Spaces:
Running
Running
File size: 5,545 Bytes
3193174 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 | import os
from pathlib import Path
from pydantic import Field, SecretStr, ValidationError, field_validator, model_validator
from pydantic_settings import BaseSettings, SettingsConfigDict
__all__ = ["FrameworkSettings", "load_env_file", "load_settings"]
class FrameworkSettings(BaseSettings):
"""
Framework settings loaded from the environment with the `RWXF_` prefix.
Key fields:
- `api_key` / `api_key_file`: secret key directly or path to a file.
- `base_url`: base URL for the LLM service.
- `model_name`: generation model identifier.
- `embedding_model`: embedding model identifier.
- `log_*`: logging parameters.
- `default_timeout`, `max_retries`: network timeouts and retries.
"""
model_config = SettingsConfigDict(env_prefix="RWXF_", extra="ignore")
api_key: SecretStr | None = Field(default=None, description="API key for LLM service")
api_key_file: Path | None = Field(
default=None,
description="Path to a file that stores the API key securely",
)
base_url: str | None = Field(default=None, description="Base URL for LLM service")
model_name: str = Field(default="gpt-4o-mini", description="LLM model identifier")
embedding_model: str = Field(
default="sentence-transformers/all-MiniLM-L6-v2",
description="Embedding model identifier",
)
embedding_normalize: bool = Field(default=True, description="Normalize embeddings")
embedding_fallback_dim: int = Field(default=384, description="Fallback dimension")
log_level: str = Field(default="INFO", description="Logging level")
log_file: str | None = Field(default=None, description="Log file path")
log_backtrace: bool = Field(default=False, description="Enable backtrace")
default_timeout: int = Field(default=60, description="Default timeout in seconds")
max_retries: int = Field(default=3, description="Max retries for LLM calls")
@field_validator("embedding_model")
@classmethod
def _validate_embedding_model(cls, value: str) -> str:
"""Validate the embedding model name."""
if value == "hash" or value.startswith("hash:"):
return value
if value.startswith(("sentence-transformers/", "sentence-transformers:")):
return value
msg = "Unsupported embedding model. Use 'sentence-transformers/<model>' or 'hash[:<dim>]'"
raise ValueError(msg)
@field_validator("*", mode="before")
@classmethod
def _handle_empty_strings(cls, value):
"""Convert empty strings to None for correct validation."""
if isinstance(value, str) and value.strip() == "":
return None
return value
@field_validator("api_key_file")
@classmethod
def _validate_api_key_file(cls, value: Path | None) -> Path | None:
"""Ensure the key file exists before reading."""
if value is None:
return None
if not value.is_file():
msg = f"API key file not found: {value}"
raise ValueError(msg)
return value
@model_validator(mode="after")
def _load_secret_key(self) -> "FrameworkSettings":
"""Load the key from a file if not set directly, and require its presence."""
if self.api_key is None and self.api_key_file is not None:
content = self.api_key_file.read_text(encoding="utf-8").strip()
if not content:
msg = "API key file is empty"
raise ValueError(msg)
object.__setattr__(self, "api_key", SecretStr(content))
if self.api_key is None:
msg = "api_key is required via RWXF_API_KEY or RWXF_API_KEY_FILE"
raise ValueError(msg)
return self
@property
def resolved_api_key(self) -> str:
"""Return the secret api_key value or raise an error if it is absent."""
if self.api_key is None:
msg = "API key is not configured"
raise RuntimeError(msg)
return self.api_key.get_secret_value()
def load_env_file(path: Path | str | None = None) -> None:
"""
Load environment variables from a .env file (if it exists).
Args:
path: Path to the .env file; defaults to the current directory.
"""
env_path = Path(path or ".env")
if not env_path.exists():
return
with env_path.open("r", encoding="utf-8") as handle:
for raw_line in handle:
line = raw_line.strip()
if not line or line.startswith("#"):
continue
if "=" not in line:
continue
key, value = line.split("=", 1)
key = key.strip()
if not key:
continue
cleaned = value.strip().strip('"').strip("'")
if key not in os.environ:
os.environ[key] = cleaned
def load_settings(path: Path | str | None = None) -> FrameworkSettings:
"""
Read the .env file (if provided), load, and validate settings.
Args:
path: Path to the .env file for pre-loading the environment.
Returns:
Validated `FrameworkSettings` instance.
Raises:
RuntimeError: if settings validation failed.
"""
load_env_file(path)
try:
settings = FrameworkSettings()
except ValidationError as exc:
messages = [err.get("msg", "invalid configuration value") for err in exc.errors()]
detail = "; ".join(messages)
raise RuntimeError(detail) from exc
return settings
|