A library for Django Rest Framework returning consistent, predictable and easy-to-parse API error messages.
This library was built with RFC7807 guidelines in mind, but with a small twist: it defines a "problem detail" as a list, but it still serves as a way to include errors in a predictable and easy-to-parse format for any API consumer. Error messages are formatted using RFC7807 keywords and DRF exception data.
Compared to other similar and popular libraries, this library is based on RFC7807 guidelines and aims to provide not only a standardized format for error details, but also human readability.
Install using the command line:
pip install drf-simple-api-errors
Add EXCEPTION_HANDLER
in your REST_FRAMEWORK
settings of your Django project settings file:
REST_FRAMEWORK = {
# ...
"EXCEPTION_HANDLER": "drf_simple_api_errors.exception_handler",
}
API error messages typically include the following keys:
"title"
(str
): A brief summary that describes the problem type"detail"
(list[str] | None
): A list of specific explanations related to the problem"invalid_params"
(list[dict] | None
): A list of dict containing details about parameters that were invalid or malformed in the request. Each dict within this list provides:"name"
(str
): The name of the parameter that was found to be invalid"reasons"
(list[str]
): A list of strings describing the specific reasons why the parameter was considered invalid or malformed
{
"title": "Error message.",
"detail": [
"error",
...
],
"invalid_params": [
{
"name": "field_name",
"reason": [
"error",
...
]
},
...
]
}
{
"title": "Error message.",
"invalid_params": [
{
"name": "field_name",
"reason": [
"error",
...
]
},
...
]
}
{
"title": "Error message.",
"detail": [
"error",
...
]
}
{
"title": "Error message."
}
Default available settings:
DRF_SIMPLE_API_ERRORS = {
"CAMELIZE": False,
"EXTRA_HANDLERS": [],
"FIELDS_SEPARATOR": ".",
}
Camel case support for Django Rest Framework exceptions JSON error responses.
If CAMELIZE
is set to True
:
{
"title": "Error message.",
"invalidParams": [
{
"name": "fieldName",
"reason": [
"error",
...
]
}
...
]
}
Support for exceptions that differ from the standard structure of the Django Rest Framework.
For instance, you may want to specify you own exception:
class AuthenticationFailed(exceptions.AuthenticationFailed):
def __init__(self, detail=None, code=None):
"""
Builds a detail dictionary for the error to give more information
to API users.
"""
detail_dict = {"detail": self.default_detail, "code": self.default_code}
if isinstance(detail, dict):
detail_dict.update(detail)
elif detail is not None:
detail_dict["detail"] = detail
if code is not None:
detail_dict["code"] = code
super().__init__(detail_dict)
Use exception in code:
def my_func():
raise AuthenticationFailed(
{
"detail": _("Error message."),
"messages": [
{
"metadata": "metadata_data",
"type": "type_name",
"message": "error message",
}
],
}
)
This will result in:
AuthenticationFailed(
{
"detail": "Error message.",
"messages": [
{
"metadata": "metadata_data",
"type": "type_name",
"message": "error message",
}
],
}
)
You can handle this by creating a handlers.py
file and specifying an handler for your use case:
def handle_exc_custom_authentication_failed(exc):
from path.to.my.exceptions import AuthenticationFailed
if isinstance(exc, AuthenticationFailed):
try:
exc.detail = exc.detail["messages"][0]["message"]
except (KeyError, IndexError):
exc.detail = exc.detail["detail"]
return exc
Then add it to the EXTRA_HANDLERS
list in this package settings:
DRF_SIMPLE_API_ERRORS = {
"EXTRA_HANDLERS": [
"path.to.my.handlers.handle_exc_custom_authentication_failed",
# ...
]
}
Support for nested dicts containing multiple fields to be flattened.
If FIELDS_SEPARATOR
is set to .
:
{
"field1": {
"field2": "value"
}
}
Will result in:
{
"field1.field2": "value"
}
All the necessary commands are included in the Makefile
.
We are using tox
and poetry
to run tests in every supported Python version.
Run test with the commands below:
make install
make test
Please open an issue.
Please use the Github Flow. In a nutshell, create a branch, commit your code, and open a pull request.