Refactor GraphRAG into standalone tool

Dockerfile

@@ -26,6 +26,7 @@ RUN python -m pip install --no-cache-dir \
 
 COPY app ./app
 COPY scripts ./scripts
+COPY tools ./tools
 
 RUN mkdir -p data
 

tools/graphrag/build_graphrag.py

@@ -5,11 +5,11 @@ import json
 from pathlib import Path
 from typing import Any
 
-from scripts.graph_db import (
+from .graph_db import (
     create_neo4j_driver,
     get_neo4j_settings,
 )
-from scripts.graphrag import (
+from .graphrag import (
     GRAPH_SCOPE,
     build_graph_payload,
 )
@@ -17,7 +17,7 @@ from scripts.graphrag import (
 
 PROJECT_ROOT = Path(
     __file__
-).resolve().parents[1]
+).resolve().parents[2]
 
 DOCUMENTS_FILE = (
     PROJECT_ROOT

tools/graphrag/graph_search.py

@@ -1283,7 +1283,7 @@ def search_graph(
     limit: int = DEFAULT_LIMIT,
     max_entities: int = DEFAULT_MAX_ENTITIES,
 ) -> dict[str, Any]:
-    from scripts.graph_db import (
+    from .graph_db import (
         create_neo4j_driver,
         get_neo4j_settings,
     )

tools/graphrag/neo4j_smoke.py

@@ -1,4 +1,4 @@
-from scripts.graph_db import verify_neo4j_connection
+from .graph_db import verify_neo4j_connection
 
 
 def main() -> None:

tools/graphrag/tests/test_graph_search.py

@@ -4,7 +4,7 @@ from typing import Any
 
 import pytest
 
-import scripts.graph_search as graph_search
+import tools.graphrag.graph_search as graph_search
 
 
 class FakeSession:

tools/graphrag/tests/test_graphrag.py

@@ -1,6 +1,6 @@
 from __future__ import annotations
 
-from scripts.graphrag import (
+from tools.graphrag.graphrag import (
     build_graph_payload,
     extract_explicit_proposal_title,
     extract_works,