full

Author: retroryan

financial_data_load/full_data_load.py

@@ -27,7 +27,7 @@
 - Azure AI Foundry deployed (run: azd up && uv run python setup_env.py)
 - Azure CLI logged in (az login --use-device-code)
 - Neo4j connection configured in .env
-- PDFs in DATA_DIR (defaults to ~/projects/workshops/workshop-financial-data)
+- PDFs in financial-data/form10k-sample (relative to this script)
 """
 
 import asyncio
@@ -58,8 +58,8 @@
 logging.basicConfig(level=logging.INFO)
 logger = logging.getLogger(__name__)
 
-# Data directory - adjust this path as needed
-DATA_DIR = Path.home() / "projects" / "workshops" / "workshop-financial-data"
+# Data directory - relative to this script
+DATA_DIR = Path(__file__).parent / "financial-data"
 PDF_DIR = DATA_DIR / "form10k-sample"
 COMPANY_CSV = DATA_DIR / "Company_Filings.csv"
 ASSET_MANAGER_CSV = DATA_DIR / "Asset_Manager_Holdings.csv"

financial_data_load/src/01_01_data_loading.py

@@ -7,9 +7,7 @@
 Run with: uv run python solutions/01_01_data_loading.py
 """
 
-from neo4j import GraphDatabase
-
-from config import Neo4jConfig
+from config import get_neo4j_driver
 
 # Sample text representing SEC 10-K filing content
 SAMPLE_TEXT = """
@@ -120,13 +118,7 @@ def show_graph_structure(driver) -> None:
 
 def main():
     """Run data loading demo."""
-    config = Neo4jConfig()
-    driver = GraphDatabase.driver(
-        config.uri,
-        auth=(config.username, config.password)
-    )
-
-    try:
+    with get_neo4j_driver() as driver:
         driver.verify_connectivity()
         print("Connected to Neo4j successfully!")
 
@@ -153,9 +145,7 @@ def main():
         # Show structure
         show_graph_structure(driver)
 
-    finally:
-        driver.close()
-        print("\nConnection closed.")
+    print("\nConnection closed.")
 
 
 if __name__ == "__main__":

financial_data_load/src/01_02_embeddings.py

@@ -9,13 +9,12 @@
 
 import asyncio
 
-from neo4j import GraphDatabase
 from neo4j_graphrag.indexes import create_vector_index
 from neo4j_graphrag.experimental.components.text_splitters.fixed_size_splitter import (
     FixedSizeSplitter,
 )
 
-from config import Neo4jConfig, get_embedder
+from config import get_neo4j_driver, get_embedder
 
 # Sample text representing SEC 10-K filing content
 SAMPLE_TEXT = """
@@ -159,13 +158,7 @@ def demo_search(driver, embedder) -> None:
 
 async def main():
     """Run embeddings demo."""
-    config = Neo4jConfig()
-    driver = GraphDatabase.driver(
-        config.uri,
-        auth=(config.username, config.password)
-    )
-
-    try:
+    with get_neo4j_driver() as driver:
         driver.verify_connectivity()
         print("Connected to Neo4j successfully!")
 
@@ -199,9 +192,7 @@ async def main():
         print("\n=== Vector Search Demo ===")
         demo_search(driver, embedder)
 
-    finally:
-        driver.close()
-        print("\n\nConnection closed.")
+    print("\n\nConnection closed.")
 
 
 if __name__ == "__main__":

financial_data_load/src/01_03_entity_extraction.py

@@ -9,10 +9,9 @@
 
 import asyncio
 
-from neo4j import GraphDatabase
 from neo4j_graphrag.experimental.pipeline.kg_builder import SimpleKGPipeline
 
-from config import Neo4jConfig, get_llm, get_embedder
+from config import get_neo4j_driver, get_llm, get_embedder
 
 # Sample text representing SEC 10-K filing content
 SAMPLE_TEXT = """
@@ -164,13 +163,7 @@ def find_chunks_for_entity(driver, entity_name: str) -> None:
 
 async def main():
     """Run entity extraction demo."""
-    config = Neo4jConfig()
-    driver = GraphDatabase.driver(
-        config.uri,
-        auth=(config.username, config.password)
-    )
-
-    try:
+    with get_neo4j_driver() as driver:
         driver.verify_connectivity()
         print("Connected to Neo4j successfully!")
 
@@ -217,9 +210,7 @@ async def main():
         find_chunks_for_entity(driver, "iPhone")
         find_chunks_for_entity(driver, "Apple")
 
-    finally:
-        driver.close()
-        print("\n\nConnection closed.")
+    print("\n\nConnection closed.")
 
 
 if __name__ == "__main__":

financial_data_load/src/01_04_full_dataset_queries.py

@@ -5,12 +5,11 @@
 containing SEC 10-K filings from multiple companies with extracted entities and relationships.
 
 Prerequisites:
-- Full dataset must be loaded (use restore_neo4j.py script)
+- Full dataset must be loaded (use full_data_load.py script)
 - Neo4j connection configured in config.py
 """
 
-from neo4j import GraphDatabase
-from config import Neo4jConfig, get_embedder
+from config import get_neo4j_driver, get_embedder
 
 
 INDEX_NAME = "chunkEmbeddings"
@@ -137,19 +136,14 @@ def show_company_products(driver, company_name: str):
 def main():
     # Setup connection
     print("Connecting to Neo4j...")
-    neo4j_config = Neo4jConfig()
-    driver = GraphDatabase.driver(
-        neo4j_config.uri,
-        auth=(neo4j_config.username, neo4j_config.password)
-    )
-    driver.verify_connectivity()
-    print("Connected to Neo4j successfully!\n")
-
-    # Initialize embedder
-    embedder = get_embedder()
-    print(f"Embedder initialized: {embedder.model}\n")
-
-    try:
+    with get_neo4j_driver() as driver:
+        driver.verify_connectivity()
+        print("Connected to Neo4j successfully!\n")
+
+        # Initialize embedder
+        embedder = get_embedder()
+        print(f"Embedder initialized: {embedder.model}\n")
+
         # Show graph summary
         print("\n" + "="*60)
         print("GRAPH SUMMARY")
@@ -203,11 +197,8 @@ def main():
         print()
         show_company_products(driver, "Microsoft")
 
-    finally:
-        # Cleanup
-        driver.close()
-        print("\n" + "="*60)
-        print("Connection closed.")
+    print("\n" + "="*60)
+    print("Connection closed.")
 
 
 if __name__ == "__main__":

financial_data_load/src/05_01_fulltext_search.py

@@ -10,7 +10,7 @@
 
 from neo4j import GraphDatabase
 
-from config import Neo4jConfig
+from config import get_neo4j_driver
 
 
 def basic_search(driver: GraphDatabase.driver, term: str) -> None:
@@ -146,13 +146,7 @@ def hybrid_search(driver: GraphDatabase.driver, keyword: str) -> None:
 
 def main() -> None:
     """Run all fulltext search examples."""
-    config = Neo4jConfig()
-    driver = GraphDatabase.driver(
-        config.uri,
-        auth=(config.username, config.password),
-    )
-
-    try:
+    with get_neo4j_driver() as driver:
         driver.verify_connectivity()
         print("Connected to Neo4j")
 
@@ -163,7 +157,7 @@ def main() -> None:
             )
             if not result.single():
                 print("\nError: Fulltext index 'search_entities' not found.")
-                print("Run: uv run python scripts/restore_neo4j.py --full-text")
+                print("Run: uv run python full_data_load.py")
                 return
 
         # Run examples
@@ -174,9 +168,6 @@ def main() -> None:
         search_with_graph_traversal(driver, "Nvidia")
         hybrid_search(driver, "Amazon")
 
-    finally:
-        driver.close()
-
 
 if __name__ == "__main__":
     main()

financial_data_load/src/05_02_hybrid_search.py

@@ -26,11 +26,10 @@
 """
 
 import neo4j
-from neo4j import GraphDatabase
 from neo4j_graphrag.retrievers import HybridRetriever, HybridCypherRetriever
 from neo4j_graphrag.types import RetrieverResultItem
 
-from config import Neo4jConfig, get_embedder
+from config import get_neo4j_driver, get_embedder
 
 # Index names
 VECTOR_INDEX = "chunkEmbeddings"
@@ -209,15 +208,9 @@ def search_method_comparison(retriever: HybridRetriever, query: str) -> None:
 
 def main() -> None:
     """Run all hybrid search examples."""
-    config = Neo4jConfig()
-    driver = GraphDatabase.driver(
-        config.uri,
-        auth=(config.username, config.password),
-    )
-
-    try:
+    with get_neo4j_driver() as driver:
         driver.verify_connectivity()
-        print(f"Connected to Neo4j: {config.uri}")
+        print("Connected to Neo4j")
 
         # Initialize embedder
         embedder = get_embedder()
@@ -232,7 +225,7 @@ def main() -> None:
             )
             if not result.single():
                 print(f"\nError: Fulltext index '{FULLTEXT_INDEX}' not found.")
-                print("Run: uv run python scripts/restore_neo4j.py --full-text")
+                print("Run: uv run python full_data_load.py")
                 return
 
         # Create HybridRetriever
@@ -263,9 +256,7 @@ def main() -> None:
         graph_enhanced_search(hybrid_cypher_retriever, "artificial intelligence")
         search_method_comparison(hybrid_retriever, "Microsoft cloud computing strategy")
 
-    finally:
-        driver.close()
-        print("\nConnection closed")
+    print("\nConnection closed")
 
 
 if __name__ == "__main__":

pyproject.toml

@@ -11,7 +11,8 @@ dependencies = [
     "python-dotenv",
     "pydantic-settings>=2.0.0",
     "neo4j>=5.0.0",
-    "neo4j-graphrag>=1.10.0",
+    # Using local fork for testing - replace with "neo4j-graphrag>=1.10.0" for production
+    "neo4j-graphrag @ file:///Users/ryanknight/projects/neo4j-graphrag-python",
     "httpx>=0.27.0",
     "openai>=1.0.0",
     "ipykernel>=6.0.0",