Hapi-format-error NPM

hapi-format-error

A plugin to structure Boom errors into a more desirable format.

It currently does the following:

Returns errors in this format:

{
  "error": {
    "message": "error message",
    "status_code": 400
  }
}

Allows for changing the default Joi validation error status code (400)
Allows for using a custom server error message
Allows for logging server errors for debugging purposes
The final list of error messages is joined with or

Options

logServerError: boolean, default true - whether server errors (status code >= 500) should be logged to stdout
serverErrorMessage: string - any custom message you want to return for server errors
joiStatusCode: integer - the status code to use instead of 400 for Joi validation errors
language: object - language templates used to format specific errors; see below for details
permeateErrorCode: boolean, default false - whether to surface the .code property from the Boom Error in the API response

Providing Error Details

Additional error details can optionally be provided in the response object. This allows a client to make decisions based on granular error messages without having to parse the human-readable message field.

Set `permeateErrorCode: true`

This will add the .code property from the Boom Error to the response object. The HTTP response will look like the following:

{
  "error": {
    "message": "error message",
    "status_code": 400,
    "code": "unauthorized"
  }
}

Language Templates (Message Formatting)

hapi-format-error can massage Joi validation errors into simpler, concise messages. It does this by applying a language template, if available. These are similar to the messages option for Joi validation; the key difference is that they support plural forms.

A simple set of language templates that handle object.unknown errors would look like this:

{
  object: {
    unknown: {
      singular: '{path} is not allowed',
      plural: 'the following parameters are not allowed: {paths}'
    }
  }
}

hapi-format-error groups messages by their error type, and then looks for matching templates. The plural template, if provided, is used when more than one error of a given type is reported. If the plural template is not provided, the singular template (required) will be applied to each individual error.

Singular templates have the following context variables available: