Skip to content
This repository has been archived by the owner on Apr 26, 2022. It is now read-only.

[pull] master from facebook:master #36

Open
wants to merge 474 commits into
base: master
Choose a base branch
from
Open

Conversation

pull[bot]
Copy link

@pull pull bot commented May 12, 2019

See Commits and Changes for more details.


Created by pull[bot]

Can you help keep this open source service alive? 💖 Please sponsor : )

@pull pull bot added the ⤵️ pull label May 12, 2019
@pull pull bot requested a review from prichodko May 12, 2019 11:39
heyimalex and others added 28 commits December 30, 2019 13:53
The old favicon was the same as the official react documentation, which is a minor annoyance during development when trying to find the tab you want. The new favicon is just the old with inverted colors.

Closes #7957
This removes `React.FC` from the base template for a Typescript project.

Long explanation for a small change: 

`React.FC` is unnecessary: it provides next to no benefits and has a few downsides.  (See below.)  I see a lot of beginners to TS+React using it, however, and I think that it's usage in this template is a contributing factor to that, as the prominence of this template makes it a de facto source of "best practice".  

### Downsides to React.FC/React.FunctionComponent

##### Provides an implicit definition of `children`

Defining a component with `React.FC` causes it to implicitly take `children` (of type `ReactNode`).  It means that all components accept children, even if they're not supposed to, allowing code like:

```ts
const App: React.FC = () => { /*... */ };
const Example = () => {
	<App><div>Unwanted children</div></App>
}
```

This isn't a run-time error, but it is a mistake and one that would be caught by Typescript if not for `React.FC`. 

##### Doesn't support generics.
I can define a generic component like:
```ts
type GenericComponentProps<T> = {
   prop: T
   callback: (t: T) => void
}
const GenericComponent = <T>(props: GenericComponentProps<T>) => {/*...*/}
```

But it's not possible when using `React.FC` - there's no way to preserve the unresolved generic `T` in the type returned by `React.FC`.

```ts
const GenericComponent: React.FC</* ??? */> = <T>(props: GenericComponentProps<T>) => {/*...*/}
```

##### Makes "component as namespace pattern" more awkward.
It's a somewhat popular pattern to use a component as a namespace for related components (usually children):

```jsx
<Select>
	<Select.Item />
</Select>
```

This is possible, but awkward, with `React.FC`:

```tsx
const  Select: React.FC<SelectProps> & { Item: React.FC<ItemProps> } = (props) => {/* ... */ }
Select.Item = (props) => { /*...*/ }
```

but "just works" without `React.FC`:

```tsx
const Select = (props: SelectProps) => {/* ... */}
Select.Item = (props) => { /*...*/ }
```

##### Doesn't work correctly with defaultProps

This is a fairly moot point as in both cases it's probably better to use ES6 default arguments, but...

```tsx
type  ComponentProps = { name: string; }

const  Component = ({ name }: ComponentProps) => (<div>
	{name.toUpperCase()} /* Safe since name is required */
</div>);
Component.defaultProps = { name: "John" };

const  Example = () => (<Component />) /* Safe to omit since name has a default value */
```
This compiles correctly.  Any approach with `React.FC` will be slightly wrong: either `React.FC<{name: string}>` will make the prop required by consumers, when it should be optional, or `React.FC<{name?: string}>` will cause `name.toUpperCase()` to be a type error.  There's no way to replicate the "internally required, externally optional" behavior which is desired.

##### It's as long, or longer than the alternative: (especially longer if you use `FunctionalComponent`):
Not a huge point, but it isn't even shorter to use `React.FC` 
```ts
const C1: React.FC<CProps> = (props) => { }
const C2 = (props: CProps) => {};
```

### Benefits of React.FC

##### Provides an explicit return type

The only benefit I really see to `React.FC` (unless you think that implicit `children` is a good thing) is that it specifies the return type, which catches mistakes like:

```ts
const Component = () => {
   return undefined; // components aren't allowed to return undefined, just `null`
}
```

In practice, I think this is fine, as it'll be caught as soon as you try to use it:

```ts
const Example = () => <Component />; // Error here, due to Component returning the wrong thing
```

But even with explicit type annotations, `React.FC` still isn't saving very much boilerplate:

```ts
const Component1 = (props: ComponentProps): ReactNode => { /*...*/ }
const Component2: FC<ComponentProps> = (props) => { /*...*/ }
```
Updates dependencies and removes babel plugins that are now covered by `@babel/preset-env`.

Co-authored-by: hdineen <hdineen@hubspot.com>
…late locally (#8092)

Co-authored-by: Ian Schmitz <ianschmitz@gmail.com>
`Auto Fix is enabled by default. Use the single string form.` warning is shown in `.vscode/settings.json` due to changes in vscode-eslint. 
As autoFix is set to default, object format in `eslint.validate` is deprecated.
Co-authored-by: Ian Schmitz <ianschmitz@gmail.com>
Co-authored-by: Ian Schmitz <ianschmitz@gmail.com>
 - babel-plugin-named-asset-import@0.3.6
 - babel-preset-react-app@9.1.1
 - cra-template-typescript@1.0.1
 - cra-template@1.0.1
 - create-react-app@3.3.1
 - eslint-config-react-app@5.2.0
 - react-app-polyfill@1.0.6
 - react-dev-utils@10.1.0
 - react-error-overlay@6.0.5
 - react-scripts@3.3.1
* Expands scope of openBrowser tab control

Adjust openChrome.applescript to allow manipulation of
other Chromium-based browsers (defaulting to Chrome).
Requires list of compatible browsers to try in openBrowser.js

* Fix typo

* Remove Safari
* setupTestFrameworkScriptFile is deprecated

__Note:_ `_setupTestFrameworkScriptFile_` _is deprecated in favor of_ `_setupFilesAfterEnv_`_.__

ref: https://jestjs.io/docs/en/configuration#setupfilesafterenv-array

* Update docusaurus/docs/running-tests.md

Co-Authored-By: Simen Bekkhus <sbekkhus91@gmail.com>

Co-authored-by: Simen Bekkhus <sbekkhus91@gmail.com>
Added the relevant topic linked directly, instead of saying it's just in the next section.
Adjusted the text so it sounds more like a recommendation rather than a requirement.
The topic linked is basically explainer to how you'd reference it using the global window object and ways to avoid linter errors. Thus I used 'reference' to give users a hint of what it the linked page would be for.
wclem4 and others added 28 commits February 22, 2021 10:53
Resolves #10411

Bumps immer version to 8.0.1 to address the prototype pollution
vulnerability with the current 7.0.9 version.
 - create-react-app@4.0.3
 - react-dev-utils@11.0.3
 - react-scripts@4.0.3
The variable is also used in dev.
* Update getProcessForPort.js

* Update getProcessForPort.js

Co-authored-by: Zhou Peng <zpbrent@gmail.com>
Co-authored-by: Dan Abramov <dan.abramov@gmail.com>
Co-authored-by: Ian Schmitz <ianschmitz@gmail.com>
Bumps [y18n](https://github.com/yargs/y18n) from 4.0.0 to 4.0.1.
- [Release notes](https://github.com/yargs/y18n/releases)
- [Changelog](https://github.com/yargs/y18n/blob/master/CHANGELOG.md)
- [Commits](https://github.com/yargs/y18n/commits)

Signed-off-by: dependabot[bot] <support@github.com>

Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com>
This just fixes a shell snippet in the readme file for this plugin
Replace the Github home link with a link to the repo's main page or a link to the source (https://github.com/CodeByZach/pace/blob/master/pace.js)
Bump immer minor version to fix `Prototype Pollution` Security issue.
Co-authored-by: Brody McKee <mrmckeb@gmail.com>
@facebook-github-bot facebook-github-bot deleted the branch strvcom:master July 7, 2021 15:44
@facebook-github-bot facebook-github-bot deleted the master branch July 7, 2021 15:44
Sign up for free to subscribe to this conversation on GitHub. Already have an account? Sign in.