Refactors index functions (#269)

* Refactors _get_hybrid_query

* Added effective search ratio to vector and hybrid searches

* Removes id warning from get_search_query

* Removed 'elementId(node) AS id' from tests

* Refactors vector and full text index search queries to support relationships

* get_search_query refactoring

* Reverted embedding_node_property change

* More reversions

* Reverted _get_hybrid_query

* Small bug fix

* Added vector relationship query tests

* Added use_parallel_runtime to _get_filtered_vector_query

* Added remove_lucene_chars

* Fixed remove_lucene_chars bugs

* Renamed remove lucene chars function

* Added retrieve_vector_index_info and retrieve_fulltext_index_info

* Fixed get_search_query bug

* Docstring updates

* Adds upsert_texts_and_vectors utility function

* Updates docs

* Added NODES_MISSING_EMBEDDINGS_QUERY

* Renamed upsert_texts_and_vectors

* Removes indexes_2

* Removes upsert_texts_and_embeddings

* Renamed IndexType to EntityType

* Renaming params for get_search_query

* Removed DistanceStrategy and updated CHANGELOG

* Hybrid retriever no longer sanitises text before embedding it

* Re-added id parameter to retrievers

* Readded deprecation warning

* Removed the _remove_lucene_chars function

* Removed unneeded lines

* Small bug fix

Author: alexthomas93

CHANGELOG.md

@@ -2,6 +2,15 @@
 
 ## Next
 
+### Added
+
+- Utility functions to retrieve metadata for vector and full-text indexes.
+- Support for effective_search_ratio parameter in vector and hybrid searches.
+
+### Changed
+
+- Refactored index-related functions for improved compatibility and functionality.
+
 ## 1.4.3
 
 ### Added
@@ -18,6 +27,7 @@
 - Refactored schema creation code to reduce duplication and improve maintainability.
 
 ### Fixed
+
 - Removed the `uuid` package from dependencies (not needed with Python 3).
 - Fixed a bug in the `AnthropicLLM` class preventing it from being used in `GraphRAG` pipeline.
 

docs/source/api.rst

@@ -390,6 +390,10 @@ Database Interaction
 
 .. autofunction:: neo4j_graphrag.indexes.async_upsert_vector_on_relationship
 
+.. autofunction:: neo4j_graphrag.indexes.retrieve_vector_index_info
+
+.. autofunction:: neo4j_graphrag.indexes.retrieve_fulltext_index_info
+
 .. autofunction:: neo4j_graphrag.schema.get_structured_schema
 
 .. autofunction:: neo4j_graphrag.schema.get_schema

src/neo4j_graphrag/indexes.py

@@ -15,7 +15,7 @@
 from __future__ import annotations
 
 import logging
-from typing import Literal, Optional
+from typing import List, Literal, Optional
 
 import neo4j
 from pydantic import ValidationError
@@ -469,3 +469,108 @@ async def async_upsert_vector_on_relationship(
         raise Neo4jInsertionError(
             f"Upserting vector to Neo4j failed: {e.message}"
         ) from e
+
+
+def _sort_by_index_name(
+    records: List[neo4j.Record], index_name: str
+) -> List[neo4j.Record]:
+    """
+    Sorts the provided list of dictionaries containing index information so
+    that any item whose 'name' key matches the given 'index_name' appears at
+    the front of the list.
+
+    Args:
+        records (List[Dict[str, Any]]): The list of records containing index
+            information to sort.
+        index_name (str): The index name to match against the 'name' key of
+            each dictionary.
+
+    Returns:
+        List[Dict[str, Any]]: A newly sorted list with items matching
+        'index_name' placed first.
+    """
+    return sorted(records, key=lambda x: x.get("name") != index_name)
+
+
+def retrieve_vector_index_info(
+    driver: neo4j.Driver, index_name: str, label_or_type: str, embedding_property: str
+) -> Optional[neo4j.Record]:
+    """
+    Check if a vector index exists in a Neo4j database and return its
+    information. If no matching index is found, returns None.
+
+    Args:
+        driver (neo4j.Driver): Neo4j Python driver instance.
+        index_name (str): The name of the index to look up.
+        label_or_type (str): The label (for nodes) or type (for relationships)
+            of the index.
+        embedding_property (str): The name of the property containing the
+            embeddings.
+
+    Returns:
+        Optional[Dict[str, Any]]:
+            A dictionary containing the first matching index's information if found,
+            or None otherwise.
+    """
+    result = driver.execute_query(
+        query_=(
+            "SHOW INDEXES YIELD name, type, entityType, labelsOrTypes, "
+            "properties, options WHERE type = 'VECTOR' AND (name = $index_name "
+            "OR (labelsOrTypes[0] = $label_or_type AND "
+            "properties[0] = $embedding_property)) "
+            "RETURN name, type, entityType, labelsOrTypes, properties, options"
+        ),
+        parameters_={
+            "index_name": index_name,
+            "label_or_type": label_or_type,
+            "embedding_property": embedding_property,
+        },
+    )
+    index_information = _sort_by_index_name(result.records, index_name)
+    if len(index_information) > 0:
+        return index_information[0]
+    else:
+        return None
+
+
+def retrieve_fulltext_index_info(
+    driver: neo4j.Driver,
+    index_name: str,
+    label_or_type: str,
+    text_properties: List[str] = [],
+) -> Optional[neo4j.Record]:
+    """
+    Check if a full text index exists in a Neo4j database and return its
+    information. If no matching index is found, returns None.
+
+    Args:
+        driver (neo4j.Driver): Neo4j Python driver instance.
+        index_name (str): The name of the index to look up.
+        label_or_type (str): The label (for nodes) or type (for relationships)
+            of the index.
+        text_properties (List[str]): The names of the text properties indexed.
+
+    Returns:
+        Optional[Dict[str, Any]]:
+            A dictionary containing the first matching index's information if found,
+            or None otherwise.
+    """
+    result = driver.execute_query(
+        query_=(
+            "SHOW INDEXES YIELD name, type, entityType, labelsOrTypes, properties, options "
+            "WHERE type = 'FULLTEXT' AND (name = $index_name "
+            "OR (labelsOrTypes = [$label_or_type] AND "
+            "properties = $text_properties)) "
+            "RETURN name, type, entityType, labelsOrTypes, properties, options"
+        ),
+        parameters_={
+            "index_name": index_name,
+            "label_or_type": label_or_type,
+            "text_properties": text_properties,
+        },
+    )
+    index_information = _sort_by_index_name(result.records, index_name)
+    if len(index_information) > 0:
+        return index_information[0]
+    else:
+        return None