATUS: Active
"""
import logging
import asyncio
import os
import re
from typing import Dict, List, Optional, Tuple, Any

# Core Module Imports
from app.core.parser import (
    read_markdown, pre_scan_markdown, normalize_frontmatter, 
    validate_required_frontmatter, NoteContext
)
from app.core.chunking import assemble_chunks
# WP-24c: Import für die deterministische UUID-Vorabberechnung
from app.core.graph.graph_utils import _mk_edge_id 

# Datenbank-Ebene (Modularisierte database-Infrastruktur)
from app.core.database.qdrant import QdrantConfig, get_client, ensure_collections, ensure_payload_indexes
from app.core.database.qdrant_points import points_for_chunks, points_for_note, points_for_edges, upsert_batch
from qdrant_client.http import models as rest 

# Services
from app.services.embeddings_client import EmbeddingsClient
from app.services.edge_registry import registry as edge_registry
from app.services.llm_service import LLMService 

# Package-Interne Imports
from .ingestion_utils import load_type_registry, resolve_note_type, get_chunk_config_by_profile
from .ingestion_db import fetch_note_payload, artifacts_missing, purge_artifacts, is_explicit_edge_present
from .ingestion_validation import validate_edge_candidate
from .ingestion_note_payload import make_note_payload
from .ingestion_chunk_payload import make_chunk_payloads

# Fallback für Edges (Struktur-Verknüpfung)
try:
    from app.core.graph.graph_derive_edges import build_edges_for_note
except ImportError:
    def build_edges_for_note(*args, **kwargs): return []

logger = logging.getLogger(__name__)

class IngestionService:
    def __init__(self, collection_prefix: str = None):
        """Initialisiert den Service und bereinigt das technische Logging."""
        from app.config import get_settings
        self.settings = get_settings()
        
        # --- LOGGING CLEANUP (Business Focus) ---
        # Unterdrückt HTTP-Bibliotheks-Lärm, erhält aber inhaltliche Service-Logs
        for lib in ["httpx", "httpcore", "qdrant_client", "urllib3", "openai"]:
            logging.getLogger(lib).setLevel(logging.WARNING)

        self.prefix = collection_prefix or self.settings.COLLECTION_PREFIX
        self.cfg = QdrantConfig.from_env()
        self.cfg.prefix = self.prefix 
        self.client = get_client(self.cfg)
        
        self.registry = load_type_registry()
        self.embedder = EmbeddingsClient()
        self.llm = LLMService() 
        
        # WP-25a: Dimensionen über das LLM-Profil auflösen
        embed_cfg = self.llm.profiles.get("embedding_expert", {})
        self.dim = embed_cfg.get("dimensions") or self.settings.VECTOR_SIZE
        
        # Festlegen des Change-Detection Modus
        self.active_hash_mode = self.settings.CHANGE_DETECTION_MODE
        
        # WP-15b: Kontext-Gedächtnis für ID-Auflösung (Global)
        self.batch_cache: Dict[str, NoteContext] = {}
        
        # WP-24c: Puffer für Phase 2 (Symmetrie-Injektion am Ende des gesamten Imports)
        self.symmetry_buffer: List[Dict[str, Any]] = []

        try:
            # Schema-Prüfung und Initialisierung
            ensure_collections(self.client, self.prefix, self.dim)
            ensure_payload_indexes(self.client, self.prefix)
        except Exception as e: 
            logger.warning(f"DB initialization warning: {e}")

    def _is_valid_note_id(self, text: str) -> bool:
        """
        WP-24c: Prüft Ziel-Strings auf fachliche Validität.
        Verhindert Müll-Kanten zu reinen System-Platzhaltern.
        """
        if not text or len(text.strip()) < 2: return False
        blacklisted = {"insight", "event", "source", "task", "project", "person", "concept", "related_to", "referenced_by"}
        if text.lower().strip() in blacklisted: return False
        if len(text) > 200: return False 
        return True

    async def run_batch(self, file_paths: List[str], vault_root: str) -> Dict[str, Any]:
        """
        WP-15b: Two-Pass Ingestion Workflow (PHASE 1).
        Füllt den Cache und verarbeitet Dateien batchweise.
        Gibt ein Dictionary zurück, um Kompatibilität zum Orchestrator zu wahren.
        """
        self.batch_cache.clear()
        logger.info(f"--- 🔍 START BATCH PHASE 1 ({len(file_paths)} Dateien) ---")

        # 1. Schritt: Pre-Scan (Context-Cache befüllen)
        for path in file_paths:
            try:
                ctx = pre_scan_markdown(path, registry=self.registry)
                if ctx:
                    self.batch_cache[ctx.note_id] = ctx
                    self.batch_cache[ctx.title] = ctx
                    fname = os.path.splitext(os.path.basename(path))[0]
                    self.batch_cache[fname] = ctx
            except Exception as e:
                logger.warning(f"  ⚠️ Pre-scan fehlgeschlagen für {path}: {e}")

        # 2. Schritt: Batch-Verarbeitung (Authority Only)
        processed_count = 0
        success_count = 0
        for p in file_paths:
            processed_count += 1
            res = await self.process_file(p, vault_root, apply=True, purge_before=True)
            if res.get("status") == "success":
                success_count += 1
            
        logger.info(f"--- ✅ Batch Phase 1 abgeschlossen ({success_count}/{processed_count}) ---")
        return {
            "status": "success", 
            "processed": processed_count, 
            "success": success_count,
            "buffered_virtuals": len(self.symmetry_buffer)
        }

    async def commit_vault_symmetries(self) -> Dict[str, Any]:
        """
        WP-24c: Führt PHASE 2 (Symmetrie-Injektion) für den gesamten Vault aus.
        Wird nach Abschluss aller Batches einmalig aufgerufen.
        Vergleicht gepufferte Kanten gegen die Instance-of-Truth in Qdrant.
        """
        if not self.symmetry_buffer:
            logger.info("⏭️ Symmetrie-Puffer ist leer. Keine Aktion erforderlich.")
            return {"status": "skipped", "reason": "buffer_empty"}

        logger.info(f"🔄 PHASE 2: Validiere {len(self.symmetry_buffer)} Symmetrie-Kanten gegen die Instance-of-Truth...")
        final_virtuals = []
        for v_edge in self.symmetry_buffer:
            # Deterministische ID der potenziellen Symmetrie berechnen
            v_id = _mk_edge_id(v_edge["kind"], v_edge["note_id"], v_edge["target_id"], v_edge.get("scope", "note"))
            
            # AUTHORITY-CHECK: Nur schreiben, wenn KEINE manuelle Kante in der DB existiert
            if not is_explicit_edge_present(self.client, self.prefix, v_id):
                final_virtuals.append(v_edge)
                # Detailliertes Logging für volle Transparenz
                logger.info(f"  🔄 [SYMMETRY] Add inverse: {v_edge['note_id']} --({v_edge['kind']})--> {v_edge['target_id']}")
            else:
                logger.debug(f"  🛡️ Schutz: Manuelle Kante belegt ID {v_id}. Symmetrie verworfen.")

        added_count = 0
        if final_virtuals:
            logger.info(f"📤 Schreibe {len(final_virtuals)} validierte Symmetrie-Kanten in den Graphen.")
            e_pts = points_for_edges(self.prefix, final_virtuals)[1]
            upsert_batch(self.client, f"{self.prefix}_edges", e_pts)
            added_count = len(final_virtuals)

        self.symmetry_buffer.clear() # Puffer nach erfolgreichem Commit leeren
        return {"status": "success", "added": added_count}

    async def process_file(self, file_path: str, vault_root: str, **kwargs) -> Dict[str, Any]:
        """
        Transformiert eine Markdown-Datei in Phase 1 (Authority First).
        Implementiert Ordner-Blacklists, Pydantic-Safety und MoE-Validierung.
        """
        apply = kwargs.get("apply", False)
        force_replace = kwargs.get("force_replace", False)
        purge_before = kwargs.get("purge_before", False)
        
        result = {"path": file_path, "status": "skipped", "changed": False, "error": None}

        try:
            # --- ORDNER-FILTER (Fix für .trash und .obsidian Junk) ---
            if any(part.startswith('.') for part in file_path.split(os.sep)):
                return {**result, "status": "skipped", "reason": "hidden_folder"}

            ingest_cfg = self.registry.get("ingestion_settings", {})
            ignore_folders = ingest_cfg.get("ignore_folders", [".trash", ".obsidian", "templates"])
            if any(folder in file_path for folder in ignore_folders):
                return {**result, "status": "skipped", "reason": "folder_blacklist"}

            # Datei einlesen und validieren
            parsed = read_markdown(file_path)
            if not parsed: return {**result, "error": "Empty file"}
            fm = normalize_frontmatter(parsed.frontmatter)
            validate_required_frontmatter(fm)
            
            note_type = resolve_note_type(self.registry, fm.get("type"))
            note_pl = make_note_payload(parsed, vault_root=vault_root, file_path=file_path, types_cfg=self.registry)
            note_id = note_pl.get("note_id")

            # --- FIX: Guard Clause gegen 'None' IDs (Verhindert Pydantic Crash)  ---
            if not note_id:
                logger.warning(f"  ⚠️ Fehlende note_id in '{file_path}'. Datei wird ignoriert.")
                return {**result, "status": "error", "error": "missing_note_id"}

            logger.info(f"📄 Bearbeite: '{note_id}' (Typ: {note_type})")

            # Change Detection & Fragment-Prüfung
            old_payload = None if force_replace else fetch_note_payload(self.client, self.prefix, note_id)
            c_miss, e_miss = artifacts_missing(self.client, self.prefix, note_id)
            if not (force_replace or not old_payload or c_miss or e_miss):
                return {**result, "status": "unchanged", "note_id": note_id}

            if not apply: 
                return {**result, "status": "dry-run", "changed": True, "note_id": note_id}

            # Chunks erzeugen und semantisch validieren (MoE)
            profile = note_pl.get("chunk_profile", "sliding_standard")
            chunk_cfg = get_chunk_config_by_profile(self.registry, profile, note_type)
            enable_smart = chunk_cfg.get("enable_smart_edge_allocation", False)
            chunks = await assemble_chunks(note_id, getattr(parsed, "body", ""), note_type, config=chunk_cfg)
            
            for ch in chunks:
                new_pool = []
                for cand in getattr(ch, "candidate_pool", []):
                    if cand.get("provenance") == "global_pool" and enable_smart:
                        # Detailliertes Business-Logging für LLM-Aktivitäten
                        target_label = cand.get('target_id') or cand.get('note_id') or "Unknown"
                        logger.info(f"  ⚖️ [VALIDATING] Relation to '{target_label}' via Expert-LLM...")
                        
                        is_valid = await validate_edge_candidate(ch.text, cand, self.batch_cache, self.llm)
                        
                        logger.info(f"  🧠 [SMART EDGE] {target_label} -> {'✅ OK' if is_valid else '❌ SKIP'}")
                        if is_valid: new_pool.append(cand)
                    else: 
                        new_pool.append(cand)
                ch.candidate_pool = new_pool

            # Embeddings und Payloads
            chunk_pls = make_chunk_payloads(fm, note_pl["path"], chunks, file_path=file_path, types_cfg=self.registry)
            vecs = await self.embedder.embed_documents([c.get("window") or "" for c in chunk_pls]) if chunk_pls else []
            
            # Kanten-Extraktion mit ID-Kanonisierung
            raw_edges = build_edges_for_note(note_id, chunk_pls, note_level_references=note_pl.get("references", []))
            
            explicit_edges = []
            for e in raw_edges:
                target_raw = e.get("target_id")
                # Auflösung von Titeln/Dateinamen zu echten IDs über den globalen Cache
                target_ctx = self.batch_cache.get(target_raw)
                target_id = target_ctx.note_id if target_ctx else target_raw
                
                if not self._is_valid_note_id(target_id): continue
                
                resolved_kind = edge_registry.resolve(e.get("kind", "related_to"), provenance=e.get("provenance", "explicit"))
                
                # Echte physische Kante markieren (Phase 1 Autorität)
                e.update({
                    "kind": resolved_kind, "target_id": target_id, 
                    "origin_note_id": note_id, "virtual": False, "confidence": 1.0
                })
                explicit_edges.append(e)
                
                # Symmetrie-Kandidat für die globale Phase 2 puffern
                inv_kind = edge_registry.get_inverse(resolved_kind)
                if inv_kind and target_id != note_id:
                    v_edge = e.copy()
                    v_edge.update({
                        "note_id": target_id, "target_id": note_id, "kind": inv_kind,
                        "virtual": True, "provenance": "structure", "confidence": 1.0,
                        "origin_note_id": note_id 
                    })
                    self.symmetry_buffer.append(v_edge)

            # 4. DB Upsert (Phase 1: Authority Commitment)
            if purge_before and old_payload: purge_artifacts(self.client, self.prefix, note_id)
            
            n_name, n_pts = points_for_note(self.prefix, note_pl, None, self.dim)
            upsert_batch(self.client, n_name, n_pts)
            if chunk_pls and vecs: 
                upsert_batch(self.client, f"{self.prefix}_chunks", points_for_chunks(self.prefix, chunk_pls, vecs)[1])
            if explicit_edges: 
                upsert_batch(self.client, f"{self.prefix}_edges", points_for_edges(self.prefix, explicit_edges)[1])
            
            logger.info(f"  ✨ Phase 1 fertig: {len(chunk_pls)} Chunks, {len(explicit_edges)} explizite Kanten.")
            return {"status": "success", "note_id": note_id, "edges_count": len(explicit_edges)}
            
        except Exception as e:
            logger.error(f"❌ Fehler bei {file_path}: {e}", exc_info=True)
            return {**result, "status": "error", "error": str(e)}

    async def create_from_text(self, markdown_content: str, filename: str, vault_root: str, folder: str = "00_Inbox") -> Dict[str, Any]:
        """Erstellt eine Note aus einem Textstream."""
        target_path = os.path.join(vault_root, folder, filename)
        os.makedirs(os.path.dirname(target_path), exist_ok=True)
        with open(target_path, "w", encoding="utf-8") as f: 
            f.write(markdown_content)
        await asyncio.sleep(0.1) 
        return await self.process_file(file_path=target_path, vault_root=vault_root, apply=True, force_replace=True, purge_before=True)