typed-configparser is an extension of the standard configparser module with support for typed configurations using dataclasses. It leverages Python's type hints and dataclasses to provide a convenient way of parsing and validating configuration files.
✓ Fully typed.
✓ Use dataclasses to parse the configuration file.
✓ Support for almost all python built-in data types - int
, float
, str
, list
, tuple
, dict
and complex data types using Union
and Optional
.
✓ Supports almost all features of dataclasses including field level init flag, post_init method, InitVars and more.
✓ Built on top of configparser
, hence retains all functionalities of configparser
.
✓ Support for optional values (optional values are automatically set to None
if not provided).
✓ Smarter defaults (see below).
You can install typed_configparser
using pip
:
pip install typed-configparser
examples/basic.py
# This is a complete example and should work as is
from typing import List
from typed_configparser import ConfigParser
from dataclasses import dataclass
@dataclass
class BASIC:
option1: int
option2: str
option3: float
option4: List[str]
config = """
[BASIC]
option1 = 10
option2 = value2
option3 = 5.2
option4 = [foo,bar]
"""
parser = ConfigParser()
parser.read_string(config)
section = parser.parse_section(using_dataclass=BASIC)
print(section)
BASIC(option1=10, option2=value2, option3=5.2, option4=['foo', 'bar'])
examples/unions_and_optionals.py
# This is a complete example and should work as is
from typing import List, Union, Optional, Dict, Tuple
from typed_configparser import ConfigParser
from dataclasses import dataclass, field
@dataclass
class DEFAULT_EXAMPLE:
option1: int
option2: Union[List[Tuple[str, str]], List[int]]
option3: Dict[str, str] = field(default_factory=lambda: {"default_key": "default_value"})
option4: Optional[float] = None
config = """
[DEFAULT]
option1 = 20
option2 = default_value2
[MY_SECTION_1]
option2 = [10,20]
option4 = 5.2
[MY_SECTION_2]
option2 = [(value2a, value2b), (value2c, value2b), (value2c, value2d)]
option3 = {key: value}
option4 = none
"""
parser = ConfigParser()
parser.read_string(config)
my_section_1 = parser.parse_section(using_dataclass=DEFAULT_EXAMPLE, section_name="MY_SECTION_1")
my_section_2 = parser.parse_section(using_dataclass=DEFAULT_EXAMPLE, section_name="MY_SECTION_2")
print(my_section_1)
print(my_section_2)
DEFAULT_EXAMPLE(option1=20, option2=[10, 20], option3={'default_key': 'default_value'}, option4=5.2)
DEFAULT_EXAMPLE(option1=20, option2=[('value2a', 'value2b'), ('value2c', 'value2b'), ('value2c', 'value2d')], option3={'key': 'value'}, option4=None)
Check example
directory for more examples.
configparser
includes sensible defaults options which allows you to declare a[DEFAULT]
section in the config file for fallback values.typed_configparser
goes a step further and allows you to set a final (last) level of defaults at dataclass level.
If you are interested in contributing to typed_configparser, please take a look at the contributing guidelines.