Revert "fix: fix multimodal example (#2178)" (#2191)

This reverts commit 8849200.

jina/resources/multimodal/pods/segmenter.py

@@ -1,13 +1,11 @@
 import os
 
 from jina import Segmenter, Crafter
-from jina.executors.decorators import single, single_multi_input
 
 
 class SimpleCrafter(Crafter):
     """Simple crafter for multimodal example."""
 
-    @single
     def craft(self, tags):
         """
         Read the data and add tags.
@@ -24,7 +22,6 @@ def craft(self, tags):
 class BiSegmenter(Segmenter):
     """Segmenter for multimodal example."""
 
-    @single_multi_input(num_data=2, flatten_output=False)
     def segment(self, text, uri):
         """
         Segment data into text and uri.

jina/types/sets/document.py

@@ -160,7 +160,7 @@ def all_embeddings(self) -> Tuple['np.ndarray', 'DocumentSet']:
                 and the documents have no embedding in a :class:`DocumentSet`.
         :rtype: A tuple of embedding in :class:`np.ndarray`
         """
-        return self.extract_docs('embedding', stack_contents=True)
+        return self.extract_docs('embedding')
 
     @property
     def all_contents(self) -> Tuple['np.ndarray', 'DocumentSet']:
@@ -170,16 +170,14 @@ def all_contents(self) -> Tuple['np.ndarray', 'DocumentSet']:
                 and the documents have no contents in a :class:`DocumentSet`.
         :rtype: A tuple of embedding in :class:`np.ndarray`
         """
-        # stack true for backward compatibility, but will not work if content is blob of different shapes
-        return self.extract_docs('content', stack_contents=True)
+        return self.extract_docs('content')
 
     def extract_docs(
-        self, *fields: str, stack_contents: bool = False
+        self, *fields: str
     ) -> Tuple[Union['np.ndarray', List['np.ndarray']], 'DocumentSet']:
         """Return in batches all the values of the fields
 
         :param fields: Variable length argument with the name of the fields to extract
-        :param stack_contents: boolean flag indicating if output lists should be stacked with `np.stack`
         :return: Returns an :class:`np.ndarray` or a list of :class:`np.ndarray` with the batches for these fields
         """
 
@@ -200,7 +198,7 @@ def extract_docs(
             for idx, c in enumerate(contents):
                 if not c:
                     continue
-                if stack_contents and not isinstance(c[0], bytes):
+                if not isinstance(c[0], bytes):
                     contents[idx] = np.stack(c)
         else:
             for doc in self:
@@ -213,7 +211,7 @@ def extract_docs(
 
             if not contents:
                 contents = None
-            elif stack_contents and not isinstance(contents[0], bytes):
+            elif not isinstance(contents[0], bytes):
                 contents = np.stack(contents)
 
         if bad_docs:

tests/integration/executors_batching/test_craft_executors_batching.py

@@ -36,13 +36,12 @@ def craft(self, text, *args, **kwargs):
         return {'text': f'{text}-crafted'}
 
 
-@pytest.mark.parametrize('stack', [False, True])
 @pytest.mark.parametrize(
     'crafter', [DummyCrafterTextSingle(), DummyCrafterTextBatching()]
 )
-def test_batching_text_one_argument(stack, crafter):
+def test_batching_text_one_argument(crafter):
     docs = DocumentSet([Document(text=f'text-{i}') for i in range(15)])
-    texts, _ = docs.extract_docs('text', stack_contents=stack)
+    texts, _ = docs.extract_docs('text')
 
     crafted_docs = crafter.craft(texts)
     for i, crafted_doc in enumerate(crafted_docs):
@@ -94,14 +93,13 @@ def craft(self, text, id, *args, **kwargs):
         return {'text': f'{text}-crafted', 'id': f'{id}-crafted'}
 
 
-@pytest.mark.parametrize('stack', [False, True])
 @pytest.mark.parametrize(
     'crafter', [DummyCrafterTextIdSingle(), DummyCrafterTextIdBatching()]
 )
-def test_batching_text_multi(stack, crafter):
+def test_batching_text_multi(crafter):
     docs = DocumentSet([Document(text=f'text-{i}', id=f'id-{i}') for i in range(15)])
     required_keys = ['text', 'id']
-    text_ids, _ = docs.extract_docs(*required_keys, stack_contents=stack)
+    text_ids, _ = docs.extract_docs(*required_keys)
 
     crafted_docs = crafter.craft(*text_ids)
 
@@ -141,6 +139,7 @@ def __init__(self, *args, **kwargs):
     @batching(batch_size=3)
     def craft(self, blob, *args, **kwargs):
         assert len(blob) == 3
+        assert blob.shape == (3, 2, 5)
         return [{'blob': b} for b in blob]
 
 
@@ -154,17 +153,18 @@ def craft(self, blob, *args, **kwargs):
         return {'blob': blob}
 
 
-@pytest.mark.parametrize('stack', [False, True])
 @pytest.mark.parametrize(
     'crafter', [DummyCrafterBlobSingle(), DummyCrafterBlobBatching()]
 )
-def test_batching_blob_one_argument(stack, crafter):
+def test_batching_blob_one_argument(crafter):
     docs = DocumentSet([Document(blob=np.array([[i] * 5, [i] * 5])) for i in range(15)])
-    texts, _ = docs.extract_docs('blob', stack_contents=stack)
+    texts, _ = docs.extract_docs('blob')
 
     crafted_docs = crafter.craft(texts)
     for i, crafted_doc in enumerate(crafted_docs):
-        np.testing.assert_equal(crafted_doc['blob'], np.array([[i] * 5, [i] * 5]))
+        np.testing.assert_equal(
+            crafted_doc['blob'], np.array([[i, i, i, i, i], [i, i, i, i, i]])
+        )
 
 
 @pytest.mark.parametrize(
@@ -200,6 +200,8 @@ def __init__(self, *args, **kwargs):
     def craft(self, blob, embedding, *args, **kwargs):
         assert len(blob) == 3
         assert len(embedding) == 3
+        assert blob.shape == (3, 2, 5)
+        assert embedding.shape == (3, 5)
         return [{'blob': b, 'embedding': e} for b, e in zip(blob, embedding)]
 
 
@@ -214,11 +216,10 @@ def craft(self, blob, embedding, *args, **kwargs):
         return {'blob': blob, 'embedding': embedding}
 
 
-@pytest.mark.parametrize('stack', [False, True])
 @pytest.mark.parametrize(
     'crafter', [DummyCrafterBlobEmbeddingSingle(), DummyCrafterBlobEmbeddingBatching()]
 )
-def test_batching_blob_multi(stack, crafter):
+def test_batching_blob_multi(crafter):
     docs = DocumentSet(
         [
             Document(
@@ -229,7 +230,7 @@ def test_batching_blob_multi(stack, crafter):
         ]
     )
     required_keys = ['blob', 'embedding']
-    text_ids, _ = docs.extract_docs(*required_keys, stack_contents=stack)
+    text_ids, _ = docs.extract_docs(*required_keys)
 
     crafted_docs = crafter.craft(*text_ids)
 
@@ -279,6 +280,8 @@ def __init__(self, *args, **kwargs):
     def craft(self, text, embedding, *args, **kwargs):
         assert len(text) == 3
         assert len(embedding) == 3
+        assert text.shape == (3,)
+        assert embedding.shape == (3, 5)
         return [
             {'text': f'{t}-crafted', 'embedding': e} for t, e in zip(text, embedding)
         ]
@@ -296,16 +299,15 @@ def craft(self, text, embedding, *args, **kwargs):
         return {'text': f'{text}-crafted', 'embedding': embedding}
 
 
-@pytest.mark.parametrize('stack', [False, True])
 @pytest.mark.parametrize(
     'crafter', [DummyCrafterTextEmbeddingSingle(), DummyCrafterTextEmbeddingBatching()]
 )
-def test_batching_mix_multi(stack, crafter):
+def test_batching_mix_multi(crafter):
     docs = DocumentSet(
         [Document(text=f'text-{i}', embedding=np.array([i] * 5)) for i in range(15)]
     )
     required_keys = ['text', 'embedding']
-    text_ids, _ = docs.extract_docs(*required_keys, stack_contents=stack)
+    text_ids, _ = docs.extract_docs(*required_keys)
 
     crafted_docs = crafter.craft(*text_ids)
 

tests/integration/executors_batching/test_encoder_executors_batching.py

@@ -39,7 +39,7 @@ def encode(self, data, *args, **kwargs):
 )
 def test_batching_encode_text(encoder):
     docs = DocumentSet([Document(text=f'text-{i}') for i in range(15)])
-    texts, _ = docs.all_contents
+    texts, _ = docs.extract_docs('text')
 
     embeds = encoder.encode(texts)
 
@@ -75,7 +75,7 @@ def encode(self, data, *args, **kwargs):
 )
 def test_batching_encode_blob(encoder):
     docs = DocumentSet([Document(blob=np.random.random((10, 20))) for _ in range(15)])
-    blob, _ = docs.all_contents
+    blob, _ = docs.extract_docs('blob')
 
     embeds = encoder.encode(blob)