Component Story Format (CSF)

Component Story Format (CSF) is the recommended way to write stories. It's an open standard based on ES6 modules that is portable beyond Storybook.

In CSF, stories and component metadata are defined as ES Modules. Every component story file consists of a required default export and one or more named exports.

Default export

The default export defines metadata about your component, including the component itself, its title (where it will show up in the navigation UI story hierarchy), decorators, and parameters.

The component field is required and used by addons for automatic prop table generation and display of other component metadata. The title field is optional and should be unique (i.e., not re-used across files).

For more examples, see writing stories.

Named story exports

With CSF, every named export in the file represents a story object by default.

The exported identifiers will be converted to "start case" using Lodash's startCase function. For example:

| Identifier | Transformation | | ---------------- | ----------------- | | name | Name | | someName | Some Name | | someNAME | Some NAME | | some_custom_NAME | Some Custom NAME | | someName1234 | Some Name 1 2 3 4 |

We recommend that all export names to start with a capital letter.

Story objects can be annotated with a few different fields to define story-level decorators and parameters, and also to define the name of the story.

Storybook's name configuration element is helpful in specific circumstances. Common use cases are names with special characters or Javascript restricted words. If not specified, Storybook defaults to the named export.

Args story inputs

Starting in SB 6.0, stories accept named inputs called Args. Args are dynamic data that are provided (and possibly updated by) Storybook and its addons.

Consider Storybook’s "Button" example of a text button that logs its click events:

Now consider the same example, re-written with args:

Or even more simply:

Not only are these versions shorter and more accessible to write than their no-args counterparts, but they are also more portable since the code doesn't depend on the actions addon specifically.

For more information on setting up Docs and Actions, see their respective documentation.

Play function

Storybook's play functions are small snippets of code executed when the story renders in the UI. They are convenient helper methods to help you test use cases that otherwise weren't possible or required user intervention.

A good use case for the play function is a form component. With previous Storybook versions, you'd write your set of stories and had to interact with the component to validate it. With Storybook's play functions, you could write the following story:

When the story renders in the UI, Storybook executes each step defined in the play function and runs the assertions without the need for user interaction.

Storybook export vs. name handling

Storybook handles named exports and the name option slightly differently. When should you use one vs. the other?

Storybook will always use the named export to determine the story ID and URL.

If you specify the name option, it will be used as the story display name in the UI. Otherwise, it defaults to the named export, processed through Storybook's storyNameFromExport and lodash.startCase functions.

it('should format CSF exports with sensible defaults', () => {
  const testCases = {
    name: 'Name',
    someName: 'Some Name',
    someNAME: 'Some NAME',
    some_custom_NAME: 'Some Custom NAME',
    someName1234: 'Some Name 1234',
    someName1_2_3_4: 'Some Name 1 2 3 4',
  };
  Object.entries(testCases).forEach(([key, val]) => {
    expect(storyNameFromExport(key)).toBe(val);
  });
});

When you want to change the name of your story, rename the CSF export. It will change the name of the story and also change the story's ID and URL.

It would be best if you used the name configuration element in the following cases:

1. You want the name to show up in the Storybook UI in a way that's not possible with a named export, e.g., reserved keywords like "default", special characters like emoji, spacing/capitalization other than what's provided by storyNameFromExport.
2. You want to preserve the Story ID independently from changing how it's displayed. Having stable Story IDs is helpful for integration with third-party tools.

Non-story exports

In some cases, you may want to export a mixture of stories and non-stories (e.g., mocked data).

You can use the optional configuration fields includeStories and excludeStories in the default export to make this possible. You can define them as an array of strings or regular expressions.

Consider the following story file:

When this file renders in Storybook, it treats ComplexStory and SimpleStory as stories and ignores the data named exports.

For this particular example, you could achieve the same result in different ways, depending on what's convenient:

• includeStories: /^[A-Z]/
• includeStories: /.*Story$/
• includeStories: ['SimpleStory', 'ComplexStory']
• excludeStories: /^[a-z]/
• excludeStories: /.*Data$/
• excludeStories: ['simpleData', 'complexData']

The first option is the recommended solution if you follow the best practice of starting story exports with an uppercase letter (i.e., use UpperCamelCase).

Upgrading from CSF 2 to CSF 3

In CSF 2, the named exports are always functions that instantiate a component, and those functions can be annotated with configuration options. For example:

This declares a Primary story for a Button that renders itself by spreading { primary: true } into the component. The default.title metadata says where to place the story in a navigation hierarchy.

Here's the CSF 3 equivalent:

Let's go through the changes individually to understand what's going on.

Spreadable story objects

In CSF 3, the named exports are objects, not functions. This allows us to reuse stories more efficiently with the JS spread operator.

Consider the following addition to the intro example, which creates a PrimaryOnDark story that renders against a dark background:

Here's the CSF 2 implementation:

export const PrimaryOnDark = Primary.bind({});
PrimaryOnDark.args = Primary.args;
PrimaryOnDark.parameters = { background: { default: 'dark' } };

Primary.bind({}) copies the story function, but it doesn't copy the annotations hanging off the function, so we must add PrimaryOnDark.args = Primary.args to inherit the args.

In CSF 3, we can spread the Primary object to carry over all its annotations:

Learn more about named story exports.

Default render functions

In CSF 3, you specify how a story renders through a render function. We can rewrite a CSF 2 example to CSF 3 through the following steps.

Let's start with a simple CSF 2 story function:

Now, let's rewrite it as a story object in CSF 3 with an explicit render function that tells the story how to render itself. Like CSF 2, this gives us full control of how we render a component or even a collection of components.

But in CSF 2, a lot of story functions are identical: take the component specified in the default export and spread args into it. What's interesting about these stories is not the function, but the args passed into the function.

CSF 3 provides default render functions for each renderer. If all you're doing is spreading args into your component—which is the most common case—you don't need to specify any render function at all:

export const Default = {};

Generate titles automatically

Finally, CSF 3 can automatically generate titles.

export default {
  title: 'components/Button',
  component: Button,
};

export default { component: Button };

You can still specify a title like in CSF 2, but if you don't specify one, it can be inferred from the story's path on disk. For more information, see the section on configuring story loading.