docarray · JoanFM · Dec 20, 2022 · Dec 19, 2022 · Dec 19, 2022 · Dec 19, 2022
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -241,7 +241,7 @@ jobs:
           pytest --suppress-no-test-exit-code --cov=docarray --cov-report=xml \
             -v -s -m "not gpu" ${{ matrix.test-path }}
           echo "::set-output name=codecov_flag::docarray"
-        timeout-minutes: 60
+        timeout-minutes: 70
         env:
           JINA_AUTH_TOKEN: "${{ secrets.JINA_AUTH_TOKEN }}"
       - name: Check codecov file

diff --git a/docarray/array/storage/opensearch/find.py b/docarray/array/storage/opensearch/find.py
@@ -166,7 +166,7 @@ def _find(
         :return: DocumentArray containing the closest documents to the query if it is a single query, otherwise a list of DocumentArrays containing
            the closest Document objects for each of the queries in `query`.
         """
-        query = np.array(query).astype(np.float)
+        query = np.array(query).astype(np.float32)
         num_rows, n_dim = ndarray.get_array_rows(query)
         if n_dim != 2:
             query = query.reshape((num_rows, -1))

diff --git a/docarray/array/storage/redis/backend.py b/docarray/array/storage/redis/backend.py
@@ -27,6 +27,7 @@ class RedisConfig:
     redis_config: Dict[str, Any] = field(default_factory=dict)
     index_text: bool = field(default=False)
     tag_indices: List[str] = field(default_factory=list)
+    language: Optional[str] = None
     batch_size: int = field(default=64)
     method: str = field(default='HNSW')
     ef_construction: Optional[int] = None
@@ -113,7 +114,9 @@ def _build_index(self, rebuild: bool = False):
                 self._client.ft(index_name=self._config.index_name).dropindex()
 
             schema = self._build_schema_from_redis_config()
-            idef = IndexDefinition(prefix=[self._doc_prefix])
+            idef = IndexDefinition(
+                prefix=[self._doc_prefix], language=self._config.language
+            )
             self._client.ft(index_name=self._config.index_name).create_index(
                 schema, definition=idef
             )

diff --git a/docs/advanced/document-store/redis.md b/docs/advanced/document-store/redis.md
@@ -130,6 +130,7 @@ The following configs can be set:
 | `method`          | Vector similarity index algorithm in Redis, either `FLAT` or `HNSW`                                                                    | `'HNSW'`                                          |
 | `index_text`      | Boolean flag indicating whether to index `.text`. `True` will enable full text search on `.text`                                       | `None`                                            |
 | `tag_indices`     | List of tags to index as text field                                                                                                    | `[]`                                              |
+| `language`        | Optional parameter for Redis text search. Refer to the [list of supported languages](https://redis.io/docs/stack/search/reference/stemming/) | `None`                                      |
 | `ef_construction` | Optional parameter for Redis HNSW algorithm                                                                                            | `200`                                             |
 | `m`               | Optional parameter for Redis HNSW algorithm                                                                                            | `16`                                              |
 | `ef_runtime`      | Optional parameter for Redis HNSW algorithm                                                                                            | `10`                                              |

diff --git a/tests/unit/array/mixins/test_find.py b/tests/unit/array/mixins/test_find.py
@@ -986,6 +986,35 @@ def test_redis_geo_filter(start_storage):
         assert distance[0][1] < 800
 
 
+def test_redis_language(start_storage):
+    n_dim = 128
+    da = DocumentArray(
+        storage='redis',
+        config={
+            'n_dim': n_dim,
+            'index_text': True,
+            'language': 'chinese',
+        },
+    )
+
+    with da:
+        da.extend(
+            [
+                Document(id='1', text='意大利和西班牙 token1 token2 token3'),
+                Document(id='2', text='法国和中国 token1 token2'),
+                Document(id='3', text='意大利和法国 token2 token3 token4'),
+            ]
+        )
+
+    results = da.find('token1')
+    assert len(results) == 2
+    assert set(results[:, 'id']) == {'1', '2'}
+
+    results = da.find('意大利')
+    assert len(results) == 2
+    assert set(results[:, 'id']) == {'1', '3'}
+
+
 @pytest.mark.parametrize('storage', ['memory'])
 @pytest.mark.parametrize('columns', [[('price', 'int')], {'price': 'int'}])
 def test_unsupported_pre_filtering(storage, start_storage, columns):

diff --git a/tests/unit/array/test_advance_indexing.py b/tests/unit/array/test_advance_indexing.py
@@ -43,7 +43,7 @@ def test_getter_int_str(docs, storage, config, start_storage):
         docs = DocumentArray(docs, storage=storage)
     # getter
     assert docs[99].text == "99"
-    assert docs[np.int(99)].text == "99"
+    assert docs[np.int32(99)].text == "99"
     assert docs[-1].text == "99"
     assert docs[0].text == "0"
     # string index