Lars/mindnet
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases 21 Wiki Activity

modified: app/services/semantic_analyzer.py

Browse Source
This commit is contained in:
Lars 2025-12-12 09:42:57 +01:00
parent 37c0f526ec
commit 3629bc3fb9
1 changed files with 14 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

20
app/services/semantic_analyzer.py
View File

@ -48,7 +48,7 @@ class SemanticAnalyzer:
user_prompt = f"Dokument-Typ: {source_type}\n\nTEXT:\n{text}" user_prompt = f"Dokument-Typ: {source_type}\n\nTEXT:\n{text}"
try: try:
# WICHTIG: Erzwingt Ollama JSON Mode # 2. LLM Call (Async)
response_json = await self.llm.generate_raw_response( response_json = await self.llm.generate_raw_response(
user_prompt, user_prompt,
system=system_prompt, system=system_prompt,
@ -59,8 +59,18 @@ class SemanticAnalyzer:
clean_json = response_json.replace("```json", "").replace("```", "").strip() clean_json = response_json.replace("```json", "").replace("```", "").strip()
data = json.loads(clean_json) data = json.loads(clean_json)
# 3a. Typsicherheit des äußeren Arrays
if not isinstance(data, list):
logger.error("SemanticAnalyzer: JSON root ist kein Array. Fehlerhafte LLM-Antwort.")
raise ValueError("Root element is not a list.")
results = [] results = []
for item in data: for item in data:
# FIX: Typsicherheit auf Item-Ebene
if not isinstance(item, dict):
logger.warning(f"SemanticAnalyzer: Ungültiges Chunk-Element ignoriert: {item}")
continue
content = item.get("content", "").strip() content = item.get("content", "").strip()
if not content: continue if not content: continue
@ -68,7 +78,7 @@ class SemanticAnalyzer:
refined_edges = [] refined_edges = []
for rel in raw_rels: for rel in raw_rels:
# FIX: Sicherstellen, dass rel ein Dictionary ist, um 'str' object has no attribute 'get' zu verhindern # Typsicherheit auf Relation-Ebene
if not isinstance(rel, dict): if not isinstance(rel, dict):
logger.warning(f"SemanticAnalyzer: Ignoriere ungültige Relation: {rel}") logger.warning(f"SemanticAnalyzer: Ignoriere ungültige Relation: {rel}")
continue continue
@ -77,7 +87,7 @@ class SemanticAnalyzer:
raw_type = rel.get("type", "related_to") raw_type = rel.get("type", "related_to")
if target: if target:
# 1. Annahme: Hole den Typ der ZIEL-Entität aus dem Index (für Matrix-Logik) # 1. Typ-Auflösung (für Matrix)
target_entity_type = self._get_target_type_from_title(target) target_entity_type = self._get_target_type_from_title(target)
# 2. Matrix-Logik anwenden: # 2. Matrix-Logik anwenden:
@ -87,7 +97,6 @@ class SemanticAnalyzer:
if final_kind not in ["related_to", "references"] and target_entity_type != "concept": if final_kind not in ["related_to", "references"] and target_entity_type != "concept":
edge_str = f"{final_kind}:{target}" edge_str = f"{final_kind}:{target}"
else: else:
# Wenn Matrix oder LLM generisch war, nutzen wir den generischen Output des LLM.
edge_str = f"{raw_type}:{target}" edge_str = f"{raw_type}:{target}"
refined_edges.append(edge_str) refined_edges.append(edge_str)
@ -100,7 +109,7 @@ class SemanticAnalyzer:
logger.error("SemanticAnalyzer: LLM lieferte KEIN valides JSON. Fallback auf Raw Text.") logger.error("SemanticAnalyzer: LLM lieferte KEIN valides JSON. Fallback auf Raw Text.")
return [SemanticChunkResult(content=text, suggested_edges=[])] return [SemanticChunkResult(content=text, suggested_edges=[])]
except Exception as e: except Exception as e:
# Wichtig: Fehler im Loggen, damit wir wissen, warum es crashte (z.B. Timeout/Ressource) # Wichtig: Wir fangen alle anderen Fehler, um den Prozess nicht abzubrechen.
logger.error(f"SemanticAnalyzer Unbehandelter Fehler: {e}") logger.error(f"SemanticAnalyzer Unbehandelter Fehler: {e}")
return [SemanticChunkResult(content=text, suggested_edges=[])] return [SemanticChunkResult(content=text, suggested_edges=[])]
@ -119,7 +128,6 @@ class SemanticAnalyzer:
if "leitbild-rituale-system" in title_lower: if "leitbild-rituale-system" in title_lower:
return "concept" return "concept"
# Fallback
return "concept" return "concept"
async def close(self): async def close(self):