Decorator parameters

The main functional component of the library is the @properties decorator, which, like the @dataclass decorator, can be used with or without brackets, so the following class definitions are similar:

from conjector import properties


@properties
class FirstClass:
    pass


@properties()
class SecondClass:
    pass


@properties(
    filename="application.yml",
    override_default=False,
    root="",
    type_cast=True,
    lazy_init=False,
)
class ThirdClass:
    pass

Each decorator parameter can be overridden using a global configuration file.

Config filename

filename - the name of a file with config. By default, it is application.yml. For now, yaml, json and toml config formats are fully supported, ini is supported but with some limitations. Config file will be searched in the same directory where is file with used decorator. Use a relative path to read the file from a parent directory.

Suppose we have the following project structure:

<project_root>
|---services
|   |   email_message_service.py
|   |   auth_service.py
|   |   .....
|---configs
|   |   service_config.toml
|   |   .....

Our file with the constants is located in the configs directory, and a class that uses these values is in email_message_service.py which is located in the services directory. Then the path to the file will be written as follows - filename="../configs/service_config.toml"

Disabling type casting

type_cast - used to know whether you want to cast config values to the field type. By default, it’s True, which means values in a config file will be cast according to the type hints. All types specified in the section supported types will be available for type casting. Also, nested types will be recursively cast. If False, type hinting is ignored, and available types are limited by a file format:

Python	JSON	YAML	TOML	INI
str	+	+	+	+
int	+	+	+	-
float	+	+	+	-
bool	+	+	+	-
none	+	+	+	-
dict	+	+	+	+
list	+	+	+	+
datetime	-	-	+	-

Override default params

override_default - used to know whether you want to override the default values of class variables. By default, it is False.

Enable lazy initialization

lazy_init - used to know whether you want to set config values immediately on the application start-up or on demand (“lazily”) after calling the method init_props(). By default, it is False. For more details read section lazy initialization

Specify root path of config

root - root key in the config. It’s the way to create “namespaces” when you work with multiple classes but use a single config file. It could be a nested value with separation by dots, for example:

example.yml:

services:
  email_service:
    key: some value
  auth_service:
    key: another value

clients:
  translation_client:
    key: value

# and so on...

services.py:

from conjector import properties


@properties(filename="example.yml", root="services.email_service")
class EmailService:
    key: str  # will store "some value"


@properties(filename="example.yml", root="services.auth_service")
class AuthService:
    key: str  # will store "another value"