Add Falkordb vectorstore Integration

llama-index-integrations/vector_stores/llama-index-vector-stores-falkordb/.gitignore

@@ -0,0 +1,29 @@
+# Python
+__pycache__/
+*.py[cod]
+*$py.class
+
+# Virtual Environment
+venv/
+env/
+.env
+
+# IDE files
+.vscode/
+.idea/
+
+# Logs
+*.log
+
+# Build files
+build/
+dist/
+*.egg-info/
+
+# Test files
+.pytest_cache/
+.coverage
+
+# OS generated files
+.DS_Store
+Thumbs.db

llama-index-integrations/vector_stores/llama-index-vector-stores-falkordb/BUILD

@@ -0,0 +1,11 @@
+load("@rules_python//python:defs.bzl", "py_library")
+
+py_library(
+    name = "falkordb",
+    srcs = ["falkordb.py"],
+    deps = [
+        "//llama_index/core:base",
+        "//llama_index/core/vector_stores:types",
+    ],
+    visibility = ["//visibility:public"],
+)

llama-index-integrations/vector_stores/llama-index-vector-stores-falkordb/Makefile

@@ -0,0 +1,20 @@
+.PHONY: all format lint test clean
+
+all: format lint test
+
+format:
+    black .
+    isort .
+
+lint:
+    flake8 .
+    mypy .
+
+test:
+    pytest tests/
+
+clean:
+    find . -type f -name "*.pyc" -delete
+    find . -type d -name "__pycache__" -delete
+    rm -rf .pytest_cache
+    rm -rf .mypy_cache

llama-index-integrations/vector_stores/llama-index-vector-stores-falkordb/README.md

@@ -0,0 +1,2 @@
+# LlamaIndex Vector_Stores Integration: Falkordb
+---

llama-index-integrations/vector_stores/llama-index-vector-stores-falkordb/falkordb-vectorstore-docs.md

@@ -0,0 +1,141 @@
+# FalkorDB Vector Store Integration for LlamaIndex
+
+## Overview
+
+The FalkorDB Vector Store integration for LlamaIndex allows you to use FalkorDB as a vector store backend for your LlamaIndex applications. This integration supports efficient storage and retrieval of vector embeddings, enabling fast similarity searches and other vector operations.
+
+## Installation
+
+To use the FalktorDB vector store with LlamaIndex, you need to install the necessary package:
+
+```bash
+pip install llama-index-vector-stores-falkordb
+```
+
+## Usage
+
+### Initializing the FalkorDB Vector Store
+
+To use the FalkorDB vector store, you first need to initialize it with your FalkorDB connection details:
+
+```python
+from llama_index.vector_stores.falkordb import FalkorDBVectorStore
+
+vector_store = FalkorDBVectorStore(
+    url="bolt://localhost:7687",
+    database="your_database_name",
+    index_name="your_index_name",
+    node_label="YourNodeLabel",
+    embedding_node_property="embedding",
+    text_node_property="text",
+    distance_strategy="cosine",
+    embedding_dimension=1536
+)
+```
+
+### Adding Documents
+
+You can add documents to the vector store using the `add` method:
+
+```python
+from llama_index.core.schema import Document
+
+documents = [
+    Document("This is the first document."),
+    Document("This is the second document.")
+]
+
+vector_store.add(documents)
+```
+
+### Querying the Vector Store
+
+To perform a similarity search, use the `query` method:
+
+```python
+from llama_index.core.vector_stores.types import VectorStoreQuery
+
+query_embedding = [0.1, 0.2, 0.3, ...]  # Your query embedding
+query = VectorStoreQuery(query_embedding=query_embedding, similarity_top_k=5)
+results = vector_store.query(query)
+
+for node, score in zip(results.nodes, results.similarities):
+    print(f"Text: {node.text}, Score: {score}")
+```
+
+## Advanced Features
+
+### Creating a New Index
+
+If you need to create a new vector index, you can use the `create_new_index` method:
+
+```python
+vector_store.create_new_index()
+```
+
+This method will create a new vector index in FalkorDB based on the parameters you provided when initializing the `FalkorDBVectorStore`.
+
+### Retrieving Existing Index
+
+To check if an index already exists and retrieve its information:
+
+```python
+exists = vector_store.retrieve_existing_index()
+if exists:
+    print("Index exists with the following properties:")
+    print(f"Node Label: {vector_store.node_label}")
+    print(f"Embedding Property: {vector_store.embedding_node_property}")
+    print(f"Embedding Dimension: {vector_store.embedding_dimension}")
+    print(f"Distance Strategy: {vector_store.distance_strategy}")
+else:
+    print("Index does not exist")
+```
+
+### Deleting Documents
+
+To delete documents from the vector store:
+
+```python
+ref_doc_id = "your_document_id"
+vector_store.delete(ref_doc_id)
+```
+
+### Using Metadata Filters
+
+You can use metadata filters when querying the vector store:
+
+```python
+from llama_index.core.vector_stores.types import MetadataFilters, MetadataFilter
+
+filters = MetadataFilters(
+    filters=[
+        MetadataFilter(key="category", value="science", operator=FilterOperator.EQ)
+    ]
+)
+
+query = VectorStoreQuery(
+    query_embedding=query_embedding,
+    similarity_top_k=5,
+    filters=filters
+)
+
+results = vector_store.query(query)
+```
+
+## Best Practices
+
+1. **Connection Management**: Ensure that you properly manage your FalkorDB connections, especially in production environments.
+2. **Index Naming**: Use descriptive names for your indexes to easily identify them in your FalkorDB instance.
+3. **Error Handling**: Implement proper error handling to manage potential issues with connections or queries.
+4. **Performance Tuning**: Adjust the `embedding_dimension` and `distance_strategy` parameters based on your specific use case and performance requirements.
+
+## Troubleshooting
+
+If you encounter issues:
+
+1. Check your FalkorDB connection details (URL, database name).
+2. Ensure that the FalkorDB server is running and accessible.
+3. Verify that the index exists and has the correct properties.
+4. Check the FalkorDB logs for any error messages.
+
+For more information on FalkorDB and its capabilities, refer to the [official FalkorDB documentation](https://falkordb.com/docs/).