"""Configuration management for LinkedIn Workflow."""
from typing import Optional
from pydantic_settings import BaseSettings, SettingsConfigDict
from pathlib import Path


class Settings(BaseSettings):
    """Application settings loaded from environment variables."""
    # API Keys
    openai_api_key: str
    perplexity_api_key: str
    apify_api_key: str

    # Supabase
    supabase_url: str
    supabase_key: str
    supabase_service_role_key: str = ""  # Required for admin operations like deleting users

    # Apify
    apify_actor_id: str = "apimaestro~linkedin-profile-posts"

    # Web Interface
    web_password: str = ""
    session_secret: str = ""

    # Development
    debug: bool = False
    log_level: str = "INFO"

    # Email Settings
    smtp_host: str = ""
    smtp_port: int = 587
    smtp_user: str = ""
    smtp_password: str = ""
    smtp_from_name: str = "LinkedIn Post System"
    email_default_recipient: str = ""

    # Writer Features (can be toggled to disable new features)
    writer_multi_draft_enabled: bool = True  # Generate multiple drafts and select best
    writer_multi_draft_count: int = 3  # Number of drafts to generate (2-5)
    writer_semantic_matching_enabled: bool = True  # Use semantically similar example posts
    writer_learn_from_feedback: bool = True  # Learn from recurring critic feedback
    writer_feedback_history_count: int = 10  # Number of past posts to analyze for patterns

    # Quality Refiner Features (NEW - post-processing improvements)
    quality_refiner_enabled: bool = True  # Enable automatic quality refinement
    quality_smart_revision_enabled: bool = True  # Enable smart revision if auto-refiner not enough
    quality_min_acceptable_score: int = 80  # Minimum score to accept (even if not perfect)
    quality_target_score: int = 85  # Target score to achieve
    quality_max_smart_revisions: int = 1  # Max LLM revisions (0 = disabled, 1 = safe)

    # User Frontend (LinkedIn OAuth via Supabase)
    user_frontend_enabled: bool = True  # Enable user frontend with LinkedIn OAuth
    supabase_redirect_url: str = ""  # OAuth Callback URL (e.g., https://linkedin.onyva.dev/auth/callback)

    # LinkedIn API (Custom OAuth for auto-posting)
    linkedin_client_id: str = ""
    linkedin_client_secret: str = ""
    linkedin_redirect_uri: str = ""  # e.g., https://yourdomain.com/settings/linkedin/callback

    # Token Encryption
    encryption_key: str = ""  # Generate with: python -c "from cryptography.fernet import Fernet; print(Fernet.generate_key().decode())"

    # MOCO Integration
    moco_api_key: str = ""    # Token für Authorization-Header
    moco_domain: str = ""     # Subdomain: {domain}.mocoapp.com

    # Redis
    redis_url: str = "redis://redis:6379/0"
    scheduler_enabled: bool = False  # True only on dedicated scheduler container

    model_config = SettingsConfigDict(
        env_file=".env",
        env_file_encoding="utf-8",
        case_sensitive=False
    )


# Global settings instance
settings = Settings()

# API pricing per 1M tokens (input, output)
API_PRICING = {
    "gpt-4o": {"input": 2.50, "output": 10.00},
    "gpt-4o-mini": {"input": 0.15, "output": 0.60},
    "sonar": {"input": 1.00, "output": 1.00},
}


def estimate_cost(model: str, prompt_tokens: int, completion_tokens: int) -> float:
    """Estimate cost in USD for an API call."""
    pricing = API_PRICING.get(model, {"input": 1.00, "output": 1.00})
    input_cost = (prompt_tokens / 1_000_000) * pricing["input"]
    output_cost = (completion_tokens / 1_000_000) * pricing["output"]
    return input_cost + output_cost