Skip to content

Azureaisearch

CognitiveSearchVectorStore module-attribute #

CognitiveSearchVectorStore = AzureAISearchVectorStore

AzureAISearchVectorStore #

Bases: BasePydanticVectorStore

Azure AI Search vector store.

Examples:

pip install llama-index-vector-stores-azureaisearch

from azure.core.credentials import AzureKeyCredential
from azure.search.documents import SearchClient
from azure.search.documents.indexes import SearchIndexClient
from llama_index.vector_stores.azureaisearch import AzureAISearchVectorStore
from llama_index.vector_stores.azureaisearch import IndexManagement, MetadataIndexFieldType

# Azure AI Search setup
search_service_api_key = "YOUR-AZURE-SEARCH-SERVICE-ADMIN-KEY"
search_service_endpoint = "YOUR-AZURE-SEARCH-SERVICE-ENDPOINT"
search_service_api_version = "2023-11-01"
credential = AzureKeyCredential(search_service_api_key)

# Index name to use
index_name = "llamaindex-vector-demo"

# Use index client to demonstrate creating an index
index_client = SearchIndexClient(
    endpoint=search_service_endpoint,
    credential=credential,
)

metadata_fields = {
    "author": "author",
    "theme": ("topic", MetadataIndexFieldType.STRING),
    "director": "director",
}

# Creating an Azure AI Search Vector Store
vector_store = AzureAISearchVectorStore(
    search_or_index_client=index_client,
    filterable_metadata_field_keys=metadata_fields,
    index_name=index_name,
    index_management=IndexManagement.CREATE_IF_NOT_EXISTS,
    id_field_key="id",
    chunk_field_key="chunk",
    embedding_field_key="embedding",
    embedding_dimensionality=1536,
    metadata_string_field_key="metadata",
    doc_id_field_key="doc_id",
    language_analyzer="en.lucene",
    vector_algorithm_type="exhaustiveKnn",
)
Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-azureaisearch/llama_index/vector_stores/azureaisearch/base.py
 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
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
class AzureAISearchVectorStore(BasePydanticVectorStore):
    """
    Azure AI Search vector store.

    Examples:
        `pip install llama-index-vector-stores-azureaisearch`

        ```python
        from azure.core.credentials import AzureKeyCredential
        from azure.search.documents import SearchClient
        from azure.search.documents.indexes import SearchIndexClient
        from llama_index.vector_stores.azureaisearch import AzureAISearchVectorStore
        from llama_index.vector_stores.azureaisearch import IndexManagement, MetadataIndexFieldType

        # Azure AI Search setup
        search_service_api_key = "YOUR-AZURE-SEARCH-SERVICE-ADMIN-KEY"
        search_service_endpoint = "YOUR-AZURE-SEARCH-SERVICE-ENDPOINT"
        search_service_api_version = "2023-11-01"
        credential = AzureKeyCredential(search_service_api_key)

        # Index name to use
        index_name = "llamaindex-vector-demo"

        # Use index client to demonstrate creating an index
        index_client = SearchIndexClient(
            endpoint=search_service_endpoint,
            credential=credential,
        )

        metadata_fields = {
            "author": "author",
            "theme": ("topic", MetadataIndexFieldType.STRING),
            "director": "director",
        }

        # Creating an Azure AI Search Vector Store
        vector_store = AzureAISearchVectorStore(
            search_or_index_client=index_client,
            filterable_metadata_field_keys=metadata_fields,
            index_name=index_name,
            index_management=IndexManagement.CREATE_IF_NOT_EXISTS,
            id_field_key="id",
            chunk_field_key="chunk",
            embedding_field_key="embedding",
            embedding_dimensionality=1536,
            metadata_string_field_key="metadata",
            doc_id_field_key="doc_id",
            language_analyzer="en.lucene",
            vector_algorithm_type="exhaustiveKnn",
        )
        ```
    """

    stores_text: bool = True
    flat_metadata: bool = True

    _index_client: SearchIndexClient = PrivateAttr()
    _index_name: Optional[str] = PrivateAttr()
    _async_index_client: AsyncSearchIndexClient = PrivateAttr()
    _search_client: SearchClient = PrivateAttr()
    _async_search_client: AsyncSearchClient = PrivateAttr()
    _embedding_dimensionality: int = PrivateAttr()
    _language_analyzer: str = PrivateAttr()
    _field_mapping: Dict[str, str] = PrivateAttr()
    _index_management: IndexManagement = PrivateAttr()
    _index_mapping: Callable[
        [Dict[str, str], Dict[str, Any]], Dict[str, str]
    ] = PrivateAttr()
    _metadata_to_index_field_map: Dict[
        str, Tuple[str, MetadataIndexFieldType]
    ] = PrivateAttr()
    _vector_profile_name: str = PrivateAttr()

    def _normalise_metadata_to_index_fields(
        self,
        filterable_metadata_field_keys: Union[
            List[str],
            Dict[str, str],
            Dict[str, Tuple[str, MetadataIndexFieldType]],
            None,
        ] = [],
    ) -> Dict[str, Tuple[str, MetadataIndexFieldType]]:
        index_field_spec: Dict[str, Tuple[str, MetadataIndexFieldType]] = {}

        if isinstance(filterable_metadata_field_keys, List):
            for field in filterable_metadata_field_keys:
                # Index field name and the metadata field name are the same
                # Use String as the default index field type
                index_field_spec[field] = (field, MetadataIndexFieldType.STRING)

        elif isinstance(filterable_metadata_field_keys, Dict):
            for k, v in filterable_metadata_field_keys.items():
                if isinstance(v, tuple):
                    # Index field name and metadata field name may differ
                    # The index field type used is as supplied
                    index_field_spec[k] = v
                else:
                    # Index field name and metadata field name may differ
                    # Use String as the default index field type
                    index_field_spec[k] = (v, MetadataIndexFieldType.STRING)

        return index_field_spec

    def _create_index_if_not_exists(self, index_name: str) -> None:
        if index_name not in self._index_client.list_index_names():
            logger.info(
                f"Index {index_name} does not exist in Azure AI Search, creating index"
            )
            self._create_index(index_name)

    async def _acreate_index_if_not_exists(self, index_name: str) -> None:
        list_index_names = set()

        async for index in self._async_index_client.list_index_names():
            list_index_names.add(index)

        if index_name not in list_index_names:
            logger.info(
                f"Index {index_name} does not exist in Azure AI Search, creating index"
            )
            await self._acreate_index(index_name)

    def _create_metadata_index_fields(self) -> List[Any]:
        """Create a list of index fields for storing metadata values."""
        from azure.search.documents.indexes.models import SimpleField

        index_fields = []

        # create search fields
        for v in self._metadata_to_index_field_map.values():
            field_name, field_type = v

            if field_type == MetadataIndexFieldType.STRING:
                index_field_type = "Edm.String"
            elif field_type == MetadataIndexFieldType.INT32:
                index_field_type = "Edm.Int32"
            elif field_type == MetadataIndexFieldType.INT64:
                index_field_type = "Edm.Int64"
            elif field_type == MetadataIndexFieldType.DOUBLE:
                index_field_type = "Edm.Double"
            elif field_type == MetadataIndexFieldType.BOOLEAN:
                index_field_type = "Edm.Boolean"

            field = SimpleField(name=field_name, type=index_field_type, filterable=True)
            index_fields.append(field)

        return index_fields

    def _create_index(self, index_name: Optional[str]) -> None:
        """
        Creates a default index based on the supplied index name, key field names and
        metadata filtering keys.
        """
        from azure.search.documents.indexes.models import (
            ExhaustiveKnnAlgorithmConfiguration,
            ExhaustiveKnnParameters,
            HnswAlgorithmConfiguration,
            HnswParameters,
            SearchableField,
            SearchField,
            SearchFieldDataType,
            SearchIndex,
            SemanticConfiguration,
            SemanticField,
            SemanticPrioritizedFields,
            SemanticSearch,
            SimpleField,
            VectorSearch,
            VectorSearchAlgorithmKind,
            VectorSearchAlgorithmMetric,
            VectorSearchProfile,
        )

        logger.info(f"Configuring {index_name} fields for Azure AI Search")
        fields = [
            SimpleField(name=self._field_mapping["id"], type="Edm.String", key=True),
            SearchableField(
                name=self._field_mapping["chunk"],
                type="Edm.String",
                analyzer_name=self._language_analyzer,
            ),
            SearchField(
                name=self._field_mapping["embedding"],
                type=SearchFieldDataType.Collection(SearchFieldDataType.Single),
                searchable=True,
                vector_search_dimensions=self._embedding_dimensionality,
                vector_search_profile_name=self._vector_profile_name,
            ),
            SimpleField(name=self._field_mapping["metadata"], type="Edm.String"),
            SimpleField(
                name=self._field_mapping["doc_id"], type="Edm.String", filterable=True
            ),
        ]
        logger.info(f"Configuring {index_name} metadata fields")
        metadata_index_fields = self._create_metadata_index_fields()
        fields.extend(metadata_index_fields)
        logger.info(f"Configuring {index_name} vector search")
        # Configure the vector search algorithms and profiles
        vector_search = VectorSearch(
            algorithms=[
                HnswAlgorithmConfiguration(
                    name="myHnsw",
                    kind=VectorSearchAlgorithmKind.HNSW,
                    # For more information on HNSw parameters, visit https://learn.microsoft.com//azure/search/vector-search-ranking#creating-the-hnsw-graph
                    parameters=HnswParameters(
                        m=4,
                        ef_construction=400,
                        ef_search=500,
                        metric=VectorSearchAlgorithmMetric.COSINE,
                    ),
                ),
                ExhaustiveKnnAlgorithmConfiguration(
                    name="myExhaustiveKnn",
                    kind=VectorSearchAlgorithmKind.EXHAUSTIVE_KNN,
                    parameters=ExhaustiveKnnParameters(
                        metric=VectorSearchAlgorithmMetric.COSINE,
                    ),
                ),
            ],
            profiles=[
                VectorSearchProfile(
                    name="myHnswProfile",
                    algorithm_configuration_name="myHnsw",
                ),
                # Add more profiles if needed
                VectorSearchProfile(
                    name="myExhaustiveKnnProfile",
                    algorithm_configuration_name="myExhaustiveKnn",
                ),
                # Add more profiles if needed
            ],
        )
        logger.info(f"Configuring {index_name} semantic search")
        semantic_config = SemanticConfiguration(
            name="mySemanticConfig",
            prioritized_fields=SemanticPrioritizedFields(
                content_fields=[SemanticField(field_name=self._field_mapping["chunk"])],
            ),
        )

        semantic_search = SemanticSearch(configurations=[semantic_config])

        index = SearchIndex(
            name=index_name,
            fields=fields,
            vector_search=vector_search,
            semantic_search=semantic_search,
        )

        logger.debug(f"Creating {index_name} search index")
        self._index_client.create_index(index)

    async def _acreate_index(self, index_name: Optional[str]) -> None:
        """
        Creates a default index based on the supplied index name, key field names and
        metadata filtering keys.
        """
        from azure.search.documents.indexes.models import (
            ExhaustiveKnnAlgorithmConfiguration,
            ExhaustiveKnnParameters,
            HnswAlgorithmConfiguration,
            HnswParameters,
            SearchableField,
            SearchField,
            SearchFieldDataType,
            SearchIndex,
            SemanticConfiguration,
            SemanticField,
            SemanticPrioritizedFields,
            SemanticSearch,
            SimpleField,
            VectorSearch,
            VectorSearchAlgorithmKind,
            VectorSearchAlgorithmMetric,
            VectorSearchProfile,
        )

        logger.info(f"Configuring {index_name} fields for Azure AI Search")
        fields = [
            SimpleField(name=self._field_mapping["id"], type="Edm.String", key=True),
            SearchableField(
                name=self._field_mapping["chunk"],
                type="Edm.String",
                analyzer_name=self._language_analyzer,
            ),
            SearchField(
                name=self._field_mapping["embedding"],
                type=SearchFieldDataType.Collection(SearchFieldDataType.Single),
                searchable=True,
                vector_search_dimensions=self._embedding_dimensionality,
                vector_search_profile_name=self._vector_profile_name,
            ),
            SimpleField(name=self._field_mapping["metadata"], type="Edm.String"),
            SimpleField(
                name=self._field_mapping["doc_id"], type="Edm.String", filterable=True
            ),
        ]
        logger.info(f"Configuring {index_name} metadata fields")
        metadata_index_fields = self._create_metadata_index_fields()
        fields.extend(metadata_index_fields)
        logger.info(f"Configuring {index_name} vector search")
        # Configure the vector search algorithms and profiles
        vector_search = VectorSearch(
            algorithms=[
                HnswAlgorithmConfiguration(
                    name="myHnsw",
                    kind=VectorSearchAlgorithmKind.HNSW,
                    # For more information on HNSw parameters, visit https://learn.microsoft.com//azure/search/vector-search-ranking#creating-the-hnsw-graph
                    parameters=HnswParameters(
                        m=4,
                        ef_construction=400,
                        ef_search=500,
                        metric=VectorSearchAlgorithmMetric.COSINE,
                    ),
                ),
                ExhaustiveKnnAlgorithmConfiguration(
                    name="myExhaustiveKnn",
                    kind=VectorSearchAlgorithmKind.EXHAUSTIVE_KNN,
                    parameters=ExhaustiveKnnParameters(
                        metric=VectorSearchAlgorithmMetric.COSINE,
                    ),
                ),
            ],
            profiles=[
                VectorSearchProfile(
                    name="myHnswProfile",
                    algorithm_configuration_name="myHnsw",
                ),
                # Add more profiles if needed
                VectorSearchProfile(
                    name="myExhaustiveKnnProfile",
                    algorithm_configuration_name="myExhaustiveKnn",
                ),
                # Add more profiles if needed
            ],
        )
        logger.info(f"Configuring {index_name} semantic search")
        semantic_config = SemanticConfiguration(
            name="mySemanticConfig",
            prioritized_fields=SemanticPrioritizedFields(
                content_fields=[SemanticField(field_name=self._field_mapping["chunk"])],
            ),
        )

        semantic_search = SemanticSearch(configurations=[semantic_config])

        index = SearchIndex(
            name=index_name,
            fields=fields,
            vector_search=vector_search,
            semantic_search=semantic_search,
        )
        logger.debug(f"Creating {index_name} search index")
        await self._async_index_client.create_index(index)

    def _validate_index(self, index_name: Optional[str]) -> None:
        if self._index_client and index_name:
            if index_name not in self._index_client.list_index_names():
                raise ValueError(
                    f"Validation failed, index {index_name} does not exist."
                )

    async def _avalidate_index(self, index_name: Optional[str]) -> None:
        list_index_names = set()

        async for index in self._async_index_client.list_index_names():
            list_index_names.add(index)

        if self._async_index_client and index_name:
            if index_name not in list_index_names:
                raise ValueError(
                    f"Validation failed, index {index_name} does not exist."
                )

    def __init__(
        self,
        search_or_index_client: Any,
        id_field_key: str,
        chunk_field_key: str,
        embedding_field_key: str,
        metadata_string_field_key: str,
        doc_id_field_key: str,
        filterable_metadata_field_keys: Optional[
            Union[
                List[str],
                Dict[str, str],
                Dict[str, Tuple[str, MetadataIndexFieldType]],
            ]
        ] = None,
        index_name: Optional[str] = None,
        index_mapping: Optional[
            Callable[[Dict[str, str], Dict[str, Any]], Dict[str, str]]
        ] = None,
        index_management: IndexManagement = IndexManagement.NO_VALIDATION,
        embedding_dimensionality: int = 1536,
        vector_algorithm_type: str = "exhaustiveKnn",
        # If we have content in other languages, it is better to enable the language analyzer to be adjusted in searchable fields.
        # https://learn.microsoft.com/en-us/azure/search/index-add-language-analyzers
        language_analyzer: str = "en.lucene",
        **kwargs: Any,
    ) -> None:
        # ruff: noqa: E501
        """
        Embeddings and documents are stored in an Azure AI Search index,
        a merge or upload approach is used when adding embeddings.
        When adding multiple embeddings the index is updated by this vector store
        in batches of 10 documents, very large nodes may result in failure due to
        the batch byte size being exceeded.

        Args:
            search_client (azure.search.documents.SearchClient):
                Client for index to populated / queried.
            id_field_key (str): Index field storing the id
            chunk_field_key (str): Index field storing the node text
            embedding_field_key (str): Index field storing the embedding vector
            metadata_string_field_key (str):
                Index field storing node metadata as a json string.
                Schema is arbitrary, to filter on metadata values they must be stored
                as separate fields in the index, use filterable_metadata_field_keys
                to specify the metadata values that should be stored in these filterable fields
            doc_id_field_key (str): Index field storing doc_id
            index_mapping:
                Optional function with definition
                (enriched_doc: Dict[str, str], metadata: Dict[str, Any]): Dict[str,str]
                used to map document fields to the AI search index fields
                (return value of function).
                If none is specified a default mapping is provided which uses
                the field keys. The keys in the enriched_doc are
                ["id", "chunk", "embedding", "metadata"]
                The default mapping is:
                    - "id" to id_field_key
                    - "chunk" to chunk_field_key
                    - "embedding" to embedding_field_key
                    - "metadata" to metadata_field_key
            *kwargs (Any): Additional keyword arguments.

        Raises:
            ImportError: Unable to import `azure-search-documents`
            ValueError: If `search_or_index_client` is not provided
            ValueError: If `index_name` is not provided and `search_or_index_client`
                is of type azure.search.documents.SearchIndexClient
            ValueError: If `index_name` is provided and `search_or_index_client`
                is of type azure.search.documents.SearchClient
            ValueError: If `create_index_if_not_exists` is true and
                `search_or_index_client` is of type azure.search.documents.SearchClient
        """
        import_err_msg = (
            "`azure-search-documents` package not found, please run "
            "`pip install azure-search-documents==11.4.0`"
        )

        try:
            import azure.search.documents  # noqa
            from azure.search.documents import SearchClient
            from azure.search.documents.indexes import SearchIndexClient
        except ImportError:
            raise ImportError(import_err_msg)

        self._index_client: SearchIndexClient = cast(SearchIndexClient, None)
        self._async_index_client: AsyncSearchIndexClient = cast(
            AsyncSearchIndexClient, None
        )
        self._search_client: SearchClient = cast(SearchClient, None)
        self._async_search_client: AsyncSearchClient = cast(AsyncSearchClient, None)
        self._embedding_dimensionality = embedding_dimensionality
        self._index_name = index_name

        if vector_algorithm_type == "exhaustiveKnn":
            self._vector_profile_name = "myExhaustiveKnnProfile"
        elif vector_algorithm_type == "hnsw":
            self._vector_profile_name = "myHnswProfile"
        else:
            raise ValueError(
                "Only 'exhaustiveKnn' and 'hnsw' are supported for vector_algorithm_type"
            )

        self._language_analyzer = language_analyzer

        # Validate search_or_index_client
        if search_or_index_client is not None:
            if isinstance(search_or_index_client, SearchIndexClient):
                # If SearchIndexClient is supplied so must index_name
                self._index_client = cast(SearchIndexClient, search_or_index_client)

                if not index_name:
                    raise ValueError(
                        "index_name must be supplied if search_or_index_client is of "
                        "type azure.search.documents.SearchIndexClient"
                    )

                self._search_client = self._index_client.get_search_client(
                    index_name=index_name
                )

            elif isinstance(search_or_index_client, AsyncSearchIndexClient):
                # If SearchIndexClient is supplied so must index_name
                self._async_index_client = cast(
                    AsyncSearchIndexClient, search_or_index_client
                )

                if not index_name:
                    raise ValueError(
                        "index_name must be supplied if search_or_index_client is of "
                        "type azure.search.documents.aio.SearchIndexClient"
                    )

                self._async_search_client = self._async_index_client.get_search_client(
                    index_name=index_name
                )

            elif isinstance(search_or_index_client, SearchClient):
                self._search_client = cast(SearchClient, search_or_index_client)

                # Validate index_name
                if index_name:
                    raise ValueError(
                        "index_name cannot be supplied if search_or_index_client "
                        "is of type azure.search.documents.SearchClient"
                    )

            elif isinstance(search_or_index_client, AsyncSearchClient):
                self._async_search_client = cast(
                    AsyncSearchClient, search_or_index_client
                )

                # Validate index_name
                if index_name:
                    raise ValueError(
                        "index_name cannot be supplied if search_or_index_client "
                        "is of type azure.search.documents.SearchClient"
                    )

            if isinstance(search_or_index_client, AsyncSearchIndexClient):
                if not self._async_index_client and not self._async_search_client:
                    raise ValueError(
                        "search_or_index_client must be of type "
                        "azure.search.documents.SearchIndexClient or "
                        "azure.search.documents.SearchClient"
                    )

            if isinstance(search_or_index_client, SearchIndexClient):
                if not self._index_client and not self._search_client:
                    raise ValueError(
                        "search_or_index_client must be of type "
                        "azure.search.documents.SearchIndexClient or "
                        "azure.search.documents.SearchClient"
                    )
        else:
            raise ValueError("search_or_index_client not specified")

        if index_management == IndexManagement.CREATE_IF_NOT_EXISTS and not (
            self._index_client or self._async_index_client
        ):
            raise ValueError(
                "index_management has value of IndexManagement.CREATE_IF_NOT_EXISTS "
                "but search_or_index_client is not of type "
                "azure.search.documents.SearchIndexClient or azure.search.documents.aio.SearchIndexClient "
            )

        self._index_management = index_management

        # Default field mapping
        field_mapping = {
            "id": id_field_key,
            "chunk": chunk_field_key,
            "embedding": embedding_field_key,
            "metadata": metadata_string_field_key,
            "doc_id": doc_id_field_key,
        }

        self._field_mapping = field_mapping

        self._index_mapping = (
            self._default_index_mapping if index_mapping is None else index_mapping
        )

        # self._filterable_metadata_field_keys = filterable_metadata_field_keys
        self._metadata_to_index_field_map = self._normalise_metadata_to_index_fields(
            filterable_metadata_field_keys
        )

        # need to do lazy init for async client
        if not isinstance(search_or_index_client, AsyncSearchIndexClient):
            if self._index_management == IndexManagement.CREATE_IF_NOT_EXISTS:
                if index_name:
                    self._create_index_if_not_exists(index_name)

            if self._index_management == IndexManagement.VALIDATE_INDEX:
                self._validate_index(index_name)

        super().__init__()

    @property
    def client(self) -> Any:
        """Get client."""
        return self._search_client

    @property
    def aclient(self) -> Any:
        """Get async client."""
        return self._async_search_client

    def _default_index_mapping(
        self, enriched_doc: Dict[str, str], metadata: Dict[str, Any]
    ) -> Dict[str, str]:
        index_doc: Dict[str, str] = {}

        for field in self._field_mapping:
            index_doc[self._field_mapping[field]] = enriched_doc[field]

        for metadata_field_name, (
            index_field_name,
            _,
        ) in self._metadata_to_index_field_map.items():
            metadata_value = metadata.get(metadata_field_name)
            if metadata_value:
                index_doc[index_field_name] = metadata_value

        return index_doc

    def add(
        self,
        nodes: List[BaseNode],
        **add_kwargs: Any,
    ) -> List[str]:
        """
        Add nodes to index associated with the configured search client.

        Args:
            nodes: List[BaseNode]: nodes with embeddings

        """
        from azure.search.documents import IndexDocumentsBatch

        if not self._search_client:
            raise ValueError("Search client not initialized")

        accumulator = IndexDocumentsBatch()
        documents = []

        ids = []
        accumulated_size = 0
        max_size = 16 * 1024 * 1024  # 16MB in bytes
        max_docs = 1000

        for node in nodes:
            logger.debug(f"Processing embedding: {node.node_id}")
            ids.append(node.node_id)

            index_document = self._create_index_document(node)
            document_size = len(
                str(node.get_content(metadata_mode=MetadataMode.NONE)).encode("utf-8")
            )
            documents.append(index_document)
            accumulated_size += document_size

            accumulator.add_upload_actions(index_document)

            if len(documents) >= max_docs or accumulated_size >= max_size:
                logger.info(
                    f"Uploading batch of size {len(documents)}, "
                    f"current progress {len(ids)} of {len(nodes)}, "
                    f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
                )
                self._search_client.index_documents(accumulator)
                accumulator.dequeue_actions()
                documents = []
                accumulated_size = 0

        # Upload remaining batch
        if documents:
            logger.info(
                f"Uploading remaining batch of size {len(documents)}, "
                f"current progress {len(ids)} of {len(nodes)}, "
                f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
            )
            self._search_client.index_documents(accumulator)

        return ids

    async def async_add(
        self,
        nodes: List[BaseNode],
        **add_kwargs: Any,
    ) -> List[str]:
        """
        Add nodes to index associated with the configured search client.

        Args:
            nodes: List[BaseNode]: nodes with embeddings

        """
        from azure.search.documents import IndexDocumentsBatch

        if not self._async_search_client:
            raise ValueError("Async Search client not initialized")

        if len(nodes) > 0:
            if self._index_management == IndexManagement.CREATE_IF_NOT_EXISTS:
                if self._index_name:
                    await self._acreate_index_if_not_exists(self._index_name)

            if self._index_management == IndexManagement.VALIDATE_INDEX:
                await self._avalidate_index(self._index_name)

        accumulator = IndexDocumentsBatch()
        documents = []

        ids = []
        accumulated_size = 0
        max_size = 16 * 1024 * 1024  # 16MB in bytes
        max_docs = 1000

        for node in nodes:
            logger.debug(f"Processing embedding: {node.node_id}")
            ids.append(node.node_id)

            index_document = self._create_index_document(node)
            document_size = len(
                str(node.get_content(metadata_mode=MetadataMode.NONE)).encode("utf-8")
            )
            documents.append(index_document)
            accumulated_size += document_size

            accumulator.add_upload_actions(index_document)

            if len(documents) >= max_docs or accumulated_size >= max_size:
                logger.info(
                    f"Uploading batch of size {len(documents)}, "
                    f"current progress {len(ids)} of {len(nodes)}, "
                    f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
                )
                await self._async_search_client.index_documents(accumulator)
                accumulator.dequeue_actions()
                documents = []
                accumulated_size = 0

        # Upload remaining batch
        if documents:
            logger.info(
                f"Uploading remaining batch of size {len(documents)}, "
                f"current progress {len(ids)} of {len(nodes)}, "
                f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
            )
            await self._async_search_client.index_documents(accumulator)

        return ids

    def _create_index_document(self, node: BaseNode) -> Dict[str, Any]:
        """Create AI Search index document from embedding result."""
        doc: Dict[str, Any] = {}
        doc["id"] = node.node_id
        doc["chunk"] = node.get_content(metadata_mode=MetadataMode.NONE) or ""
        doc["embedding"] = node.get_embedding()
        doc["doc_id"] = node.ref_doc_id

        node_metadata = node_to_metadata_dict(
            node,
            remove_text=True,
            flat_metadata=self.flat_metadata,
        )

        doc["metadata"] = json.dumps(node_metadata)

        return self._index_mapping(doc, node_metadata)

    def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """
        Delete documents from the AI Search Index
        with doc_id_field_key field equal to ref_doc_id.
        """
        # Locate documents to delete
        filter = f'{self._field_mapping["doc_id"]} eq \'{ref_doc_id}\''
        results = self._search_client.search(search_text="*", filter=filter)

        logger.debug(f"Searching with filter {filter}")

        docs_to_delete = []
        for result in results:
            doc = {}
            doc["id"] = result[self._field_mapping["id"]]
            logger.debug(f"Found document to delete: {doc}")
            docs_to_delete.append(doc)

        if len(docs_to_delete) > 0:
            logger.debug(f"Deleting {len(docs_to_delete)} documents")
            self._search_client.delete_documents(docs_to_delete)

    async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
        """
        Delete documents from the AI Search Index
        with doc_id_field_key field equal to ref_doc_id.
        """
        # Locate documents to delete
        filter = f'{self._field_mapping["doc_id"]} eq \'{ref_doc_id}\''

        results = await self._async_search_client.search(search_text="*", filter=filter)

        logger.debug(f"Searching with filter {filter}")

        docs_to_delete = []

        for result in results:
            doc = {}
            doc["id"] = result[self._field_mapping["id"]]
            logger.debug(f"Found document to delete: {doc}")
            docs_to_delete.append(doc)

        if len(docs_to_delete) > 0:
            logger.debug(f"Deleting {len(docs_to_delete)} documents")
            await self._search_client.delete_documents(docs_to_delete)

    def _create_odata_filter(self, metadata_filters: MetadataFilters) -> str:
        """Generate an OData filter string using supplied metadata filters."""
        odata_filter: List[str] = []
        for f in metadata_filters.legacy_filters():
            if not isinstance(f, ExactMatchFilter):
                raise NotImplementedError(
                    "Only `ExactMatchFilter` filters are supported"
                )

            # Raise error if filtering on a metadata field that lacks a mapping to
            # an index field
            metadata_mapping = self._metadata_to_index_field_map.get(f.key)

            if not metadata_mapping:
                raise ValueError(
                    f"Metadata field '{f.key}' is missing a mapping to an index field, "
                    "provide entry in 'filterable_metadata_field_keys' for this "
                    "vector store"
                )

            index_field = metadata_mapping[0]

            if len(odata_filter) > 0:
                odata_filter.append(f" {metadata_filters.condition.value} ")
            if isinstance(f.value, str):
                escaped_value = "".join([("''" if s == "'" else s) for s in f.value])
                odata_filter.append(f"{index_field} eq '{escaped_value}'")
            else:
                odata_filter.append(f"{index_field} eq {f.value}")

        odata_expr = "".join(odata_filter)

        logger.info(f"Odata filter: {odata_expr}")

        return odata_expr

    def query(self, query: VectorStoreQuery, **kwargs: Any) -> VectorStoreQueryResult:
        odata_filter = None
        if query.filters is not None:
            odata_filter = self._create_odata_filter(query.filters)
        azure_query_result_search: AzureQueryResultSearchBase = (
            AzureQueryResultSearchDefault(
                query, self._field_mapping, odata_filter, self._search_client
            )
        )
        if query.mode == VectorStoreQueryMode.SPARSE:
            azure_query_result_search = AzureQueryResultSearchSparse(
                query, self._field_mapping, odata_filter, self._search_client
            )
        elif query.mode == VectorStoreQueryMode.HYBRID:
            azure_query_result_search = AzureQueryResultSearchHybrid(
                query, self._field_mapping, odata_filter, self._search_client
            )
        elif query.mode == VectorStoreQueryMode.SEMANTIC_HYBRID:
            azure_query_result_search = AzureQueryResultSearchSemanticHybrid(
                query, self._field_mapping, odata_filter, self._search_client
            )
        return azure_query_result_search.search()

    async def aquery(
        self, query: VectorStoreQuery, **kwargs: Any
    ) -> VectorStoreQueryResult:
        odata_filter = None
        if query.filters is not None:
            odata_filter = self._create_odata_filter(query.filters)
        azure_query_result_search: AzureQueryResultSearchBase = (
            AzureQueryResultSearchDefault(
                query, self._field_mapping, odata_filter, self._async_search_client
            )
        )
        if query.mode == VectorStoreQueryMode.SPARSE:
            azure_query_result_search = AzureQueryResultSearchSparse(
                query, self._field_mapping, odata_filter, self._async_search_client
            )
        elif query.mode == VectorStoreQueryMode.HYBRID:
            azure_query_result_search = AzureQueryResultSearchHybrid(
                query, self._field_mapping, odata_filter, self._async_search_client
            )
        elif query.mode == VectorStoreQueryMode.SEMANTIC_HYBRID:
            azure_query_result_search = AzureQueryResultSearchSemanticHybrid(
                query, self._field_mapping, odata_filter, self._async_search_client
            )
        return await azure_query_result_search.asearch()

client property #

client: Any

Get client.

aclient property #

aclient: Any

Get async client.

add #

add(nodes: List[BaseNode], **add_kwargs: Any) -> List[str]

Add nodes to index associated with the configured search client.

Parameters:

Name Type Description Default
nodes List[BaseNode]

List[BaseNode]: nodes with embeddings

required
Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-azureaisearch/llama_index/vector_stores/azureaisearch/base.py
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
def add(
    self,
    nodes: List[BaseNode],
    **add_kwargs: Any,
) -> List[str]:
    """
    Add nodes to index associated with the configured search client.

    Args:
        nodes: List[BaseNode]: nodes with embeddings

    """
    from azure.search.documents import IndexDocumentsBatch

    if not self._search_client:
        raise ValueError("Search client not initialized")

    accumulator = IndexDocumentsBatch()
    documents = []

    ids = []
    accumulated_size = 0
    max_size = 16 * 1024 * 1024  # 16MB in bytes
    max_docs = 1000

    for node in nodes:
        logger.debug(f"Processing embedding: {node.node_id}")
        ids.append(node.node_id)

        index_document = self._create_index_document(node)
        document_size = len(
            str(node.get_content(metadata_mode=MetadataMode.NONE)).encode("utf-8")
        )
        documents.append(index_document)
        accumulated_size += document_size

        accumulator.add_upload_actions(index_document)

        if len(documents) >= max_docs or accumulated_size >= max_size:
            logger.info(
                f"Uploading batch of size {len(documents)}, "
                f"current progress {len(ids)} of {len(nodes)}, "
                f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
            )
            self._search_client.index_documents(accumulator)
            accumulator.dequeue_actions()
            documents = []
            accumulated_size = 0

    # Upload remaining batch
    if documents:
        logger.info(
            f"Uploading remaining batch of size {len(documents)}, "
            f"current progress {len(ids)} of {len(nodes)}, "
            f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
        )
        self._search_client.index_documents(accumulator)

    return ids

async_add async #

async_add(nodes: List[BaseNode], **add_kwargs: Any) -> List[str]

Add nodes to index associated with the configured search client.

Parameters:

Name Type Description Default
nodes List[BaseNode]

List[BaseNode]: nodes with embeddings

required
Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-azureaisearch/llama_index/vector_stores/azureaisearch/base.py
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
async def async_add(
    self,
    nodes: List[BaseNode],
    **add_kwargs: Any,
) -> List[str]:
    """
    Add nodes to index associated with the configured search client.

    Args:
        nodes: List[BaseNode]: nodes with embeddings

    """
    from azure.search.documents import IndexDocumentsBatch

    if not self._async_search_client:
        raise ValueError("Async Search client not initialized")

    if len(nodes) > 0:
        if self._index_management == IndexManagement.CREATE_IF_NOT_EXISTS:
            if self._index_name:
                await self._acreate_index_if_not_exists(self._index_name)

        if self._index_management == IndexManagement.VALIDATE_INDEX:
            await self._avalidate_index(self._index_name)

    accumulator = IndexDocumentsBatch()
    documents = []

    ids = []
    accumulated_size = 0
    max_size = 16 * 1024 * 1024  # 16MB in bytes
    max_docs = 1000

    for node in nodes:
        logger.debug(f"Processing embedding: {node.node_id}")
        ids.append(node.node_id)

        index_document = self._create_index_document(node)
        document_size = len(
            str(node.get_content(metadata_mode=MetadataMode.NONE)).encode("utf-8")
        )
        documents.append(index_document)
        accumulated_size += document_size

        accumulator.add_upload_actions(index_document)

        if len(documents) >= max_docs or accumulated_size >= max_size:
            logger.info(
                f"Uploading batch of size {len(documents)}, "
                f"current progress {len(ids)} of {len(nodes)}, "
                f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
            )
            await self._async_search_client.index_documents(accumulator)
            accumulator.dequeue_actions()
            documents = []
            accumulated_size = 0

    # Upload remaining batch
    if documents:
        logger.info(
            f"Uploading remaining batch of size {len(documents)}, "
            f"current progress {len(ids)} of {len(nodes)}, "
            f"accumulated size {accumulated_size / (1024 * 1024):.2f} MB"
        )
        await self._async_search_client.index_documents(accumulator)

    return ids

delete #

delete(ref_doc_id: str, **delete_kwargs: Any) -> None

Delete documents from the AI Search Index with doc_id_field_key field equal to ref_doc_id.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-azureaisearch/llama_index/vector_stores/azureaisearch/base.py
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
def delete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """
    Delete documents from the AI Search Index
    with doc_id_field_key field equal to ref_doc_id.
    """
    # Locate documents to delete
    filter = f'{self._field_mapping["doc_id"]} eq \'{ref_doc_id}\''
    results = self._search_client.search(search_text="*", filter=filter)

    logger.debug(f"Searching with filter {filter}")

    docs_to_delete = []
    for result in results:
        doc = {}
        doc["id"] = result[self._field_mapping["id"]]
        logger.debug(f"Found document to delete: {doc}")
        docs_to_delete.append(doc)

    if len(docs_to_delete) > 0:
        logger.debug(f"Deleting {len(docs_to_delete)} documents")
        self._search_client.delete_documents(docs_to_delete)

adelete async #

adelete(ref_doc_id: str, **delete_kwargs: Any) -> None

Delete documents from the AI Search Index with doc_id_field_key field equal to ref_doc_id.

Source code in llama-index-integrations/vector_stores/llama-index-vector-stores-azureaisearch/llama_index/vector_stores/azureaisearch/base.py
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
async def adelete(self, ref_doc_id: str, **delete_kwargs: Any) -> None:
    """
    Delete documents from the AI Search Index
    with doc_id_field_key field equal to ref_doc_id.
    """
    # Locate documents to delete
    filter = f'{self._field_mapping["doc_id"]} eq \'{ref_doc_id}\''

    results = await self._async_search_client.search(search_text="*", filter=filter)

    logger.debug(f"Searching with filter {filter}")

    docs_to_delete = []

    for result in results:
        doc = {}
        doc["id"] = result[self._field_mapping["id"]]
        logger.debug(f"Found document to delete: {doc}")
        docs_to_delete.append(doc)

    if len(docs_to_delete) > 0:
        logger.debug(f"Deleting {len(docs_to_delete)} documents")
        await self._search_client.delete_documents(docs_to_delete)