backend/app/config.py

"""
Configuration management for SLM Code Engine
"""
from pathlib import Path
from typing import Optional
from pydantic_settings import BaseSettings
from pydantic import Field


class Settings(BaseSettings):
    """Application settings with environment variable support"""

    # API Configuration
    api_host: str = Field(default="0.0.0.0", env="API_HOST")
    api_port: int = Field(default=8000, env="API_PORT")
    api_workers: int = Field(default=1, env="API_WORKERS")
    debug: bool = Field(default=True, env="DEBUG")

    # Groq Configuration
    groq_api_key: Optional[str] = Field(default=None, env="GROQ_API_KEY")

    # Localization
    language: str = Field(
        default="en",
        env="LANGUAGE",
        description="Language for responses (e.g., 'en', 'fr')",
    )

    # Project paths
    project_root: Path = Path(__file__).parent.parent.parent
    models_dir: Path = Field(default_factory=lambda: Path(__file__).parent.parent.parent / "models")
    data_dir: Path = Field(default_factory=lambda: Path(__file__).parent.parent.parent / "data")
    cache_dir: Path = Field(default_factory=lambda: Path(__file__).parent.parent.parent / "data" / "cache")

    # Models Configuration
    starcoder_model: str = Field(default="phi-2.Q4_K_M.gguf", env="STARCODER_MODEL")
    codet5_model: str = Field(default="codet5-small", env="CODET5_MODEL")
    embedding_model: str = Field(default="all-MiniLM-L6-v2", env="EMBEDDING_MODEL")

    # Model inference settings
    max_tokens: int = Field(default=2048, env="MAX_TOKENS")
    temperature: float = Field(default=0.2, env="TEMPERATURE")
    n_ctx: int = Field(default=4096, env="N_CTX")  # Context window
    n_threads: Optional[int] = Field(default=None, env="N_THREADS")  # CPU threads

    # Database
    db_path: Path = Field(default_factory=lambda: Path(__file__).parent.parent.parent / "data" / "usage.db")

    # Sandbox Configuration
    sandbox_enabled: bool = Field(default=True, env="SANDBOX_ENABLED")
    sandbox_timeout: int = Field(default=30, env="SANDBOX_TIMEOUT")  # seconds
    sandbox_memory_limit: str = Field(default="512m", env="SANDBOX_MEMORY_LIMIT")

    # Orchestrator Configuration
    router_threshold: float = Field(default=0.7, env="ROUTER_THRESHOLD")  # Confidence threshold
    enable_automata_first: bool = Field(default=True, env="ENABLE_AUTOMATA_FIRST")
    enable_automata_first: bool = Field(default=True, env="ENABLE_AUTOMATA_FIRST")
    enable_rag: bool = Field(default=True, env="ENABLE_RAG")  # Enable RAG context enrichment
    enable_distillation: bool = Field(default=True, env="ENABLE_DISTILLATION")  # Enable data collection

    # Logging
    log_level: str = Field(default="INFO", env="LOG_LEVEL")
    log_file: Optional[Path] = Field(default=None, env="LOG_FILE")

    class Config:
        # Look for .env in project root
        env_file = str(Path(__file__).parent.parent.parent / ".env")
        env_file_encoding = "utf-8"
        case_sensitive = False
        extra = "ignore"  # Ignore extra fields in .env

    def __init__(self, **kwargs):
        super().__init__(**kwargs)
        # Ensure directories exist
        self.models_dir.mkdir(parents=True, exist_ok=True)
        self.data_dir.mkdir(parents=True, exist_ok=True)
        self.cache_dir.mkdir(parents=True, exist_ok=True)

    @property
    def starcoder_path(self) -> Path:
        """Get full path to StarCoder model (currently using Phi-2)"""
        return self.models_dir / "phi-2" / self.starcoder_model

    @property
    def codet5_path(self) -> Path:
        """Get full path to CodeT5 model"""
        return self.models_dir / "codet5-small"


# Global settings instance
settings = Settings()