src/utils/response_formatter.py

from typing import Dict, List
import re

class ResponseFormatter:
    @staticmethod
    def clean_text(text: str) -> str:
        """Clean up raw text by removing excessive whitespace and common boilerplate"""
        # Remove multiple newlines and spaces
        text = re.sub(r'\n+', '\n', text)
        text = re.sub(r'\s+', ' ', text)
        
        # Remove common website elements
        text = re.sub(r'SUBSCRIBE|RECENT|POPULAR|TRENDY', '', text, flags=re.I)
        text = re.sub(r'Copyright © \d{4}.*', '', text)
        text = re.sub(r'Privacy Policy|Terms of Service', '', text)
        
        return text.strip()

    @staticmethod
    def format_sources(sources: List[Dict]) -> str:
        """Format source URLs into readable references"""
        if not sources:
            return ""
        
        formatted_sources = "\n\nSources:\n"
        for i, source in enumerate(sources, 1):
            formatted_sources += f"{i}. {source['url']}\n"
        return formatted_sources

    @staticmethod
    def format_response(api_response: Dict) -> str:
        """Convert API response to natural language"""
        if "error" in api_response:
            return f"Sorry, I encountered an error: {api_response['error']}"
            
        if "response" not in api_response:
            return "I couldn't find any relevant information."
            
        # Clean and format the main response
        clean_response = ResponseFormatter.clean_text(api_response["response"])
        
        # Add sources if available
        if "sources" in api_response:
            clean_response += ResponseFormatter.format_sources(api_response["sources"])
            
        return clean_response