Better handle empty results from tantivy (#155)

Closes #154

---------

Co-authored-by: Chang She <chang@lancedb.com>

python/lancedb/fts.py

@@ -118,6 +118,8 @@ def search_index(
     query = index.parse_query(query)
     # get top results
     results = searcher.search(query, limit)
+    if results.count == 0:
+        return tuple(), tuple()
     return tuple(
         zip(
             *[

python/lancedb/query.py

@@ -164,6 +164,8 @@ class LanceFtsQueryBuilder(LanceQueryBuilder):
         index = tantivy.Index.open(index_path)
         # get the scores and doc ids
         row_ids, scores = search_index(index, self._query, self._limit)
+        if len(row_ids) == 0:
+            return pd.DataFrame()
         scores = pa.array(scores)
         output_tbl = self._table.to_lance().take(row_ids, columns=self._columns)
         output_tbl = output_tbl.append_column("score", scores)

python/pyproject.toml

@@ -37,7 +37,7 @@ repository = "https://github.com/lancedb/lancedb"
 
 [project.optional-dependencies]
 tests = [
-    "pytest"
+    "pytest", "pytest-mock"
 ]
 dev = [
     "ruff", "pre-commit", "black"

python/tests/test_fts.py

@@ -82,3 +82,10 @@ def test_create_index_multiple_columns(tmp_path, table):
     assert len(df) == 10
     assert "text" in df.columns
     assert "text2" in df.columns
+
+
+def test_empty_rs(tmp_path, table, mocker):
+    table.create_fts_index(["text", "text2"])
+    mocker.patch("lancedb.fts.search_index", return_value=([], []))
+    df = table.search("puppy").limit(10).to_df()
+    assert len(df) == 0