From 1d962230394dd4a28d315d58e04746901d9c213a Mon Sep 17 00:00:00 2001
From: tatavishnurao <tatavishnurao@gmail.com>
Date: Tue, 19 May 2026 18:19:59 +0530
Subject: [PATCH] fix: align local_knowledge_search notebook with sochdb SDK
 v0.5.7 API

Replace stale API references in 0_local_knowledge_search_walkthrough.ipynb:

- HnswIndex -> VectorIndex
- m= -> max_connections=
- metric= kwarg removed (not a constructor param)
- insert_batch_with_ids() -> insert_batch()
- search() returns List[Tuple[int,float]], not two arrays
---
 0_local_knowledge_search_walkthrough.ipynb | 24 ++++++++++++----------
 1 file changed, 13 insertions(+), 11 deletions(-)

diff --git a/0_local_knowledge_search_walkthrough.ipynb b/0_local_knowledge_search_walkthrough.ipynb
index 65a4f1e..c60f849 100644
--- a/0_local_knowledge_search_walkthrough.ipynb
+++ b/0_local_knowledge_search_walkthrough.ipynb
@@ -40,7 +40,7 @@
     "from pathlib import Path\n",
     "\n",
     "import numpy as np\n",
-    "from sochdb import Database, HnswIndex"
+    "from sochdb import Database, VectorIndex"
    ]
   },
   {
@@ -198,9 +198,9 @@
      ]
     }
    ],
-   "source": [
-    "index = HnswIndex(dimension=DIMENSION, m=16, ef_construction=100, metric='cosine')\n",
-    "inserted = index.insert_batch_with_ids(ids, vectors)\n",
+"source": [
+     "index = VectorIndex(dimension=DIMENSION, max_connections=16, ef_construction=100)\n",
+     "inserted = index.insert_batch(ids, vectors)\n",
     "print(f'Indexed {inserted} document embeddings')"
    ]
   },
@@ -247,13 +247,15 @@
      "output_type": "execute_result"
     }
    ],
-   "source": [
-    "query = 'How do I access internal tools securely from my laptop?'\n",
-    "query_vec = embed_text(query)\n",
-    "result_ids, distances = index.search(query_vec, k=3)\n",
-    "\n",
-    "results = []\n",
-    "for doc_id, distance in zip(result_ids.tolist(), distances.tolist()):\n",
+"source": [
+     "query = 'How do I access internal tools securely from my laptop?'\n",
+     "query_vec = embed_text(query)\n",
+     "results = index.search(query_vec, k=3)\n",
+     "result_ids = [r[0] for r in results]\n",
+     "distances = [r[1] for r in results]\n",
+     "\n",
+     "results = []\n",
+     "for doc_id, distance in zip(result_ids, distances):\n",
     "    payload = db.get(f'docs/{doc_id}'.encode('utf-8'))\n",
     "    if payload is None:\n",
     "        continue\n",