docarray · JohannesMessner · Nov 22, 2022 · Aug 26, 2022 · Aug 30, 2022 · Oct 4, 2022
diff --git a/.github/workflows/ci.yml b/.github/workflows/ci.yml
@@ -187,7 +187,7 @@ jobs:
           pytest --suppress-no-test-exit-code --cov=docarray --cov-report=xml \
             -v -s -m "not gpu" ${{ matrix.test-path }}
           echo "codecov_flag=docarray" >> $GITHUB_OUTPUT
-        timeout-minutes: 45
+        timeout-minutes: 60
         env:
           JINA_AUTH_TOKEN: "${{ secrets.JINA_AUTH_TOKEN }}"
       - name: Check codecov file
@@ -238,7 +238,7 @@ jobs:
           pytest --suppress-no-test-exit-code --cov=docarray --cov-report=xml \
             -v -s -m "not gpu" ${{ matrix.test-path }}
           echo "::set-output name=codecov_flag::docarray"
-        timeout-minutes: 40
+        timeout-minutes: 60
         env:
           JINA_AUTH_TOKEN: "${{ secrets.JINA_AUTH_TOKEN }}"
       - name: Check codecov file

diff --git a/docarray/array/document.py b/docarray/array/document.py
@@ -12,11 +12,13 @@
     from docarray.array.weaviate import DocumentArrayWeaviate
     from docarray.array.elastic import DocumentArrayElastic
     from docarray.array.redis import DocumentArrayRedis
+    from docarray.array.milvus import DocumentArrayMilvus
     from docarray.array.storage.sqlite import SqliteConfig
     from docarray.array.storage.annlite import AnnliteConfig
     from docarray.array.storage.weaviate import WeaviateConfig
     from docarray.array.storage.elastic import ElasticConfig
     from docarray.array.storage.redis import RedisConfig
+    from docarray.array.storage.milvus import MilvusConfig
 
 
 class DocumentArray(AllMixins, BaseDocumentArray):
@@ -140,6 +142,16 @@ def __new__(
         """Create a Redis-powered DocumentArray object."""
         ...
 
+    @overload
+    def __new__(
+        cls,
+        _docs: Optional['DocumentArraySourceType'] = None,
+        storage: str = 'milvus',
+        config: Optional[Union['MilvusConfig', Dict]] = None,
+    ) -> 'DocumentArrayMilvus':
+        """Create a Milvus-powered DocumentArray object."""
+        ...
+
     def __enter__(self):
         self._exit_stack = ExitStack()
         # Ensure that we sync the data to the storage backend when exiting the context manager
@@ -184,6 +196,10 @@ def __new__(cls, *args, storage: str = 'memory', **kwargs):
                 from .redis import DocumentArrayRedis
 
                 instance = super().__new__(DocumentArrayRedis)
+            elif storage == 'milvus':
+                from .milvus import DocumentArrayMilvus
+
+                instance = super().__new__(DocumentArrayMilvus)
 
             else:
                 raise ValueError(f'storage=`{storage}` is not supported.')

diff --git a/docarray/array/milvus.py b/docarray/array/milvus.py
@@ -0,0 +1,46 @@
+from .document import DocumentArray
+
+from .storage.milvus import StorageMixins, MilvusConfig
+
+__all__ = ['MilvusConfig', 'DocumentArrayMilvus']
+
+
+class DocumentArrayMilvus(StorageMixins, DocumentArray):
+    """
+    DocumentArray that stores Documents in a `Milvus <https://milvus.io//>`_ vector search engine.
+
+    .. note::
+        This DocumentArray requires `pymilvus`. You can install it via `pip install "docarray[milvus]"`.
+
+        To use Milvus as storage backend, a Milvus service needs to be running on your machine.
+
+    With this implementation, :meth:`match` and :meth:`find` perform fast (approximate) vector search.
+    Additionally, search with filters is supported.
+
+    Example usage:
+
+    .. code-block:: python
+
+        from docarray import DocumentArray
+
+        # connect to running Milvus service with default configuration (address: http://localhost:19530)
+        da = DocumentArray(storage='milvus', config={'n_dim': 10})
+
+        # connect to a previously persisted DocumentArrayMilvus by specifying collection_name, host, and port
+        da = DocumentArray(
+            storage='milvus',
+            config={
+                'collection_name': 'persisted',
+                'host': 'localhost',
+                'port': '19530',
+                'n_dim': 10,
+            },
+        )
+
+
+    .. seealso::
+        For further details, see our :ref:`user guide <milvus>`.
+    """
+
+    def __new__(cls, *args, **kwargs):
+        return super().__new__(cls)
diff --git a/docarray/array/mixins/find.py b/docarray/array/mixins/find.py
@@ -96,7 +96,7 @@ def find(
         limit: Optional[Union[int, float]] = 20,
         metric_name: Optional[str] = None,
         exclude_self: bool = False,
-        filter: Optional[Dict] = None,
+        filter: Union[Dict, str, None] = None,
         only_id: bool = False,
         index: str = 'text',
         on: Optional[str] = None,

diff --git a/docarray/array/storage/base/seqlike.py b/docarray/array/storage/base/seqlike.py
@@ -14,13 +14,14 @@ def _update_subindices_append_extend(self, value):
                 if len(docs_selector) > 0:
                     da.extend(docs_selector)
 
-    def insert(self, index: int, value: 'Document'):
+    def insert(self, index: int, value: 'Document', **kwargs):
         """Insert `doc` at `index`.
 
         :param index: Position of the insertion.
         :param value: The doc needs to be inserted.
+        :param kwargs: Additional Arguments that are passed to the Document Store. This has no effect for in-memory DocumentArray.
         """
-        self._set_doc_by_id(value.id, value)
+        self._set_doc_by_id(value.id, value, **kwargs)
         self._offset2ids.insert(index, value.id)
 
     def append(self, value: 'Document', **kwargs):

diff --git a/docarray/array/storage/milvus/__init__.py b/docarray/array/storage/milvus/__init__.py
@@ -0,0 +1,12 @@
+from abc import ABC
+
+from .backend import BackendMixin, MilvusConfig
+from .find import FindMixin
+from .getsetdel import GetSetDelMixin
+from .seqlike import SequenceLikeMixin
+
+__all__ = ['StorageMixins', 'MilvusConfig']
+
+
+class StorageMixins(FindMixin, BackendMixin, GetSetDelMixin, SequenceLikeMixin, ABC):
+    ...