-
Notifications
You must be signed in to change notification settings - Fork 1
/
azure_search_helper.py
275 lines (256 loc) · 9.54 KB
/
azure_search_helper.py
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
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
import logging
from typing import Union
from langchain.vectorstores.azuresearch import AzureSearch
from azure.core.credentials import AzureKeyCredential
from azure.identity import DefaultAzureCredential
from azure.search.documents import SearchClient
from azure.search.documents.indexes import SearchIndexClient
from azure.search.documents.indexes.models import (
ExhaustiveKnnAlgorithmConfiguration,
ExhaustiveKnnParameters,
HnswAlgorithmConfiguration,
HnswParameters,
SearchableField,
SearchField,
SearchFieldDataType,
SearchIndex,
SemanticConfiguration,
SemanticField,
SemanticPrioritizedFields,
SemanticSearch,
SimpleField,
VectorSearch,
VectorSearchAlgorithmKind,
VectorSearchAlgorithmMetric,
VectorSearchProfile,
)
from ..helpers.azure_computer_vision_client import AzureComputerVisionClient
from .llm_helper import LLMHelper
from .env_helper import EnvHelper
logger = logging.getLogger(__name__)
class AzureSearchHelper:
_search_dimension: int | None = None
_image_search_dimension: int | None = None
def __init__(self):
self.llm_helper = LLMHelper()
self.env_helper = EnvHelper()
search_credential = self._search_credential()
self.search_client = self._create_search_client(search_credential)
self.search_index_client = self._create_search_index_client(search_credential)
self.azure_computer_vision_client = AzureComputerVisionClient(self.env_helper)
def _search_credential(self):
if self.env_helper.is_auth_type_keys():
return AzureKeyCredential(self.env_helper.AZURE_SEARCH_KEY)
else:
return DefaultAzureCredential()
def _create_search_client(
self, search_credential: Union[AzureKeyCredential, DefaultAzureCredential]
) -> SearchClient:
return SearchClient(
endpoint=self.env_helper.AZURE_SEARCH_SERVICE,
index_name=self.env_helper.AZURE_SEARCH_INDEX,
credential=search_credential,
)
def _create_search_index_client(
self, search_credential: Union[AzureKeyCredential, DefaultAzureCredential]
):
return SearchIndexClient(
endpoint=self.env_helper.AZURE_SEARCH_SERVICE, credential=search_credential
)
def get_search_client(self) -> SearchClient:
self.create_index()
return self.search_client
@property
def search_dimensions(self) -> int:
if AzureSearchHelper._search_dimension is None:
AzureSearchHelper._search_dimension = len(
self.llm_helper.get_embedding_model().embed_query("Text")
)
return AzureSearchHelper._search_dimension
@property
def image_search_dimensions(self) -> int:
if AzureSearchHelper._image_search_dimension is None:
AzureSearchHelper._image_search_dimension = len(
self.azure_computer_vision_client.vectorize_text("Text")
)
return AzureSearchHelper._image_search_dimension
def create_index(self):
fields = [
SimpleField(
name="id",
type=SearchFieldDataType.String,
key=True,
filterable=True,
),
SearchableField(
name="content",
type=SearchFieldDataType.String,
),
SearchField(
name="content_vector",
type=SearchFieldDataType.Collection(SearchFieldDataType.Single),
searchable=True,
vector_search_dimensions=self.search_dimensions,
vector_search_profile_name="myHnswProfile",
),
SearchableField(
name="metadata",
type=SearchFieldDataType.String,
),
SearchableField(
name="title",
type=SearchFieldDataType.String,
facetable=True,
filterable=True,
),
SearchableField(
name="source",
type=SearchFieldDataType.String,
filterable=True,
),
SimpleField(
name="chunk",
type=SearchFieldDataType.Int32,
filterable=True,
),
SimpleField(
name="offset",
type=SearchFieldDataType.Int32,
filterable=True,
),
]
if self.env_helper.USE_ADVANCED_IMAGE_PROCESSING:
logger.info("Adding image_vector field to index")
fields.append(
SearchField(
name="image_vector",
type=SearchFieldDataType.Collection(SearchFieldDataType.Single),
searchable=True,
vector_search_dimensions=self.image_search_dimensions,
vector_search_profile_name="myHnswProfile",
),
)
index = SearchIndex(
name=self.env_helper.AZURE_SEARCH_INDEX,
fields=fields,
semantic_search=(
SemanticSearch(
configurations=[
SemanticConfiguration(
name=self.env_helper.AZURE_SEARCH_SEMANTIC_SEARCH_CONFIG,
prioritized_fields=SemanticPrioritizedFields(
title_field=None,
content_fields=[SemanticField(field_name="content")],
),
)
]
)
),
vector_search=VectorSearch(
algorithms=[
HnswAlgorithmConfiguration(
name="default",
parameters=HnswParameters(
metric=VectorSearchAlgorithmMetric.COSINE
),
kind=VectorSearchAlgorithmKind.HNSW,
),
ExhaustiveKnnAlgorithmConfiguration(
name="default_exhaustive_knn",
kind=VectorSearchAlgorithmKind.EXHAUSTIVE_KNN,
parameters=ExhaustiveKnnParameters(
metric=VectorSearchAlgorithmMetric.COSINE
),
),
],
profiles=[
VectorSearchProfile(
name="myHnswProfile",
algorithm_configuration_name="default",
),
VectorSearchProfile(
name="myExhaustiveKnnProfile",
algorithm_configuration_name="default_exhaustive_knn",
),
],
),
)
if self._index_not_exists(self.env_helper.AZURE_SEARCH_INDEX):
logger.info(
f"Creating or updating index {self.env_helper.AZURE_SEARCH_INDEX}"
)
self.search_index_client.create_index(index)
def _index_not_exists(self, index_name: str) -> bool:
return index_name not in [
name for name in self.search_index_client.list_index_names()
]
def get_conversation_logger(self):
fields = [
SimpleField(
name="id",
type=SearchFieldDataType.String,
key=True,
filterable=True,
),
SimpleField(
name="conversation_id",
type=SearchFieldDataType.String,
filterable=True,
facetable=True,
),
SearchableField(
name="content",
type=SearchFieldDataType.String,
),
SearchField(
name="content_vector",
type=SearchFieldDataType.Collection(SearchFieldDataType.Single),
searchable=True,
vector_search_dimensions=self.search_dimensions,
vector_search_profile_name="myHnswProfile",
),
SearchableField(
name="metadata",
type=SearchFieldDataType.String,
),
SimpleField(
name="type",
type=SearchFieldDataType.String,
facetable=True,
filterable=True,
),
SimpleField(
name="user_id",
type=SearchFieldDataType.String,
filterable=True,
facetable=True,
),
SimpleField(
name="sources",
type=SearchFieldDataType.Collection(SearchFieldDataType.String),
filterable=True,
facetable=True,
),
SimpleField(
name="created_at",
type=SearchFieldDataType.DateTimeOffset,
filterable=True,
),
SimpleField(
name="updated_at",
type=SearchFieldDataType.DateTimeOffset,
filterable=True,
),
]
return AzureSearch(
azure_search_endpoint=self.env_helper.AZURE_SEARCH_SERVICE,
azure_search_key=(
self.env_helper.AZURE_SEARCH_KEY
if self.env_helper.is_auth_type_keys()
else None
),
index_name=self.env_helper.AZURE_SEARCH_CONVERSATIONS_LOG_INDEX,
embedding_function=self.llm_helper.get_embedding_model().embed_query,
fields=fields,
user_agent="langchain chatwithyourdata-sa",
)