How dipanshu1146/waterrag works
This is a wastewater-document RAG application: users upload documents, which the backend extracts, chunks, and indexes for later question answering; the UI also supports document listing and deletion. The graph distinguishes upload orchestration, ingestion/extraction, retrieval, and storage. The README is empty, so capabilities and relationships are based on the supplied source excerpts and SOURCE INDEX; unsampled route behavior and integrations are not expanded beyond documented references.
Main entry points: search_index.py Index operations · upload.py Upload management · ask.py Question answering · pipeline.py Ingestion pipeline
Components 18
User interface
- Web application
frontend/src/App.jsx
API workflows
- FastAPI application
src/api/main.py - Upload management
src/api/routes/upload.py - Question answering
src/api/routes/ask.py
Ingestion and extraction
- Ingestion pipeline
src/ingestion/pipeline.py - Document loading
src/ingestion/load_documents.py - PDF table extraction
src/extraction/pdf_table_extractor.py - Table row formatting
src/extraction/table_to_text.py - Document extraction
src/extraction/document_extractor.py
Question retrieval
- Hybrid retrieval
src/vectorstore/general_search.py - Filtered retrieval
src/vectorstore/filter_index.py
Index and storage
- Index operations
src/vectorstore/search_index.py - Cloud storage adapter
src/vectorstore/cloud_store.py - Runtime configuration
src/config.py - Document vector index
Other components
- Document user
- Cloud database
- Cloud file storage
Connections 25
- Document user → Web application: uses
- Web application → Upload management: uploads files
frontend/src/App.jsx - Web application → Upload management: manages documents
frontend/src/App.jsx - FastAPI application → Upload management: registers route
src/api/main.py - FastAPI application → Question answering: registers route
src/api/main.py - Upload management → Ingestion pipeline: indexes upload
src/api/routes/upload.py - Upload management → Index operations: indexes and deletes
src/api/routes/upload.py - Ingestion pipeline → Document loading: loads documents
src/ingestion/pipeline.py - Ingestion pipeline → PDF table extraction: extracts PDF tables
src/ingestion/pipeline.py - Ingestion pipeline → Table row formatting: formats table rows
src/ingestion/pipeline.py - Document loading → PDF table extraction: extracts tables
src/ingestion/load_documents.py - Question answering → Hybrid retrieval: retrieves broadly
src/api/routes/ask.py - Question answering → Filtered retrieval: filters retrieval
src/api/routes/ask.py - Question answering → Index operations: queries index
src/api/routes/ask.py - Hybrid retrieval → Index operations: semantic search
src/vectorstore/general_search.py - Filtered retrieval → Index operations: queries index
src/vectorstore/filter_index.py - Index operations → Document loading: loads Excel records
src/vectorstore/search_index.py - Index operations → Cloud storage adapter: uses cloud backend
src/vectorstore/search_index.py - Hybrid retrieval → Cloud storage adapter: uses cloud backend
src/vectorstore/general_search.py - Hybrid retrieval → Runtime configuration: reads settings
src/vectorstore/general_search.py - Filtered retrieval → Runtime configuration: reads settings
src/vectorstore/filter_index.py - Index operations → Runtime configuration: reads settings
src/vectorstore/search_index.py - Index operations → Document vector index: reads and writes
src/vectorstore/search_index.py - Cloud storage adapter → Cloud database: stores vectors
src/vectorstore/cloud_store.py - Cloud storage adapter → Cloud file storage: stores files
src/vectorstore/cloud_store.py