feat: support vector search with distance thresholds (#1993)

Signed-off-by: BubbleCal <bubble-cal@outlook.com>

python/python/tests/test_query.py

@@ -94,6 +94,73 @@ def test_with_row_id(table: lancedb.table.Table):
     assert rs["_rowid"].to_pylist() == [0, 1]
 
 
+def test_distance_range(table: lancedb.table.Table):
+    q = [0, 0]
+    rs = table.search(q).to_arrow()
+    dists = rs["_distance"].to_pylist()
+    min_dist = dists[0]
+    max_dist = dists[-1]
+
+    res = table.search(q).distance_range(upper_bound=min_dist).to_arrow()
+    assert len(res) == 0
+
+    res = table.search(q).distance_range(lower_bound=max_dist).to_arrow()
+    assert len(res) == 1
+    assert res["_distance"].to_pylist() == [max_dist]
+
+    res = table.search(q).distance_range(upper_bound=max_dist).to_arrow()
+    assert len(res) == 1
+    assert res["_distance"].to_pylist() == [min_dist]
+
+    res = table.search(q).distance_range(lower_bound=min_dist).to_arrow()
+    assert len(res) == 2
+    assert res["_distance"].to_pylist() == [min_dist, max_dist]
+
+
+@pytest.mark.asyncio
+async def test_distance_range_async(table_async: AsyncTable):
+    q = [0, 0]
+    rs = await table_async.query().nearest_to(q).to_arrow()
+    dists = rs["_distance"].to_pylist()
+    min_dist = dists[0]
+    max_dist = dists[-1]
+
+    res = (
+        await table_async.query()
+        .nearest_to(q)
+        .distance_range(upper_bound=min_dist)
+        .to_arrow()
+    )
+    assert len(res) == 0
+
+    res = (
+        await table_async.query()
+        .nearest_to(q)
+        .distance_range(lower_bound=max_dist)
+        .to_arrow()
+    )
+    assert len(res) == 1
+    assert res["_distance"].to_pylist() == [max_dist]
+
+    res = (
+        await table_async.query()
+        .nearest_to(q)
+        .distance_range(upper_bound=max_dist)
+        .to_arrow()
+    )
+    assert len(res) == 1
+    assert res["_distance"].to_pylist() == [min_dist]
+
+    res = (
+        await table_async.query()
+        .nearest_to(q)
+        .distance_range(lower_bound=min_dist)
+        .to_arrow()
+    )
+    assert len(res) == 2
+    assert res["_distance"].to_pylist() == [min_dist, max_dist]
+
+
 def test_vector_query_with_no_limit(table):
     with pytest.raises(ValueError):
         LanceVectorQueryBuilder(table, [0, 0], "vector").limit(0).select(

python/python/tests/test_remote_db.py

@@ -306,6 +306,8 @@ def test_query_sync_minimal():
             "k": 10,
             "prefilter": False,
             "refine_factor": None,
+            "lower_bound": None,
+            "upper_bound": None,
             "ef": None,
             "vector": [1.0, 2.0, 3.0],
             "nprobes": 20,
@@ -348,6 +350,8 @@ def test_query_sync_maximal():
             "refine_factor": 10,
             "vector": [1.0, 2.0, 3.0],
             "nprobes": 5,
+            "lower_bound": None,
+            "upper_bound": None,
             "ef": None,
             "filter": "id > 0",
             "columns": ["id", "name"],
@@ -449,6 +453,8 @@ def test_query_sync_hybrid():
                 "refine_factor": None,
                 "vector": [0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
                 "nprobes": 20,
+                "lower_bound": None,
+                "upper_bound": None,
                 "ef": None,
                 "with_row_id": True,
                 "version": None,