"""Configuration constants for the application."""
from __future__ import annotations
import os
from typing import Dict, Any

# Agent Configuration
class AgentConfig:
    """Configuration for agent behavior."""
    # Optimization cycles
    OPTIMIZATION_CYCLES = int(os.getenv("OPTIMIZATION_CYCLES", "3"))
    
    # Character limits
    RESUME_MAX_CHARS = int(os.getenv("RESUME_MAX_CHARS", "8000"))
    COVER_LETTER_MAX_CHARS = int(os.getenv("COVER_LETTER_MAX_CHARS", "4000"))
    
    # Keyword extraction
    JOB_KEYWORDS_COUNT = int(os.getenv("JOB_KEYWORDS_COUNT", "40"))
    RESUME_KEYWORDS_COUNT = int(os.getenv("RESUME_KEYWORDS_COUNT", "25"))
    COVER_KEYWORDS_COUNT = int(os.getenv("COVER_KEYWORDS_COUNT", "20"))
    MAX_NEW_KEYWORDS = int(os.getenv("MAX_NEW_KEYWORDS", "30"))
    
    # Consistency checking
    MAX_CONTRADICTION_FIXES = int(os.getenv("MAX_CONTRADICTION_FIXES", "8"))
    
    # Text processing
    SKILL_DISPLAY_LIMIT = int(os.getenv("SKILL_DISPLAY_LIMIT", "8"))
    DISTILL_MAX_POINTS = int(os.getenv("DISTILL_MAX_POINTS", "12"))


# LLM Configuration
class LLMConfig:
    """Configuration for LLM providers."""
    PROVIDER = os.getenv("LLM_PROVIDER", "openai").lower()
    MODEL = os.getenv("LLM_MODEL")
    
    # Model defaults by provider
    DEFAULT_MODELS = {
        "openai": "gpt-4o-mini",
        "anthropic": "claude-3-5-sonnet-latest",
        "gemini": "gemini-1.5-flash"
    }
    
    # Token limits
    RESUME_MAX_TOKENS = int(os.getenv("RESUME_MAX_TOKENS", "1200"))
    COVER_MAX_TOKENS = int(os.getenv("COVER_MAX_TOKENS", "800"))
    DEFAULT_MAX_TOKENS = int(os.getenv("DEFAULT_MAX_TOKENS", "1200"))
    
    # Temperature
    TEMPERATURE = float(os.getenv("LLM_TEMPERATURE", "0.4"))


# API Configuration
class APIConfig:
    """Configuration for external APIs."""
    # LinkedIn OAuth
    LINKEDIN_CLIENT_ID = os.getenv("LINKEDIN_CLIENT_ID", "")
    LINKEDIN_CLIENT_SECRET = os.getenv("LINKEDIN_CLIENT_SECRET", "")
    LINKEDIN_REDIRECT_URI = os.getenv("LINKEDIN_REDIRECT_URI", "http://localhost:8501")
    MOCK_MODE = os.getenv("MOCK_MODE", "true").lower() == "true"
    
    # Tavily Research
    TAVILY_API_KEY = os.getenv("TAVILY_API_KEY")
    TAVILY_MAX_RESULTS = int(os.getenv("TAVILY_MAX_RESULTS", "5"))
    
    # Timeouts
    HTTP_TIMEOUT = float(os.getenv("HTTP_TIMEOUT", "20.0"))
    
    # Retry configuration
    MAX_RETRIES = int(os.getenv("MAX_RETRIES", "3"))
    RETRY_BACKOFF = float(os.getenv("RETRY_BACKOFF", "1.0"))


# Memory Configuration
class MemoryConfig:
    """Configuration for memory storage."""
    BASE_DIR = os.getenv("MEMORY_BASE_DIR", "/workspace/memory/data")
    
    # File limits
    MAX_PATH_LENGTH = int(os.getenv("MAX_PATH_LENGTH", "255"))
    
    # Cleanup
    AUTO_CLEANUP_DAYS = int(os.getenv("AUTO_CLEANUP_DAYS", "30"))


# Security Configuration
class SecurityConfig:
    """Security-related configuration."""
    # Input validation
    MAX_INPUT_LENGTH = int(os.getenv("MAX_INPUT_LENGTH", "10000"))
    MAX_JOB_ID_LENGTH = int(os.getenv("MAX_JOB_ID_LENGTH", "100"))
    
    # Rate limiting (requests per minute)
    RATE_LIMIT_PER_USER = int(os.getenv("RATE_LIMIT_PER_USER", "10"))
    
    # Session
    SESSION_TIMEOUT_MINUTES = int(os.getenv("SESSION_TIMEOUT_MINUTES", "60"))


# UI Configuration
class UIConfig:
    """User interface configuration."""
    # Streamlit
    PAGE_TITLE = "Job Application Assistant"
    LAYOUT = "wide"
    
    # Display limits
    MAX_PREVIEW_LENGTH = int(os.getenv("MAX_PREVIEW_LENGTH", "3000"))
    MAX_SUGGESTED_JOBS = int(os.getenv("MAX_SUGGESTED_JOBS", "5"))
    
    # Gradio
    GRADIO_PORT = int(os.getenv("PORT", "7860"))
    GRADIO_SERVER = "0.0.0.0"


# Probability Scoring Weights
class ScoringConfig:
    """Configuration for probability scoring."""
    # Resume scoring
    RESUME_COVERAGE_WEIGHT = float(os.getenv("RESUME_COVERAGE_WEIGHT", "0.7"))
    RESUME_CONCISENESS_WEIGHT = float(os.getenv("RESUME_CONCISENESS_WEIGHT", "0.3"))
    
    # Cover letter scoring  
    COVER_COVERAGE_WEIGHT = float(os.getenv("COVER_COVERAGE_WEIGHT", "0.6"))
    COVER_CONCISENESS_WEIGHT = float(os.getenv("COVER_CONCISENESS_WEIGHT", "0.4"))


# Default Values
class Defaults:
    """Default values for various components."""
    USER_ID = "default_user"
    
    # Mock profile
    MOCK_USER_NAME = "Alex Candidate"
    MOCK_USER_HEADLINE = "Senior Software Engineer"
    MOCK_USER_EMAIL = "alex@example.com"
    MOCK_USER_LOCATION = "Remote"
    MOCK_USER_SKILLS = [
        "Python", "AWS", "Docker", "Kubernetes", 
        "PostgreSQL", "Data Engineering"
    ]


def get_config() -> Dict[str, Any]:
    """Get all configuration as a dictionary."""
    return {
        "agent": {
            "optimization_cycles": AgentConfig.OPTIMIZATION_CYCLES,
            "resume_max_chars": AgentConfig.RESUME_MAX_CHARS,
            "cover_letter_max_chars": AgentConfig.COVER_LETTER_MAX_CHARS,
        },
        "llm": {
            "provider": LLMConfig.PROVIDER,
            "model": LLMConfig.MODEL,
            "temperature": LLMConfig.TEMPERATURE,
        },
        "api": {
            "mock_mode": APIConfig.MOCK_MODE,
            "http_timeout": APIConfig.HTTP_TIMEOUT,
        },
        "security": {
            "max_input_length": SecurityConfig.MAX_INPUT_LENGTH,
            "rate_limit": SecurityConfig.RATE_LIMIT_PER_USER,
        }
    }


# Export main config classes
__all__ = [
    "AgentConfig",
    "LLMConfig", 
    "APIConfig",
    "MemoryConfig",
    "SecurityConfig",
    "UIConfig",
    "ScoringConfig",
    "Defaults",
    "get_config"
]