Doc fixes

Author: Sagar

.gitignore

@@ -104,5 +104,4 @@ confidential_uploads/
 private_cases/
 temp_analysis/
 
-TECHNICAL_DOCUMENTATION.md
 jurisfind-backend_key.pem

README.md

@@ -39,8 +39,7 @@ def create_app():
             "http://localhost:5173",  # Vite dev server
             "http://localhost:3000",  # React dev server
             "http://20.186.113.106",  # Azure VM public IP
-            # Add your Azure Static Web App URL here after deploying:
-            # "https://your-app.azurestaticapps.net",
+            "https://blue-cliff-0dfeb910f.2.azurestaticapps.net",  # Azure Static Web App
         ],
         allow_credentials=True,
         allow_methods=["*"],

api/main.py

@@ -0,0 +1,45 @@
+"""
+Run this once to upload FAISS index files to Azure Blob Storage.
+Usage: python upload_to_blob.py
+"""
+import os
+from azure.storage.blob import BlobServiceClient
+
+# ── Paste your connection string here ──────────────────────────────────────────
+CONNECTION_STRING = "YOUR_CONNECTION_STRING_HERE"
+CONTAINER_NAME = "data"
+# ───────────────────────────────────────────────────────────────────────────────
+
+FILES = {
+    "faiss_store/legal_cases.index": r"e:\Sagar\AI_ML_DL_DS\projects\legal-case\JurisFind\api\data\faiss_store\legal_cases.index",
+    "faiss_store/id2name.json":      r"e:\Sagar\AI_ML_DL_DS\projects\legal-case\JurisFind\api\data\faiss_store\id2name.json",
+}
+
+def upload():
+    client = BlobServiceClient.from_connection_string(CONNECTION_STRING)
+    container = client.get_container_client(CONTAINER_NAME)
+
+    # Create container if it doesn't exist
+    try:
+        container.create_container()
+        print(f"Created container: {CONTAINER_NAME}")
+    except Exception:
+        print(f"Container '{CONTAINER_NAME}' already exists.")
+
+    for blob_name, local_path in FILES.items():
+        size_mb = os.path.getsize(local_path) / (1024 * 1024)
+        print(f"Uploading {blob_name} ({size_mb:.1f} MB)...")
+        with open(local_path, "rb") as f:
+            container.upload_blob(blob_name, f, overwrite=True)
+        print(f"  ✓ Done")
+
+    print("\nAll files uploaded successfully!")
+    print(f"Blobs in container '{CONTAINER_NAME}':")
+    for blob in container.list_blobs():
+        print(f"  - {blob.name}")
+
+if __name__ == "__main__":
+    if CONNECTION_STRING == "YOUR_CONNECTION_STRING_HERE":
+        print("ERROR: Set your CONNECTION_STRING in this script first.")
+    else:
+        upload()

api/upload_to_blob.py

@@ -2,55 +2,108 @@
 
 ## Overview
 
-Legal Case provides a REST API built with FastAPI. Interactive documentation available at `/docs` when running locally.
+JurisFind provides a REST API built with FastAPI. Interactive documentation available at `/docs` when running locally or in production.
 
 ## Base URL
+
+Local development:
 ```
 http://localhost:8000/api
 ```
 
+Production:
+```
+http://20.186.113.106/api
+```
+
+Interactive Swagger UI (production): `http://20.186.113.106/docs`
+Interactive Swagger UI (local): `http://localhost:8000/docs`
+
 ## Endpoints
 
 ### Health Check
 - **GET** `/health`
-- **Response**: `{"status": "healthy"}`
+- **Response**:
+  ```json
+  {"status": "healthy", "message": "Legal case search service is running", "total_cases": 46456}
+  ```
 
 ### Search
 - **POST** `/search`
 - **Body**:
+  ```json
+  {"query": "breach of contract", "top_k": 5}
+  ```
+- **Response**:
   ```json
   {
-    "query": "contract terms",
-    "top_k": 5
+    "success": true,
+    "query": "breach of contract",
+    "results": [
+      {"filename": "case_001.pdf", "score": 0.94, "similarity_percentage": 94.3}
+    ],
+    "total_results": 5
   }
   ```
-- **Response**: Array of search results with scores and metadata
 
-### Document Analysis
-- **POST** `/analyze`
+### Unified Document Analysis
+- **POST** `/unified/analyze`
 - **Body**:
   ```json
-  {
-    "filename": "document.pdf",
-    "query": "summarize key points"
-  }
+  {"filename": "case_001.pdf", "source": "database"}
   ```
-- **Response**: AI-generated analysis
+  `source` is `"database"` for PDFs from the main index or `"uploaded"` for a confidential upload.
+- **Response**: `{success, filename, text_length, embedding_status, summary, message}`
 
-### Chat
-- **POST** `/chat`
+### Unified Document Q&A
+- **POST** `/unified/ask`
 - **Body**:
   ```json
-  {
-    "message": "What are the main arguments?",
-    "context": "document.pdf"
-  }
+  {"filename": "case_001.pdf", "question": "What was the ruling?", "source": "database"}
   ```
-- **Response**: Conversational response
+- **Response**: `{success, filename, question, answer}`
+
+### Serve PDF
+- **GET** `/pdf/{filename}`
+- **Response**: PDF binary stream (`application/pdf`). Downloads from Azure Blob Storage, or falls back to local `api/data/pdfs/`.
 
 ### Document Statistics
 - **GET** `/document-stats/{filename}`
-- **Response**: Document metadata and statistics
+- **Response**: `{success, filename, stats}` with embedding metadata for the analyzed document.
+
+### Upload Confidential PDF
+- **POST** `/upload-confidential-pdf`
+- **Body**: `multipart/form-data` with a `file` field containing the PDF
+- **Response**: `{success, filename, message}`
+- The file is saved to the ephemeral `confidential_tmp` Docker volume and never reaches Blob Storage.
+
+### Retrieve Similar Cases
+- **POST** `/retrieve-similar-cases?filename={name}&top_k=5`
+- Finds cases in the main FAISS index semantically similar to the uploaded confidential PDF.
+- **Response**: `{success, filename, similar_cases, total_found}`
+
+### Cleanup Confidential Session
+- **DELETE** `/cleanup-confidential/{filename}`
+- **Response**: `{success, message}` — deletes the uploaded PDF and its embeddings.
+
+### Legal Chatbot
+- **POST** `/legal-chat`
+- **Body**:
+  ```json
+  {"question": "What constitutes fair use under copyright law?"}
+  ```
+- **Response**: `{success, response, is_legal, domain_filtered}`
+  - `is_legal`: `false` if the question was detected as non-legal and filtered.
+
+### Additional Utility Endpoints
+- **GET** `/list-pdfs` — list all available PDFs in Blob or local storage
+- **POST** `/analyze-document?filename={name}` — analyze a single database PDF (legacy route)
+- **POST** `/ask-question` `{filename, question}` — Q&A on a database PDF (legacy route)
+- **DELETE** `/cleanup-embeddings/{filename}` — remove cached embeddings for a document
+- **DELETE** `/legal-chat/clear` — reset chatbot conversation history
+- **GET** `/legal-chat/stats` — chatbot session statistics
+- **POST** `/upload-pdf-to-azure` `multipart/form-data` — upload a PDF to Blob Storage directly
+- **POST** `/generate-embeddings-from-azure` — rebuild FAISS index from Blob PDFs (admin operation)
 
 ## Authentication
 
@@ -73,8 +126,7 @@ Error responses include:
 
 ## Rate Limiting
 
-- Default: 100 requests per minute per IP
-- Configurable in environment variables
+Nginx enforces 30 requests per minute per IP by default. Requests exceeding this limit receive HTTP 429. The limit is configured in `nginx.conf` and can be adjusted on the VM.
 
 ## CORS
 

docs/api_reference.md

@@ -2,7 +2,7 @@
 
 ## System Overview
 
-Legal Case uses a microservices architecture with three main layers:
+JurisFind uses a microservices architecture with three main layers:
 
 ```
 ┌─────────────────┐    ┌─────────────────┐    ┌─────────────────┐