feat: Add comprehensive Azure OpenAI support

This commit adds full Azure OpenAI integration to the proxy server, allowing
users to route requests to Azure OpenAI deployments using the azure/ model prefix.

## Key Features Added:

### Server Integration (server.py):
- Added Azure environment variables (AZURE_OPENAI_ENDPOINT, AZURE_OPENAI_API_KEY, AZURE_API_VERSION, AZURE_DEPLOYMENT_NAME)
- Enhanced API key selection logic to detect azure/ model prefix
- Proper Azure credential and endpoint configuration for LiteLLM
- Updated model validation to include Azure models
- Azure models use OpenAI-compatible message processing
- Fixed undefined endpoint variable issue in log_request_beautifully function
- Added Azure model mapping for both small and big models
- Improved error detail serialization for exceptions

### Environment Configuration (.env.example):
- Added Azure OpenAI configuration section with examples
- Clear documentation on azure/ model format usage
- No default values for version/deployment (explicit configuration required)

### Test Suite:
- test_azure_openai.py: Direct Azure OpenAI connectivity test using LiteLLM
- test_azure_proxy.py: Proxy server Azure routing test

## Usage:
Set Azure configuration in .env file, then use model format: azure/your-deployment-name

## Testing:
- ✅ Direct Azure connection via LiteLLM
- ✅ Proxy routing to Azure OpenAI
- ✅ Token usage tracking
- ✅ Error handling and validation
- ✅ Response parsing and formatting

## Technical Details:
- Uses same LiteLLM library for consistency
- Explicit configuration (no defaults for version/deployment)
- Proper error messages and validation
- Clean separation from other providers
- Compatible with existing OpenAI message processing

🤖 Generated with [Claude Code](https://claude.ai/code)

Co-Authored-By: Claude <[email protected]>

Author: 2 people

.env.example

@@ -18,4 +18,12 @@ PREFERRED_PROVIDER="openai"
 # Example Google mapping:
 # PREFERRED_PROVIDER="google"
 # BIG_MODEL="gemini-2.5-pro-preview-03-25"
-# SMALL_MODEL="gemini-2.0-flash" 
+# SMALL_MODEL="gemini-2.0-flash" 
+
+# Azure OpenAI Configuration (optional)
+# Uncomment and set these if you want to use Azure OpenAI
+# Use model format: azure/your-deployment-name in requests
+# AZURE_OPENAI_ENDPOINT="https://your-resource.openai.azure.com"
+# AZURE_OPENAI_API_KEY="your-azure-openai-api-key"
+# AZURE_API_VERSION="2025-01-01-preview"
+# AZURE_DEPLOYMENT_NAME="gpt-4"