replace Keycloak authentication with token-based inference API authentication

Author: gopal-raj-suresh

CodeTranslation/README.md

@@ -1,7 +1,7 @@
 ## Code Translation
 
 A full-stack code translation application that converts code between programming languages using AI.
-The system integrates a FastAPI backend powered by CodeLlama-34b-instruct, alongside a modern React + Vite + Tailwind CSS frontend for an intuitive translation experience.
+The system integrates a FastAPI backend, alongside a modern React + Vite + Tailwind CSS frontend for an intuitive translation experience.
 
 ## Table of Contents
 
@@ -17,7 +17,7 @@ The system integrates a FastAPI backend powered by CodeLlama-34b-instruct, along
 
 ## Project Overview
 
-The **Code Translation** application demonstrates how large language models can be used to translate code between different programming languages. It accepts source code in one language, processes it through CodeLlama-34b-instruct, and returns translated code in the target language. This project integrates seamlessly with cloud-hosted APIs or local model endpoints, offering flexibility for research, enterprise, or educational use.
+The **Code Translation** application demonstrates how large language models can be used to translate code between different programming languages. It accepts source code in one language, processes it and returns translated code in the target language. This project integrates seamlessly with cloud-hosted APIs or local model endpoints, offering flexibility for research, enterprise, or educational use.
 
 ---
 
@@ -27,9 +27,8 @@ The **Code Translation** application demonstrates how large language models can
 
 - Code translation between 6 languages (Java, C, C++, Python, Rust, Go)
 - PDF code extraction with pattern recognition
-- CodeLlama-34b-instruct for accurate translations
 - Enterprise inference endpoints
-- Keycloak authentication for secure API access
+- Token-based authentication for inference API
 - Comprehensive error handling and logging
 - File validation and size limits
 - CORS enabled for web integration
@@ -70,8 +69,7 @@ Below is the architecture as it consists of a server that waits for code input o
       end
 
       subgraph "External Services"
-          E[Keycloak Auth]
-          F[CodeLlama-34b Model]
+          E[CodeLlama-34b Model]
       end
 
       A1 --> B
@@ -80,31 +78,29 @@ Below is the architecture as it consists of a server that waits for code input o
       B --> C
       C -->|Extracted Code| B
       B --> D
-      D -->|Get Token| E
-      E -->|Access Token| D
-      D -->|Translate Code + Token| F
-      F -->|Translated Code| D
+      D -->|Translate Code + Token| E
+      E -->|Translated Code| D
       D --> B
       B --> A
 
       style A fill:#e1f5ff
       style B fill:#fff4e1
-      style F fill:#e1ffe1
+      style E fill:#e1ffe1
 ```
 
-This application is built with enterprise inference capabilities using Keycloak for authentication and CodeLlama-34b-instruct for code translation.
+This application is built with enterprise inference capabilities using token-based authentication and CodeLlama-34b-instruct for code translation.
 
 **Service Components:**
 
 1. **React Web UI (Port 3000)** -  Provides side-by-side code comparison interface with language selection, PDF upload, and real-time translation results
 
-2. **FastAPI Backend (Port 5001)** -  Handles code validation, PDF extraction, Keycloak authentication, and orchestrates code translation through CodeLlama model
+2. **FastAPI Backend (Port 5001)** -  Handles code validation, PDF extraction, inference API authentication, and orchestrates code translation through CodeLlama model
 
 **Typical Flow:**
 
 1. User enters code or uploads a PDF through the web UI.
 2. The backend validates the input and extracts code if needed.
-3. The backend authenticates with Keycloak and calls CodeLlama model.
+3. The backend authenticates with inference API and calls CodeLlama model.
 4. The model translates the code to the target language.
 5. The translated code is returned and displayed to the user.
 6. User can copy the translated code with one click.
@@ -118,7 +114,13 @@ This application is built with enterprise inference capabilities using Keycloak
 Before you begin, ensure you have the following installed:
 
 - **Docker and Docker Compose**
-- **Enterprise inference endpoint access** (Keycloak authentication)
+- **Enterprise inference endpoint access** (token-based authentication)
+
+### Required API Configuration
+
+**For Inference Service (Code Translation):**
+- INFERENCE_API_ENDPOINT: URL of the deployed model inference service
+- INFERENCE_API_TOKEN: API key / bearer token used to authenticate requests
 
 ### Verify Docker Installation
 
@@ -139,7 +141,7 @@ docker ps
 ### Clone the Repository
 
 ```bash
-git clone https://github.com/opea-project/GenAIExamples.git
+git clone https://github.com/cld2labs/GenAIExamples.git
 cd GenAIExamples/CodeTranslation
 ```
 
@@ -153,13 +155,9 @@ cat > .env << EOF
 # Backend API Configuration
 BACKEND_PORT=5001
 
-# Required - Enterprise/Keycloak Configuration
-BASE_URL=https://api.example.com
-KEYCLOAK_CLIENT_ID=api
-KEYCLOAK_CLIENT_SECRET=your_client_secret
-
-# Required - Model Configuration
-INFERENCE_MODEL_ENDPOINT=CodeLlama-34b-Instruct-hf
+# Inference API Configuration
+INFERENCE_API_ENDPOINT=https://your-api-endpoint.com/deployment
+INFERENCE_API_TOKEN=your-pre-generated-token-here
 INFERENCE_MODEL_NAME=codellama/CodeLlama-34b-Instruct-hf
 
 # LLM Settings
@@ -181,13 +179,9 @@ Or manually create `.env` with:
 # Backend API Configuration
 BACKEND_PORT=5001
 
-# Required - Enterprise/Keycloak Configuration
-BASE_URL=https://api.example.com
-KEYCLOAK_CLIENT_ID=api
-KEYCLOAK_CLIENT_SECRET=your_client_secret
-
-# Required - Model Configuration
-INFERENCE_MODEL_ENDPOINT=CodeLlama-34b-Instruct-hf
+# Inference API Configuration
+INFERENCE_API_ENDPOINT=https://your-api-endpoint.com/deployment
+INFERENCE_API_TOKEN=your-pre-generated-token-here
 INFERENCE_MODEL_NAME=codellama/CodeLlama-34b-Instruct-hf
 
 # LLM Settings
@@ -209,7 +203,7 @@ CORS_ALLOW_ORIGINS=["http://localhost:5173", "http://localhost:3000"]
 Start both API and UI services together with Docker Compose:
 
 ```bash
-# From the CodeTranslation directory
+# From the code-translation directory
 docker compose up --build
 
 # Or run in detached mode (background)

CodeTranslation/TROUBLESHOOTING.md

@@ -9,16 +9,16 @@ This document contains all common issues encountered during development and thei
 
 ### API Common Issues
 
-#### "API client not initialized. Check Keycloak configuration."
+#### "API client not initialized. Check inference API configuration."
 
 **Solution**:
 
 1. Create a `.env` file in the root directory
-2. Add your Keycloak credentials:
+2. Add your inference API credentials:
    ```
-   BASE_URL=https://api.example.com
-   KEYCLOAK_CLIENT_ID=api
-   KEYCLOAK_CLIENT_SECRET=your_client_secret
+   INFERENCE_API_ENDPOINT=https://your-api-endpoint.com/deployment
+   INFERENCE_API_TOKEN=your-pre-generated-token-here
+   INFERENCE_MODEL_NAME=codellama/CodeLlama-34b-Instruct-hf
    ```
 3. Restart the server
 
@@ -68,10 +68,11 @@ This document contains all common issues encountered during development and thei
 
 **Solution**:
 
-1. Verify Keycloak authentication is working (check `/health` endpoint)
+1. Verify inference API authentication is working (check `/health` endpoint)
 2. Check if the model endpoint is accessible
-3. Try with simpler code first
-4. Check server logs for API errors
+3. Verify INFERENCE_API_TOKEN is valid and not expired
+4. Try with simpler code first
+5. Check server logs for API errors
 
 #### "No module named 'pypdf'"
 

CodeTranslation/api/config.py

@@ -8,20 +8,11 @@
 # Load environment variables from .env file
 load_dotenv()
 
-# Custom API Configuration for Keycloak
-BASE_URL = os.getenv("BASE_URL", "https://api.example.com")
-KEYCLOAK_REALM = os.getenv("KEYCLOAK_REALM", "master")
-KEYCLOAK_CLIENT_ID = os.getenv("KEYCLOAK_CLIENT_ID", "api")
-KEYCLOAK_CLIENT_SECRET = os.getenv("KEYCLOAK_CLIENT_SECRET")
-
-# Model Configuration for CodeLlama-34b-instruct
-INFERENCE_MODEL_ENDPOINT = os.getenv("INFERENCE_MODEL_ENDPOINT", "CodeLlama-34b-Instruct")
+# Inference API Configuration
+INFERENCE_API_ENDPOINT = os.getenv("INFERENCE_API_ENDPOINT")
+INFERENCE_API_TOKEN = os.getenv("INFERENCE_API_TOKEN")
 INFERENCE_MODEL_NAME = os.getenv("INFERENCE_MODEL_NAME", "codellama/CodeLlama-34b-Instruct-hf")
 
-# Validate required configuration
-if not KEYCLOAK_CLIENT_SECRET:
-    raise ValueError("KEYCLOAK_CLIENT_SECRET must be set in environment variables")
-
 # Application Settings
 APP_TITLE = "Code Translation API"
 APP_DESCRIPTION = "AI-powered code translation service using CodeLlama-34b-instruct"

CodeTranslation/api/models.py

@@ -60,7 +60,7 @@ class HealthResponse(BaseModel):
     """Response model for health check"""
     status: str = Field(..., description="Health status")
     model_configured: bool = Field(..., description="Whether model is configured")
-    keycloak_authenticated: bool = Field(..., description="Whether Keycloak auth is successful")
+    inference_authenticated: bool = Field(..., description="Whether inference API auth is successful")
 
 
 class SupportedLanguagesResponse(BaseModel):

CodeTranslation/api/server.py

@@ -33,7 +33,7 @@ async def lifespan(app: FastAPI):
     try:
         api_client = get_api_client()
         app.state.api_client = api_client
-        logger.info("✓ API client initialized with Keycloak authentication")
+        logger.info("API client initialized with inference endpoint")
     except Exception as e:
         logger.error(f"Failed to initialize API client: {str(e)}")
         app.state.api_client = None
@@ -81,7 +81,7 @@ def health_check():
     return HealthResponse(
         status="healthy",
         model_configured=bool(config.INFERENCE_MODEL_NAME),
-        keycloak_authenticated=app.state.api_client is not None and app.state.api_client.is_authenticated()
+        inference_authenticated=app.state.api_client is not None and app.state.api_client.is_authenticated()
     )
 
 
@@ -105,7 +105,7 @@ def translate_code_endpoint(request: TranslateRequest):
     if not app.state.api_client:
         raise HTTPException(
             status_code=status.HTTP_503_SERVICE_UNAVAILABLE,
-            detail="API client not initialized. Check Keycloak configuration."
+            detail="API client not initialized. Check inference API configuration."
         )
 
     # Validate languages
@@ -144,7 +144,7 @@ def translate_code_endpoint(request: TranslateRequest):
                 detail="Translation failed. No output received from model."
             )
 
-        logger.info(f"✓ Successfully translated code")
+        logger.info(f"Successfully translated code")
 
         return TranslateResponse(
             translated_code=translated_code,
@@ -196,7 +196,7 @@ async def upload_pdf(file: UploadFile = File(...)):
                 detail="No code content could be extracted from the PDF"
             )
 
-        logger.info(f"✓ Successfully extracted code from PDF: {file.filename}")
+        logger.info(f"Successfully extracted code from PDF: {file.filename}")
 
         return UploadPdfResponse(
             message=f"Successfully extracted code from '{file.filename}'",

CodeTranslation/api/services/api_client.py

@@ -1,9 +1,8 @@
 """
-API Client for Keycloak authentication and API calls
+API Client for inference API calls
 """
 
 import logging
-import requests
 import httpx
 from typing import Optional
 import config
@@ -13,54 +12,23 @@
 
 class APIClient:
     """
-    Client for handling Keycloak authentication and API calls
+    Client for handling inference API calls
     """
 
     def __init__(self):
-        self.base_url = config.BASE_URL
-        self.token = None
-        self.http_client = None
-        self._authenticate()
-
-    def _authenticate(self) -> None:
-        """
-        Authenticate and obtain access token from Keycloak
-        """
-        token_url = f"{self.base_url}/token"
-        payload = {
-            "grant_type": "client_credentials",
-            "client_id": config.KEYCLOAK_CLIENT_ID,
-            "client_secret": config.KEYCLOAK_CLIENT_SECRET,
-        }
-
-        try:
-            response = requests.post(token_url, data=payload, verify=False)
-
-            if response.status_code == 200:
-                self.token = response.json().get("access_token")
-                logger.info(f"✓ Access token obtained: {self.token[:20]}..." if self.token else "Failed to get token")
-
-                # Create httpx client with SSL verification disabled
-                self.http_client = httpx.Client(verify=False)
-
-            else:
-                logger.error(f"Error obtaining token: {response.status_code} - {response.text}")
-                raise Exception(f"Authentication failed: {response.status_code}")
-
-        except Exception as e:
-            logger.error(f"Error during authentication: {str(e)}")
-            raise
+        self.endpoint = config.INFERENCE_API_ENDPOINT
+        self.token = config.INFERENCE_API_TOKEN
+        self.http_client = httpx.Client(verify=False) if self.token else None
 
     def get_inference_client(self):
         """
         Get OpenAI-style client for code generation inference
-        Uses CodeLlama-34b-instruct endpoint
         """
         from openai import OpenAI
 
         return OpenAI(
             api_key=self.token,
-            base_url=f"{self.base_url}/{config.INFERENCE_MODEL_ENDPOINT}/v1",
+            base_url=f"{self.endpoint}/v1",
             http_client=self.http_client
         )
 

CodeTranslation/docker-compose.yaml

@@ -1,5 +1,3 @@
-version: '3.8'
-
 services:
   backend:
     build:
@@ -11,10 +9,8 @@ services:
     env_file:
       - .env
     environment:
-      - BASE_URL=${BASE_URL}
-      - KEYCLOAK_CLIENT_ID=${KEYCLOAK_CLIENT_ID}
-      - KEYCLOAK_CLIENT_SECRET=${KEYCLOAK_CLIENT_SECRET}
-      - INFERENCE_MODEL_ENDPOINT=${INFERENCE_MODEL_ENDPOINT}
+      - INFERENCE_API_ENDPOINT=${INFERENCE_API_ENDPOINT}
+      - INFERENCE_API_TOKEN=${INFERENCE_API_TOKEN}
       - INFERENCE_MODEL_NAME=${INFERENCE_MODEL_NAME}
       - LLM_TEMPERATURE=${LLM_TEMPERATURE:-0.2}
       - LLM_MAX_TOKENS=${LLM_MAX_TOKENS:-4096}