Skip to content

Neo4jRetriever

Neo4jEmbeddingRetriever ¤

A component for retrieving documents from Neo4jDocumentStore.

Retrieving documents assuming documents have been previously indexed
from haystack import Document, Pipeline
from haystack.components.embedders import SentenceTransformersTextEmbedder

from neo4j_haystack import Neo4jDocumentStore, Neo4jEmbeddingRetriever

model_name = "sentence-transformers/all-MiniLM-L6-v2"

# Document store with default credentials
document_store = Neo4jDocumentStore(
    url="bolt://localhost:7687",
    embedding_dim=384, # same as the embedding model
)

pipeline = Pipeline()
pipeline.add_component("text_embedder", SentenceTransformersTextEmbedder(model=model_name))
pipeline.add_component("retriever", Neo4jEmbeddingRetriever(document_store=document_store))
pipeline.connect("text_embedder.embedding", "retriever.query_embedding")

result = pipeline.run(
    data={
        "text_embedder": {"text": "Query to be embedded"},
        "retriever": {
            "top_k": 5,
            "filters": {"field": "release_date", "operator": "==", "value": "2018-12-09"},
        },
    }
)

# Obtain retrieved documents from pipeline execution
documents: List[Document] = result["retriever"]["documents"]
Source code in src/neo4j_haystack/components/neo4j_retriever.py 
 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
@component
class Neo4jEmbeddingRetriever:
    """
    A component for retrieving documents from Neo4jDocumentStore.

    ```py title="Retrieving documents assuming documents have been previously indexed"
    from haystack import Document, Pipeline
    from haystack.components.embedders import SentenceTransformersTextEmbedder

    from neo4j_haystack import Neo4jDocumentStore, Neo4jEmbeddingRetriever

    model_name = "sentence-transformers/all-MiniLM-L6-v2"

    # Document store with default credentials
    document_store = Neo4jDocumentStore(
        url="bolt://localhost:7687",
        embedding_dim=384, # same as the embedding model
    )

    pipeline = Pipeline()
    pipeline.add_component("text_embedder", SentenceTransformersTextEmbedder(model=model_name))
    pipeline.add_component("retriever", Neo4jEmbeddingRetriever(document_store=document_store))
    pipeline.connect("text_embedder.embedding", "retriever.query_embedding")

    result = pipeline.run(
        data={
            "text_embedder": {"text": "Query to be embedded"},
            "retriever": {
                "top_k": 5,
                "filters": {"field": "release_date", "operator": "==", "value": "2018-12-09"},
            },
        }
    )

    # Obtain retrieved documents from pipeline execution
    documents: List[Document] = result["retriever"]["documents"]
    ```
    """

    def __init__(
        self,
        document_store: Neo4jDocumentStore,
        filters: Optional[Dict[str, Any]] = None,
        top_k: int = 10,
        scale_score: bool = True,
        return_embedding: bool = False,
    ):
        """
        Create a Neo4jEmbeddingRetriever component.

        Args:
            document_store: An instance of `Neo4jDocumentStore`.
            filters: A dictionary with filters to narrow down the search space.
            top_k: The maximum number of documents to retrieve.
            scale_score: Whether to scale the scores of the retrieved documents or not.
            return_embedding: Whether to return the embedding of the retrieved Documents.

        Raises:
            ValueError: If `document_store` is not an instance of `Neo4jDocumentStore`.
        """

        if not isinstance(document_store, Neo4jDocumentStore):
            msg = "document_store must be an instance of Neo4jDocumentStore"
            raise ValueError(msg)

        self._document_store = document_store

        self._filters = filters
        self._top_k = top_k
        self._scale_score = scale_score
        self._return_embedding = return_embedding

    def to_dict(self) -> Dict[str, Any]:
        """
        Serialize this component to a dictionary.
        """
        data = default_to_dict(
            self,
            document_store=self._document_store,
            filters=self._filters,
            top_k=self._top_k,
            scale_score=self._scale_score,
            return_embedding=self._return_embedding,
        )
        data["init_parameters"]["document_store"] = self._document_store.to_dict()

        return data

    @classmethod
    def from_dict(cls, data: Dict[str, Any]) -> "Neo4jEmbeddingRetriever":
        """
        Deserialize this component from a dictionary.
        """
        document_store = Neo4jDocumentStore.from_dict(data["init_parameters"]["document_store"])
        data["init_parameters"]["document_store"] = document_store
        return default_from_dict(cls, data)

    @component.output_types(documents=List[Document])
    def run(
        self,
        query_embedding: List[float],
        filters: Optional[Dict[str, Any]] = None,
        top_k: Optional[int] = None,
        scale_score: Optional[bool] = None,
        return_embedding: Optional[bool] = None,
    ):
        """
        Run the Embedding Retriever on the given input data.

        Args:
            query_embedding: Embedding of the query.
            filters: A dictionary with filters to narrow down the search space.
            top_k: The maximum number of documents to return.
            scale_score: Whether to scale the scores of the retrieved documents or not.
            return_embedding: Whether to return the embedding of the retrieved Documents.

        Returns:
            The retrieved documents.
        """
        docs = self._document_store.query_by_embedding(
            query_embedding=query_embedding,
            filters=filters or self._filters,
            top_k=top_k or self._top_k,
            scale_score=scale_score or self._scale_score,
            return_embedding=return_embedding or self._return_embedding,
        )

        return {"documents": docs}

__init__ ¤

__init__(
    document_store: Neo4jDocumentStore,
    filters: Optional[Dict[str, Any]] = None,
    top_k: int = 10,
    scale_score: bool = True,
    return_embedding: bool = False,
)

Parameters:

  • document_store (Neo4jDocumentStore) –

    An instance of Neo4jDocumentStore.

  • filters (Optional[Dict[str, Any]], default: None ) –

    A dictionary with filters to narrow down the search space.

  • top_k (int, default: 10 ) –

    The maximum number of documents to retrieve.

  • scale_score (bool, default: True ) –

    Whether to scale the scores of the retrieved documents or not.

  • return_embedding (bool, default: False ) –

    Whether to return the embedding of the retrieved Documents.

Raises:

  • ValueError –

    If document_store is not an instance of Neo4jDocumentStore.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
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
def __init__(
    self,
    document_store: Neo4jDocumentStore,
    filters: Optional[Dict[str, Any]] = None,
    top_k: int = 10,
    scale_score: bool = True,
    return_embedding: bool = False,
):
    """
    Create a Neo4jEmbeddingRetriever component.

    Args:
        document_store: An instance of `Neo4jDocumentStore`.
        filters: A dictionary with filters to narrow down the search space.
        top_k: The maximum number of documents to retrieve.
        scale_score: Whether to scale the scores of the retrieved documents or not.
        return_embedding: Whether to return the embedding of the retrieved Documents.

    Raises:
        ValueError: If `document_store` is not an instance of `Neo4jDocumentStore`.
    """

    if not isinstance(document_store, Neo4jDocumentStore):
        msg = "document_store must be an instance of Neo4jDocumentStore"
        raise ValueError(msg)

    self._document_store = document_store

    self._filters = filters
    self._top_k = top_k
    self._scale_score = scale_score
    self._return_embedding = return_embedding

to_dict ¤

to_dict() -> Dict[str, Any]

Serialize this component to a dictionary.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
def to_dict(self) -> Dict[str, Any]:
    """
    Serialize this component to a dictionary.
    """
    data = default_to_dict(
        self,
        document_store=self._document_store,
        filters=self._filters,
        top_k=self._top_k,
        scale_score=self._scale_score,
        return_embedding=self._return_embedding,
    )
    data["init_parameters"]["document_store"] = self._document_store.to_dict()

    return data

from_dict classmethod ¤

from_dict(data: Dict[str, Any]) -> Neo4jEmbeddingRetriever

Deserialize this component from a dictionary.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
103
104
105
106
107
108
109
110
@classmethod
def from_dict(cls, data: Dict[str, Any]) -> "Neo4jEmbeddingRetriever":
    """
    Deserialize this component from a dictionary.
    """
    document_store = Neo4jDocumentStore.from_dict(data["init_parameters"]["document_store"])
    data["init_parameters"]["document_store"] = document_store
    return default_from_dict(cls, data)

run ¤

run(
    query_embedding: List[float],
    filters: Optional[Dict[str, Any]] = None,
    top_k: Optional[int] = None,
    scale_score: Optional[bool] = None,
    return_embedding: Optional[bool] = None,
)

Run the Embedding Retriever on the given input data.

Parameters:

  • query_embedding (List[float]) –

    Embedding of the query.

  • filters (Optional[Dict[str, Any]], default: None ) –

    A dictionary with filters to narrow down the search space.

  • top_k (Optional[int], default: None ) –

    The maximum number of documents to return.

  • scale_score (Optional[bool], default: None ) –

    Whether to scale the scores of the retrieved documents or not.

  • return_embedding (Optional[bool], default: None ) –

    Whether to return the embedding of the retrieved Documents.

Returns:

  • –

    The retrieved documents.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
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
@component.output_types(documents=List[Document])
def run(
    self,
    query_embedding: List[float],
    filters: Optional[Dict[str, Any]] = None,
    top_k: Optional[int] = None,
    scale_score: Optional[bool] = None,
    return_embedding: Optional[bool] = None,
):
    """
    Run the Embedding Retriever on the given input data.

    Args:
        query_embedding: Embedding of the query.
        filters: A dictionary with filters to narrow down the search space.
        top_k: The maximum number of documents to return.
        scale_score: Whether to scale the scores of the retrieved documents or not.
        return_embedding: Whether to return the embedding of the retrieved Documents.

    Returns:
        The retrieved documents.
    """
    docs = self._document_store.query_by_embedding(
        query_embedding=query_embedding,
        filters=filters or self._filters,
        top_k=top_k or self._top_k,
        scale_score=scale_score or self._scale_score,
        return_embedding=return_embedding or self._return_embedding,
    )

    return {"documents": docs}

Neo4jDynamicDocumentRetriever ¤

A component for retrieving Documents from Neo4j database using plain Cypher query.

This component gives flexible way to retrieve data from Neo4j by running arbitrary Cypher query along with query parameters. Query parameters can be supplied in a pipeline from other components (or pipeline data).

See the following documentation on how to compose Cypher queries with parameters:

Above are resources which will help understand better Cypher query syntax and parameterization. Under the hood Neo4j Python Driver is used to query database and fetch results. You might be interested in the following documentation:

Note

Please consider data types mappings in Cypher query when working with parameters. Neo4j Python Driver handles type conversions/mappings. Specifically you can figure out in the documentation of the driver how to work with temporal types (e.g. DateTime).

Query execution results will be mapped/converted to haystack.Document type. See more details in the RETURN clause documentation. There are two ways how Documents are being composed from query results.

(1) Converting documents from nodes

Convert Neo4j `node` to `haystack.Document`
client_config = Neo4jClientConfig(
    "bolt://localhost:7687", database="neo4j", username="neo4j", password="passw0rd"
)

retriever = Neo4jDynamicDocumentRetriever(
    client_config=client_config, doc_node_name="doc", verify_connectivity=True
)

result = retriever.run(
    query="MATCH (doc:Document) WHERE doc.year > $year OR doc.year is NULL RETURN doc",
    parameters={"year": 2020}
)
documents: List[Document] = result["documents"]

Please notice how doc_node_name attribute assumes "doc" node is going to be returned from the query. Neo4jDynamicDocumentRetriever will convert properties of the node (e.g. id, content etc) to haystack.Document type.

(2) Converting documents from query output keys (e.g. column aliases)

You might want to run a complex query which aggregates information from multiple sources (nodes) in Neo4j. In such case you can compose final Document from several dta points.

Convert Neo4j `node` to `haystack.Document`
# Configuration with default settings
client_config=Neo4jClientConfig()

retriever = Neo4jDynamicDocumentRetriever(client_config=client_config, compose_doc_from_result=True)

result = retriever.run(
    query=(
        "MATCH (doc:Document) "
        "WHERE doc.year > $year OR doc.year is NULL "
        "RETURN doc.id as id, doc.content as content, doc.year as year"
    ),
    parameters={"year": 2020},
)
documents: List[Document] = result["documents"]

The above will produce Documents with id, content and year(meta) fields. Please notice compose_doc_from_result is set to True to enable such Document construction behavior.

Below is an example of a pipeline which explores all ways how parameters could be supplied to the Neo4jDynamicDocumentRetriever component in the pipeline.

@component
class YearProvider:
    @component.output_types(year_start=int, year_end=int)
    def run(self, year_start: int, year_end: int):
        return {"year_start": year_start, "year_end": year_end}

# Configuration with default settings
client_config=Neo4jClientConfig()

retriever = Neo4jDynamicDocumentRetriever(
    client_config=client_config,
    runtime_parameters=["year_start", "year_end"],
)

query = (
    "MATCH (doc:Document) "
    "WHERE (doc.year >= $year_start and doc.year <= $year_end) AND doc.month = $month"
    "RETURN doc LIMIT $num_return"
)

pipeline = Pipeline()
pipeline.add_component("year_provider", YearProvider())
pipeline.add_component("retriever", retriever)
pipeline.connect("year_provider.year_start", "retriever.year_start")
pipeline.connect("year_provider.year_end", "retriever.year_end")

result = pipeline.run(
    data={
        "year_provider": {"year_start": 2020, "year_end": 2021},
        "retriever": {
            "query": query,
            "parameters": {
                "month": "02",
                "num_return": 2,
            },
        },
    }
)

documents = result["retriever"]["documents"]

Please notice the following from the example above:

  • runtime_parameters is a list of parameter names which are going to be input slots when connecting components in a pipeline. In our case year_start and year_end parameters flow from the year_provider component into retriever. The query uses those parameters in the WHERE clause.
  • pipeline.run specifies additional parameters to the retriever component which can be referenced in the query. If parameter names clash those provided in the pipeline's data take precedence.
Source code in src/neo4j_haystack/components/neo4j_retriever.py 
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
@component
class Neo4jDynamicDocumentRetriever:
    """
    A component for retrieving Documents from Neo4j database using plain Cypher query.

    This component gives flexible way to retrieve data from Neo4j by running arbitrary Cypher query along with query
    parameters. Query parameters can be supplied in a pipeline from other components (or pipeline data).

    See the following documentation on how to compose Cypher queries with parameters:

    - [Overview of Cypher query syntax](https://neo4j.com/docs/cypher-manual/current/queries/)
    - [Cypher Query Parameters](https://neo4j.com/docs/cypher-manual/current/syntax/parameters/)

    Above are resources which will help understand better Cypher query syntax and parameterization. Under the hood
    [Neo4j Python Driver](https://neo4j.com/docs/python-manual/current/) is used to query database and fetch results.
    You might be interested in the following documentation:

    - [Query the database](https://neo4j.com/docs/python-manual/current/query-simple/)
    - [Query parameters](https://neo4j.com/docs/python-manual/current/query-simple/#query-parameters)
    - [Data types and mapping to Cypher types](https://neo4j.com/docs/python-manual/current/data-types/)

    Note:
        Please consider data types mappings in Cypher query when working with parameters. Neo4j Python Driver handles
        type conversions/mappings. Specifically you can figure out in the documentation of the driver how to work with
        temporal types (e.g. `DateTime`).

    Query execution results will be mapped/converted to `haystack.Document` type. See more details in the
    [RETURN clause](https://neo4j.com/docs/cypher-manual/current/clauses/return/) documentation. There are two
    ways how Documents are being composed from query results.

    (1) Converting documents from [nodes](https://neo4j.com/docs/cypher-manual/current/clauses/return/#return-nodes)

    ```py title="Convert Neo4j `node` to `haystack.Document`"
    client_config = Neo4jClientConfig(
        "bolt://localhost:7687", database="neo4j", username="neo4j", password="passw0rd"
    )

    retriever = Neo4jDynamicDocumentRetriever(
        client_config=client_config, doc_node_name="doc", verify_connectivity=True
    )

    result = retriever.run(
        query="MATCH (doc:Document) WHERE doc.year > $year OR doc.year is NULL RETURN doc",
        parameters={"year": 2020}
    )
    documents: List[Document] = result["documents"]
    ```

    Please notice how `doc_node_name` attribute assumes `"doc"` node is going to be returned from the query.
    `Neo4jDynamicDocumentRetriever` will convert properties of the node (e.g. `id`, `content` etc) to
    `haystack.Document` type.

    (2) Converting documents from query output keys (e.g. [column aliases](https://neo4j.com/docs/cypher-manual/current/clauses/return/#return-column-alias))

    You might want to run a complex query which aggregates information from multiple sources (nodes) in Neo4j. In such
    case you can compose final Document from several dta points.

    ```py title="Convert Neo4j `node` to `haystack.Document`"
    # Configuration with default settings
    client_config=Neo4jClientConfig()

    retriever = Neo4jDynamicDocumentRetriever(client_config=client_config, compose_doc_from_result=True)

    result = retriever.run(
        query=(
            "MATCH (doc:Document) "
            "WHERE doc.year > $year OR doc.year is NULL "
            "RETURN doc.id as id, doc.content as content, doc.year as year"
        ),
        parameters={"year": 2020},
    )
    documents: List[Document] = result["documents"]
    ```

    The above will produce Documents with `id`, `content` and `year`(meta) fields. Please notice
    `compose_doc_from_result` is set to `True` to enable such Document construction behavior.

    Below is an example of a pipeline which explores all ways how parameters could be supplied to the
    `Neo4jDynamicDocumentRetriever` component in the pipeline.

    ```py
    @component
    class YearProvider:
        @component.output_types(year_start=int, year_end=int)
        def run(self, year_start: int, year_end: int):
            return {"year_start": year_start, "year_end": year_end}

    # Configuration with default settings
    client_config=Neo4jClientConfig()

    retriever = Neo4jDynamicDocumentRetriever(
        client_config=client_config,
        runtime_parameters=["year_start", "year_end"],
    )

    query = (
        "MATCH (doc:Document) "
        "WHERE (doc.year >= $year_start and doc.year <= $year_end) AND doc.month = $month"
        "RETURN doc LIMIT $num_return"
    )

    pipeline = Pipeline()
    pipeline.add_component("year_provider", YearProvider())
    pipeline.add_component("retriever", retriever)
    pipeline.connect("year_provider.year_start", "retriever.year_start")
    pipeline.connect("year_provider.year_end", "retriever.year_end")

    result = pipeline.run(
        data={
            "year_provider": {"year_start": 2020, "year_end": 2021},
            "retriever": {
                "query": query,
                "parameters": {
                    "month": "02",
                    "num_return": 2,
                },
            },
        }
    )

    documents = result["retriever"]["documents"]
    ```

    Please notice the following from the example above:

    - `runtime_parameters` is a list of parameter names which are going to be input slots when connecting components
        in a pipeline. In our case `year_start` and `year_end` parameters flow from the `year_provider` component into
        `retriever`. The `query` uses those parameters in the `WHERE` clause.
    - `pipeline.run` specifies additional parameters to the `retriever` component which can be referenced in the
        `query`. If parameter names clash those provided in the pipeline's data take precedence.
    """

    def __init__(
        self,
        client_config: Neo4jClientConfig,
        runtime_parameters: Optional[List[str]] = None,
        doc_node_name: Optional[str] = "doc",
        compose_doc_from_result: Optional[bool] = False,
        verify_connectivity: Optional[bool] = False,
    ):
        """
        Create a Neo4jDynamicDocumentRetriever component.

        Args:
            client_config: Neo4j client configuration to connect to database (e.g. credentials and connection settings).
            runtime_parameters: list of input parameters/slots for connecting components in a pipeline.
            doc_node_name: the name of the variable which is returned from Cypher query which contains Document
                attributes (e.g. `id`, `content`, `meta` fields).
            compose_doc_from_result: If `True` Document attributes will be constructed from Cypher query outputs (keys).
                `doc_node_name` setting will be ignored in this case.
            verify_connectivity: If `True` will verify connectivity with Neo4j database configured by `client_config`.

        Raises:
            ComponentError: In case neither `compose_doc_from_result` nor `doc_node_name` are defined.
        """
        if not compose_doc_from_result and not doc_node_name:
            raise ComponentError(
                "Please specify how Document is being composed out of Neo4j query response. "
                "With `compose_doc_from_result` set to `True` documents will be created out of properties/keys "
                "returned by the query."
            )

        self._client_config = client_config
        self._runtime_parameters = runtime_parameters or []
        self._doc_node_name = doc_node_name
        self._compose_doc_from_result = compose_doc_from_result
        self._verify_connectivity = verify_connectivity

        self._neo4j_client = Neo4jClient(client_config)

        # setup inputs
        run_input_slots = {"query": str, "parameters": Optional[Dict[str, Any]]}
        kwargs_input_slots = {param: Optional[Any] for param in self._runtime_parameters}
        component.set_input_types(self, **run_input_slots, **kwargs_input_slots)

        # setup outputs
        component.set_output_types(self, documents=List[Document])

        if verify_connectivity:
            self._neo4j_client.verify_connectivity()

    def to_dict(self) -> Dict[str, Any]:
        """
        Serialize this component to a dictionary.
        """
        data = default_to_dict(
            self,
            runtime_parameters=self._runtime_parameters,
            doc_node_name=self._doc_node_name,
            compose_doc_from_result=self._compose_doc_from_result,
            verify_connectivity=self._verify_connectivity,
        )

        data["init_parameters"]["client_config"] = self._client_config.to_dict()

        return data

    @classmethod
    def from_dict(cls, data: Dict[str, Any]) -> "Neo4jDynamicDocumentRetriever":
        """
        Deserialize this component from a dictionary.
        """
        client_config = Neo4jClientConfig.from_dict(data["init_parameters"]["client_config"])
        data["init_parameters"]["client_config"] = client_config
        return default_from_dict(cls, data)

    def run(self, query: str, parameters: Optional[Dict[str, Any]] = None, **kwargs: Dict[str, Any]):
        """
        Runs the arbitrary Cypher `query` with `parameters` and returns Documents.

        Args:
            query: Cypher query to run.
            parameters: Cypher query parameters which can be used as placeholders in the `query`.
            kwargs: Arbitrary parameters supplied in a pipeline execution from other component's output slots, e.g.
                `pipeline.connect("year_provider.year_start", "retriever.year_start")`, where `year_start` will be part
                of `kwargs`.

        Returns:
            Retrieved documents.
        """
        kwargs = kwargs or {}
        parameters = parameters or {}
        parameters_combined = {**kwargs, **parameters}

        documents: List[Document] = []
        neo4j_query_result = self._neo4j_client.query_nodes(query, parameters_combined)

        for record in neo4j_query_result:
            data = record.data()
            document_dict = data if self._compose_doc_from_result else data.get(cast(str, self._doc_node_name))
            documents.append(Document.from_dict(document_dict))

        return {"documents": documents}

__init__ ¤

__init__(
    client_config: Neo4jClientConfig,
    runtime_parameters: Optional[List[str]] = None,
    doc_node_name: Optional[str] = "doc",
    compose_doc_from_result: Optional[bool] = False,
    verify_connectivity: Optional[bool] = False,
)

Parameters:

  • client_config (Neo4jClientConfig) –

    Neo4j client configuration to connect to database (e.g. credentials and connection settings).

  • runtime_parameters (Optional[List[str]], default: None ) –

    list of input parameters/slots for connecting components in a pipeline.

  • doc_node_name (Optional[str], default: 'doc' ) –

    the name of the variable which is returned from Cypher query which contains Document attributes (e.g. id, content, meta fields).

  • compose_doc_from_result (Optional[bool], default: False ) –

    If True Document attributes will be constructed from Cypher query outputs (keys). doc_node_name setting will be ignored in this case.

  • verify_connectivity (Optional[bool], default: False ) –

    If True will verify connectivity with Neo4j database configured by client_config.

Raises:

  • ComponentError –

    In case neither compose_doc_from_result nor doc_node_name are defined.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
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
def __init__(
    self,
    client_config: Neo4jClientConfig,
    runtime_parameters: Optional[List[str]] = None,
    doc_node_name: Optional[str] = "doc",
    compose_doc_from_result: Optional[bool] = False,
    verify_connectivity: Optional[bool] = False,
):
    """
    Create a Neo4jDynamicDocumentRetriever component.

    Args:
        client_config: Neo4j client configuration to connect to database (e.g. credentials and connection settings).
        runtime_parameters: list of input parameters/slots for connecting components in a pipeline.
        doc_node_name: the name of the variable which is returned from Cypher query which contains Document
            attributes (e.g. `id`, `content`, `meta` fields).
        compose_doc_from_result: If `True` Document attributes will be constructed from Cypher query outputs (keys).
            `doc_node_name` setting will be ignored in this case.
        verify_connectivity: If `True` will verify connectivity with Neo4j database configured by `client_config`.

    Raises:
        ComponentError: In case neither `compose_doc_from_result` nor `doc_node_name` are defined.
    """
    if not compose_doc_from_result and not doc_node_name:
        raise ComponentError(
            "Please specify how Document is being composed out of Neo4j query response. "
            "With `compose_doc_from_result` set to `True` documents will be created out of properties/keys "
            "returned by the query."
        )

    self._client_config = client_config
    self._runtime_parameters = runtime_parameters or []
    self._doc_node_name = doc_node_name
    self._compose_doc_from_result = compose_doc_from_result
    self._verify_connectivity = verify_connectivity

    self._neo4j_client = Neo4jClient(client_config)

    # setup inputs
    run_input_slots = {"query": str, "parameters": Optional[Dict[str, Any]]}
    kwargs_input_slots = {param: Optional[Any] for param in self._runtime_parameters}
    component.set_input_types(self, **run_input_slots, **kwargs_input_slots)

    # setup outputs
    component.set_output_types(self, documents=List[Document])

    if verify_connectivity:
        self._neo4j_client.verify_connectivity()

to_dict ¤

to_dict() -> Dict[str, Any]

Serialize this component to a dictionary.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
def to_dict(self) -> Dict[str, Any]:
    """
    Serialize this component to a dictionary.
    """
    data = default_to_dict(
        self,
        runtime_parameters=self._runtime_parameters,
        doc_node_name=self._doc_node_name,
        compose_doc_from_result=self._compose_doc_from_result,
        verify_connectivity=self._verify_connectivity,
    )

    data["init_parameters"]["client_config"] = self._client_config.to_dict()

    return data

from_dict classmethod ¤

from_dict(data: Dict[str, Any]) -> Neo4jDynamicDocumentRetriever

Deserialize this component from a dictionary.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
342
343
344
345
346
347
348
349
@classmethod
def from_dict(cls, data: Dict[str, Any]) -> "Neo4jDynamicDocumentRetriever":
    """
    Deserialize this component from a dictionary.
    """
    client_config = Neo4jClientConfig.from_dict(data["init_parameters"]["client_config"])
    data["init_parameters"]["client_config"] = client_config
    return default_from_dict(cls, data)

run ¤

run(
    query: str,
    parameters: Optional[Dict[str, Any]] = None,
    **kwargs: Dict[str, Any]
)

Runs the arbitrary Cypher query with parameters and returns Documents.

Parameters:

  • query (str) –

    Cypher query to run.

  • parameters (Optional[Dict[str, Any]], default: None ) –

    Cypher query parameters which can be used as placeholders in the query.

  • kwargs (Dict[str, Any], default: {} ) –

    Arbitrary parameters supplied in a pipeline execution from other component's output slots, e.g. pipeline.connect("year_provider.year_start", "retriever.year_start"), where year_start will be part of kwargs.

Returns:

  • –

    Retrieved documents.

Source code in src/neo4j_haystack/components/neo4j_retriever.py 
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
def run(self, query: str, parameters: Optional[Dict[str, Any]] = None, **kwargs: Dict[str, Any]):
    """
    Runs the arbitrary Cypher `query` with `parameters` and returns Documents.

    Args:
        query: Cypher query to run.
        parameters: Cypher query parameters which can be used as placeholders in the `query`.
        kwargs: Arbitrary parameters supplied in a pipeline execution from other component's output slots, e.g.
            `pipeline.connect("year_provider.year_start", "retriever.year_start")`, where `year_start` will be part
            of `kwargs`.

    Returns:
        Retrieved documents.
    """
    kwargs = kwargs or {}
    parameters = parameters or {}
    parameters_combined = {**kwargs, **parameters}

    documents: List[Document] = []
    neo4j_query_result = self._neo4j_client.query_nodes(query, parameters_combined)

    for record in neo4j_query_result:
        data = record.data()
        document_dict = data if self._compose_doc_from_result else data.get(cast(str, self._doc_node_name))
        documents.append(Document.from_dict(document_dict))

    return {"documents": documents}