fix(python)!: use null for missing linear combination scores

When `LinearCombinationReranker` uses `return_score="all"` and either the
vector or full-text side has no rows, the unavailable score column is now a
typed Arrow `float32` null column instead of a column of NaNs. Existing scores
from the non-empty side and their descending `_relevance_score` ordering are
preserved.

Regression coverage exercises both missing sides through the hybrid query path.
It checks the missing column's `float32` type, null count and row count, exact
preservation of the original present-side values, and descending relevance
ordering. The existing empty-result, score-column override, and `return_all`
cases remain covered.

The exact head `db33a068263cd69ee578f95a375f0060dcbb4e21` passed all 15
applicable check runs and all 4 applicable workflow runs, with no commit
statuses. The hosted macOS Python run completed with 3019 passed, 126 skipped,
and 70 deselected; the two new cases and nine adjacent regression cases passed.
The retained CI collector reached the all-success snapshot but exited 1 while
writing its final metadata because of a harness `TypeError`; it did not rerun
or invalidate the product or CI results.

BREAKING CHANGE: For `return_score="all"` with an empty vector or full-text
side, the missing side's score column now contains typed nulls instead of NaN.
Consumers that distinguish these values will observe different results.

python/python/lancedb/rerankers/linear_combination.py

@@ -3,7 +3,6 @@
 
 
 from collections import defaultdict
-from numpy import nan
 import pyarrow as pa
 
 from .base import Reranker
@@ -66,7 +65,7 @@ class LinearCombinationReranker(Reranker):
             elif self.score == "all":
                 results = results.append_column(
                     "_distance",
-                    pa.array([nan] * len(fts_results), type=pa.float32()),
+                    pa.nulls(len(fts_results), type=pa.float32()),
                 )
             return results
 
@@ -87,7 +86,7 @@ class LinearCombinationReranker(Reranker):
             elif self.score == "all":
                 results = results.append_column(
                     "_score",
-                    pa.array([nan] * len(vector_results), type=pa.float32()),
+                    pa.nulls(len(vector_results), type=pa.float32()),
                 )
             return results
         results = defaultdict()

python/python/tests/test_rerankers.py

@@ -1150,6 +1150,52 @@ def test_linear_combination_missing_fts_is_penalised():
     )
 
 
+@pytest.mark.parametrize("empty_side", ["vector", "fts"])
+def test_linear_combination_empty_side_scores_are_null(empty_side):
+    """With return_score="all", the missing side's score column is null, not NaN."""
+    from lancedb.query import LanceHybridQueryBuilder
+
+    vector_results = pa.table(
+        {
+            "_rowid": pa.array([1, 2], pa.uint64()),
+            "_distance": pa.array([0.1, 0.2], pa.float32()),
+        }
+    )
+    fts_results = pa.table(
+        {
+            "_rowid": pa.array([1, 2], pa.uint64()),
+            "_score": pa.array([5.0, 4.0], pa.float32()),
+        }
+    )
+    if empty_side == "vector":
+        expected_present = fts_results["_score"].to_pylist()
+        vector_results = vector_results.slice(0, 0)
+        missing = "_distance"
+        present = "_score"
+    else:
+        expected_present = vector_results["_distance"].to_pylist()
+        fts_results = fts_results.slice(0, 0)
+        missing = "_score"
+        present = "_distance"
+
+    result = LanceHybridQueryBuilder._combine_hybrid_results(
+        fts_results=fts_results,
+        vector_results=vector_results,
+        norm="score",
+        fts_query="query",
+        reranker=LinearCombinationReranker(return_score="all"),
+        limit=10,
+        with_row_ids=True,
+    )
+
+    assert result[missing].type == pa.float32()
+    assert result[missing].null_count == len(result) == 2
+    assert result[present].to_pylist() == expected_present
+    assert result["_relevance_score"].to_pylist() == sorted(
+        result["_relevance_score"].to_pylist(), reverse=True
+    )
+
+
 @pytest.mark.skipif(
     os.environ.get("WATSONX_API_KEY") is None
     or (