When you have an error-ish but what you really want is an Error.
There are three main use cases for Errorish:
Error has a message, name, and stack properties.Error class to store an identifying label, a source error, and/or associated data.Exception is an Error extending class with additional label, error and data fields.ensure ensures any is an Error, otherwise creating one -it can optionally include a normalization step, enabled by default.normalize ensures an Error has a message, name, and stack properties -filling them if they're not defined.capture runs Error.captureStackTrace if running in V8 to clean up the error stack trace.ExceptionSee documentation for Exception.
Exception is an Error extending class that can store an identifying label, the source error that caused it and/or additional associated data. Exception also comes with several static and instance methods.
import { Exception } from 'errorish';
try {
try {
throw new Error('Source');
} catch (err) {
// throws with label
throw new Exception(['label', 'message'], err, { code: 401 });
}
} catch (err) {
// throws without label
throw new Exception(err.message, err, { code: 500 })
}
ensureEnsure will return its first argument if an instance of Error is passed as such, otherwise instantiating and returning an Exception.
import { ensure } from 'errorish';
ensure('foo'); // Error: foo
ensure(new Error('foo')); // Error: foo
ensure({ message: 'foo' }); // Error: foo
normalizeSee documentation for normalize.
Normalization fills an error's message, name, and stack property when empty. It's performed by default by ensure, but it can also be run independently.
import { normalize } from 'errorish';
normalize(new Error()); // Error: An error occurred
normalize(new Error(), { message: 'Foo bar' }); // Error: Foo bar
captureSee documentation for capture.
Captures the stack trace on Node and Chromium browsers.
import { capture } from 'errorish';
capture(new Error());