A simple tool for loading YAML
and JSON
configuration/settings using
pydantic2
.
There is also a version for pydantic1
, see release/v1
. Major
versions of this package will match the major version of the respective
pydantic
release.
This project can be helpful for projects that have large configuration files,
nested configuration files, or for those of us who don't like writing large .env
files. It is also worth noting that due to the backwards compatability between
YAML
and JSON
that this will also parse JSON
configuration.
This can also be helpful when writing out application settings in kubernetes
/helm, where most configuration is written as YAML
. In such a case we may
want to validate/store our settings as YAML
as writing JSON
and
JSON
strings can be compersome due to syntax error in larger documents.
Install using pip
:
pip install yaml-settings-pydantic
First, it is worth reading the pydantic_settings
docs about additional sources: https://docs.pydantic.dev/latest/usage/pydantic_settings/
Additionally see the example in ./tests/examples/__init__.py
. It is gaurenteed to
work as its contents are tested. It contains information on how to write nested
configurations.
There are three classes worth knowing about:
YamlSettingsConfigDict
--pydantic_settings.SetttingsConfigDict
extended to include the fields used byCreateYamlSettings
.CreateYamlSettings
-- The pydanticPydanticBaseSettingsSource
that will analyze your class for the following class variables:- Files to be used -- under
__env_yaml_files__
ormodel_config.yaml_files
. - The reload settings -- under
__env_yaml_reload__
ormodel_config.yaml_reload
.
CreateYamlSettings
does not have to be used at all, but can be helpful if you don't want to useBaseYamlSettings
for any reason.- Files to be used -- under
BaseYamlSettings
-- Use this directly as done in the example below. This is 'the easy way'.
The shortest possible example is as follows:
from yaml_settings_pydantic import BaseYamlSettings
class MySettings(BaseYamlSettings):
__env_yaml_files__ = "settings.yaml"
setttingOne: str
settingTwo: str
...
...
Note that the above example can also be written like
from yaml_settings_pydantic import BaseYamlSettings, YamlSettingsConfigDict
class MySettings(BaseYamlSettings):
model_config = YamlSettingsConfigDict(yaml_files="settings.yaml")
setttingOne: str
settingTwo: str
...
...
which is more like pydantic v2. The 'dunder' specifications will take priority over their equivalent model_config specifications. These map as follows:
+-----------------------+------------------+
| dunder | model_config |
+-----------------------+------------------+
| __env_yaml_files__ | yaml_files |
| __env_yaml_reload__ | yaml_reload |
+-----------------------+------------------+