From eafa0fad859c38e66f1eea451d719005f88a5d47 Mon Sep 17 00:00:00 2001
From: Lars <Lars@stommer.de>
Date: Sun, 16 Nov 2025 17:36:09 +0100
Subject: [PATCH] Dateien nach "tests" hochladen

---
 tests/diag_payload_indexes.py    | 73 +++++++++++++-------------------
 tests/ensure_indexes_and_show.py | 18 +++-----
 2 files changed, 36 insertions(+), 55 deletions(-)

diff --git a/tests/diag_payload_indexes.py b/tests/diag_payload_indexes.py
index 204b891..1418e62 100644
--- a/tests/diag_payload_indexes.py
+++ b/tests/diag_payload_indexes.py
@@ -1,56 +1,31 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 """
-scripts/diag_payload_indexes.py
-Listet die Payload-Indizes (payload_schema) für alle Mindnet-Collections.
+scripts/diag_payload_indexes.py  (v1.1)
 
-Nutzt client.get_collection(name) und zeigt:
-- payload_schema (Feld -> Schema/Typ)
-- Vector-Konfiguration (size, distance, named vectors)
-- Segment-Infos (optional komprimiert)
-
-Aufruf:
-  python3 -m scripts.diag_payload_indexes
-  python3 -m scripts.diag_payload_indexes --raw   # vollständiges JSON
-
-Hinweis: In Qdrant werden die Payload-Indizes im Collection-Info-Endpunkt unter
-`result.payload_schema` geliefert.
+Zeigt payload_schema (Indizes) je Collection.
+WICHTIG: Einige Qdrant-Versionen liefern payload_schema nur, wenn
+`with_payload_schema=true` gesetzt wird. Daher setzen wir das Flag explizit.
 """
 from __future__ import annotations
 import argparse, json
 from qdrant_client.http import models as rest
 from app.core.qdrant import QdrantConfig, get_client, collection_names
 
-def compact_schema(ps: dict) -> dict:
+def compact_schema(ps: dict | None) -> dict:
+    if not isinstance(ps, dict):
+        return {}
     out = {}
-    for k, v in (ps or {}).items():
-        # v ist z.B. {"type":"keyword"} oder {"type":"text"} etc.
-        if isinstance(v, dict):
-            out[k] = v.get("type") or v
+    for k, v in ps.items():
+        if isinstance(v, dict) and "type" in v:
+            out[k] = v["type"]
         else:
             out[k] = v
     return out
 
-def get_info(client, name: str) -> dict:
-    info = client.get_collection(collection_name=name)
-    # Das SDK liefert .dict(). Wir extrahieren das Wesentliche.
-    d = info.dict() if hasattr(info, "dict") else info
-    result = d.get("result", {})
-    vectors = result.get("vectors") or {}
-    if isinstance(vectors, dict) and "config" in vectors:
-        vectors = vectors.get("config")
-    payload_schema = result.get("payload_schema") or {}
-    segments = result.get("segments_count")
-    return {
-        "name": name,
-        "vectors": vectors,
-        "payload_schema": compact_schema(payload_schema),
-        "segments_count": segments,
-    }
-
 def main():
     ap = argparse.ArgumentParser()
-    ap.add_argument("--raw", action="store_true", help="Gibt das rohe get_collection-Result zurück")
+    ap.add_argument("--raw", action="store_true")
     args = ap.parse_args()
 
     cfg = QdrantConfig.from_env()
@@ -58,14 +33,24 @@ def main():
     notes, chunks, edges = collection_names(cfg.prefix)
 
     cols = [notes, chunks, edges]
-    if args.raw:
-        for name in cols:
-            info = client.get_collection(collection_name=name)
-            d = info.dict() if hasattr(info, "dict") else info
-            print(json.dumps({"collection": name, "raw": d}, ensure_ascii=False))
-    else:
-        out = [ get_info(client, name) for name in cols ]
-        print(json.dumps({"prefix": cfg.prefix, "collections": out}, ensure_ascii=False, indent=2))
+    result = []
+    for name in cols:
+        info = client.get_collection(collection_name=name, with_payload_schema=True)
+        d = info.model_dump() if hasattr(info, "model_dump") else (info.dict() if hasattr(info, "dict") else info)
+        payload_schema = (d.get("result") or {}).get("payload_schema")
+        vectors = (d.get("result") or {}).get("vectors")
+        if isinstance(vectors, dict) and "config" in vectors:
+            vectors = vectors.get("config")
+        if args.raw:
+            result.append({"collection": name, "raw": d})
+        else:
+            result.append({
+                "name": name,
+                "vectors": vectors,
+                "payload_schema": compact_schema(payload_schema),
+                "segments_count": (d.get("result") or {}).get("segments_count"),
+            })
+    print(json.dumps({"prefix": cfg.prefix, "collections": result}, ensure_ascii=False, indent=2))
 
 if __name__ == "__main__":
     main()
diff --git a/tests/ensure_indexes_and_show.py b/tests/ensure_indexes_and_show.py
index 3b9aa8e..a8183d5 100644
--- a/tests/ensure_indexes_and_show.py
+++ b/tests/ensure_indexes_and_show.py
@@ -1,11 +1,9 @@
 #!/usr/bin/env python3
 # -*- coding: utf-8 -*-
 """
-scripts/ensure_indexes_and_show.py
-Erzwingt/prüft die Index-Erstellung via ensure_payload_indexes(...) und zeigt danach das payload_schema.
-
-Aufruf:
-  python3 -m scripts.ensure_indexes_and_show
+tests/ensure_indexes_and_show.py  (v1.0)
+Alias für die Test-Suite: ruft ensure_payload_indexes(...) auf
+und zeigt danach das payload_schema (mit with_payload_schema=True).
 """
 from __future__ import annotations
 import json
@@ -15,14 +13,12 @@ def main():
     cfg = QdrantConfig.from_env()
     client = get_client(cfg)
     ensure_payload_indexes(client, cfg.prefix)
-    # Danach anzeigen
     notes, chunks, edges = collection_names(cfg.prefix)
-    cols = [notes, chunks, edges]
     res = {}
-    for name in cols:
-        info = client.get_collection(collection_name=name)
-        d = info.dict() if hasattr(info, "dict") else info
-        res[name] = d.get("result", {}).get("payload_schema")
+    for name in (notes, chunks, edges):
+        info = client.get_collection(collection_name=name, with_payload_schema=True)
+        d = info.model_dump() if hasattr(info, "model_dump") else (info.dict() if hasattr(info, "dict") else info)
+        res[name] = (d.get("result") or {}).get("payload_schema")
     print(json.dumps(res, ensure_ascii=False, indent=2))
 
 if __name__ == "__main__":