This repository was archived by the owner on Sep 9, 2026. It is now read-only.
Repository navigation
Expand file tree
/
Copy pathtest_array_from_to_bytes.py
More file actions
148 lines (124 loc) · 4.74 KB
/
Copy pathtest_array_from_to_bytes.py
File metadata and controls
148 lines (124 loc) · 4.74 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
import pytest
from docarray import BaseDoc, DocList, DocVec
from docarray.documents import ImageDoc
from docarray.typing import NdArray, TorchTensor
class MyDoc(BaseDoc):
embedding: NdArray
text: str
image: ImageDoc
@pytest.mark.parametrize(
'protocol', ['pickle-array', 'protobuf-array', 'protobuf', 'pickle']
)
@pytest.mark.parametrize('compress', ['lz4', 'bz2', 'lzma', 'zlib', 'gzip', None])
@pytest.mark.parametrize('show_progress', [False, True])
@pytest.mark.parametrize('array_cls', [DocList, DocVec])
def test_from_to_bytes(protocol, compress, show_progress, array_cls):
da = array_cls[MyDoc](
[
MyDoc(
embedding=[1, 2, 3, 4, 5], text='hello', image=ImageDoc(url='aux.png')
),
MyDoc(embedding=[5, 4, 3, 2, 1], text='hello world', image=ImageDoc()),
]
)
bytes_da = da.to_bytes(
protocol=protocol, compress=compress, show_progress=show_progress
)
da2 = array_cls[MyDoc].from_bytes(
bytes_da, protocol=protocol, compress=compress, show_progress=show_progress
)
assert len(da2) == 2
assert len(da) == len(da2)
for d1, d2 in zip(da, da2):
assert d1.embedding.tolist() == d2.embedding.tolist()
assert d1.text == d2.text
assert d1.image.url == d2.image.url
assert da[1].image.url is None
assert da2[1].image.url is None
@pytest.mark.parametrize(
'protocol', ['pickle-array', 'protobuf-array', 'protobuf', 'pickle']
)
@pytest.mark.parametrize('compress', ['lz4', 'bz2', 'lzma', 'zlib', 'gzip', None])
@pytest.mark.parametrize('show_progress', [False, True]) # [False, True])
@pytest.mark.parametrize('array_cls', [DocList, DocVec])
def test_from_to_base64(protocol, compress, show_progress, array_cls):
da = array_cls[MyDoc](
[
MyDoc(
embedding=[1, 2, 3, 4, 5], text='hello', image=ImageDoc(url='aux.png')
),
MyDoc(embedding=[5, 4, 3, 2, 1], text='hello world', image=ImageDoc()),
]
)
bytes_da = da.to_base64(
protocol=protocol, compress=compress, show_progress=show_progress
)
da2 = array_cls[MyDoc].from_base64(
bytes_da, protocol=protocol, compress=compress, show_progress=show_progress
)
assert len(da2) == 2
assert len(da) == len(da2)
for d1, d2 in zip(da, da2):
assert d1.embedding.tolist() == d2.embedding.tolist()
assert d1.text == d2.text
assert d1.image.url == d2.image.url
assert da[1].image.url is None
assert da2[1].image.url is None
# test_from_to_base64('protobuf', 'lz4', False, DocVec)
class MyTensorTypeDocNdArray(BaseDoc):
embedding: NdArray
text: str
image: ImageDoc
class MyTensorTypeDocTorchTensor(BaseDoc):
embedding: TorchTensor
text: str
image: ImageDoc
@pytest.mark.parametrize(
'doc_type, tensor_type',
[(MyTensorTypeDocNdArray, NdArray), (MyTensorTypeDocTorchTensor, TorchTensor)],
)
@pytest.mark.parametrize('protocol', ['protobuf-array', 'pickle-array'])
def test_from_to_base64_tensor_type(doc_type, tensor_type, protocol):
da = DocVec[doc_type](
[
doc_type(
embedding=[1, 2, 3, 4, 5], text='hello', image=ImageDoc(url='aux.png')
),
doc_type(embedding=[5, 4, 3, 2, 1], text='hello world', image=ImageDoc()),
],
tensor_type=tensor_type,
)
bytes_da = da.to_base64(protocol=protocol)
da2 = DocVec[doc_type].from_base64(
bytes_da, tensor_type=tensor_type, protocol=protocol
)
assert da2.tensor_type == tensor_type
assert isinstance(da2.embedding, tensor_type)
@pytest.mark.parametrize('tensor_type', [NdArray, TorchTensor])
def test_from_to_bytes_tensor_type(tensor_type):
da = DocVec[MyDoc](
[
MyDoc(
embedding=[1, 2, 3, 4, 5], text='hello', image=ImageDoc(url='aux.png')
),
MyDoc(embedding=[5, 4, 3, 2, 1], text='hello world', image=ImageDoc()),
],
tensor_type=tensor_type,
)
bytes_da = da.to_bytes()
da2 = DocVec[MyDoc].from_bytes(bytes_da, tensor_type=tensor_type)
assert da2.tensor_type == tensor_type
assert isinstance(da2.embedding, tensor_type)
def test_union_type_error(tmp_path):
from typing import Union
from docarray.documents import TextDoc
class CustomDoc(BaseDoc):
ud: Union[TextDoc, ImageDoc] = TextDoc(text='union type')
docs = DocList[CustomDoc]([CustomDoc(ud=TextDoc(text='union type'))])
with pytest.raises(ValueError):
docs.from_bytes(docs.to_bytes())
class BasisUnion(BaseDoc):
ud: Union[int, str]
docs_basic = DocList[BasisUnion]([BasisUnion(ud="hello")])
docs_copy = DocList[BasisUnion].from_bytes(docs_basic.to_bytes())
assert docs_copy == docs_basic