env/runtime_config.py

from __future__ import annotations

import os
from dataclasses import dataclass
from typing import Any


@dataclass(frozen=True)
class RuntimeConfig:
    seed: int = 42
    difficulty: str = "easy"
    use_quantizer: bool = False
    quant_mode: str = "full"
    quant_every_n_steps: int = 1
    embedding_dim: int = 16
    quant_bits: int = 3
    distortion_lambda: float = 0.2
    inner_product_lambda: float = 0.1

    @staticmethod
    def _to_bool(value: str | None, default: bool = False) -> bool:
        if value is None:
            return default
        return value.strip().lower() in {"1", "true", "yes", "on"}

    @classmethod
    def from_env(cls) -> "RuntimeConfig":
        quant_mode = os.getenv("OPENENV_QUANT_MODE", "full").lower()
        if quant_mode not in {"off", "full", "throttle", "status", "hybrid"}:
            quant_mode = "full"

        quant_every_n_steps = max(1, int(os.getenv("OPENENV_QUANT_EVERY_N_STEPS", 1)))
        seed_raw = os.getenv("OPENENV_SEED", os.getenv("ENV_SEED", 42))

        return cls(
            seed=int(seed_raw),
            difficulty=os.getenv("OPENENV_DIFFICULTY", "easy"),
            use_quantizer=cls._to_bool(os.getenv("OPENENV_USE_QUANTIZER"), False),
            quant_mode=quant_mode,
            quant_every_n_steps=quant_every_n_steps,
            embedding_dim=max(1, int(os.getenv("OPENENV_EMBEDDING_DIM", 16))),
            quant_bits=max(1, int(os.getenv("OPENENV_QUANT_BITS", 3))),
            distortion_lambda=float(os.getenv("OPENENV_DISTORTION_LAMBDA", 0.2)),
            inner_product_lambda=float(os.getenv("OPENENV_INNER_PRODUCT_LAMBDA", 0.1)),
        )

    def to_env_kwargs(self) -> dict[str, Any]:
        return {
            "difficulty": self.difficulty,
            "seed": self.seed,
            "use_quantizer": self.use_quantizer,
            "quant_mode": self.quant_mode,
            "quant_every_n_steps": self.quant_every_n_steps,
            "embedding_dim": self.embedding_dim,
            "quant_bits": self.quant_bits,
            "distortion_lambda": self.distortion_lambda,
            "inner_product_lambda": self.inner_product_lambda,
        }

    def as_dict(self) -> dict[str, Any]:
        return self.to_env_kwargs()