react-native/ReactAndroid
Tim Yung f090840f45 RN: Revert D8666509
Summary:
Reverts D8666509. Unfortunately, I misunderstood `setClipChildren` on Android.

When set on a `ViewGroup`, `setClipChildren` configures whether its //children// — not itself — are clipped to their bounds. This is unlike `overflow` (as it behaves on iOS) which configures whether the view itself is clipped to its bounds.

But they are definitely related. In theory, I think we could implement `overflow` using `setClipChildren` by:

- Setting `setClipChildren(false)` by default. (This part, I got right.)
- When `overflow` is set to `hidden` on a `View`, we create an extra `ViewGroup` (child) within the normal `ViewGroup` (parent). Then, we can set `setClipChildren(true)` on the parent `ViewGroup` which will cause the child `ViewGroup` to be clipped to its bounds.

However, I think the tricky thing will be to create the child `ViewGroup` without incurring unintentional side effects.

I need to decide whether or not this is worth trying. The alternative is to add a new `clipChildren` boolean prop that is Android-only, but I really hate further bifurcating the platform. But for now, I am reverting my mistake.

Reviewed By: achen1

Differential Revision: D8690551

fbshipit-source-id: 1ba3bbcc5458ffbd5c475430ea0382b3fd0916b2
2018-06-29 12:18:27 -07:00
..
libs use android build-tools 26.0.2 and set compileSdk to 26 (#19662) 2018-06-13 10:39:55 -07:00
src RN: Revert D8666509 2018-06-29 12:18:27 -07:00
.npmignore
DevExperience.md
README.md Fixing link to Android build guide. 2018-05-27 15:17:55 -07:00
build.gradle android targetSdkVersion is 26 (#19944) 2018-06-28 11:49:32 -07:00
gradle.properties
proguard-rules.pro remove fresco proguard rules 2018-05-14 11:18:05 -07:00
release.gradle Migrate Android sources to MIT license header 2018-05-31 15:37:30 -07:00

README.md

Building React Native for Android

See the docs on the website.

Running tests

When you submit a pull request CircleCI will automatically run all tests. To run tests locally, see Testing.