Summary:
@public
The info about bridge modules (such as id, name, queue, methods...) was spread
across arrays & dictionaries on the bridge, move it into a specific class.
It also removes a lot of information that was statically cached, and now have
the same lifecycle of the bridge.
Also moved RCTModuleMethod, RCTFrameUpdate and RCTBatchedBridge into it's own
files, for organization sake.
NOTE: This diff seems huge, but most of it was just moving code :)
Test Plan:
Tested UIExplorer & UIExplorer tests, Catalyst, MAdMan and Groups. Everything
looks fine.
Summary:
Added remapping of style textAlign property to textAlignment (which is the name of the same property on text input components)
Closes https://github.com/facebook/react-native/pull/772
Github Author: Daryl Rowland <daryl@cloudyclear.com>
Test Plan: Not sure. I tried cleaned and rebuilt and my TextInput works, though! Do we need to set up any unit tests or examples for this?
Summary:
@public
We have a function that detects whether a give file is to be treated as a node_modules. If so it doesn't have access to the haste module map. There is an exception to this rule which is a few modules that are allowed to do that. Currently thats react-native, react-tools, and parse.
The current implementation had a bug where if you had `react-native` (or react-tools etc) in the name before the actual package root then the detection will be off. This fixes the problem by starting from the `lastIndexOf('node_modules')` directory, that way nothing confuses us.
Test Plan:
./runJestTests.sh
export OSS, patch, run e2e test
Summary:
@public
The current output of console.log is extremely bad. If you pass NaN, it shows up as null (super confusing I know -_-), if you pass a cyclical object, it just says that it is cyclic and that's it. It doesn't print up the first few levels which are NOT cyclical and would be really helpful.
It turns out that nodejs console.log pretty printer is really awesome and can be easily extracted as a few hundred lines. This is going to be such a productivity boost that I think it's the right tradeoff to embed it like this
Test Plan:
```
var a = {kikoo: {lol: 1}}
a.kikoo.nice = a;
console.log(a);
> { kikoo: { lol: 1, nice: [Circular] } }
console.log(NaN)
> NaN
```
Summary:
### TL/DR:
```
a="function() {return [22]}"
a.substring(a.indexOf("{")+1,a.indexOf("}")-1) // "return [22"
a.substring(a.indexOf("{")+1,a.indexOf("}")) // "return [22]"
```
### In long: why it is broken now and why it worked before:
I've installed latest iOS 9 and started to see really strange issues when code is minified:
```
Invariant Violation: Application app has not been registered."
2015-06-18 16:29:05.898 [error][tid:com.facebook.React.JavaScript] "Error: Unexpected identifier 'transformMatrix'. Expected ']' to end a subscript expression
```
After some investigation it turns out that new Safari returned a bit different string representation for a MatrixOps.unroll. On old safari:
`function(e,t,n,r,o,i,a,s,u,c,l,p,d,h,f,m,g){t=e[0],n=e[1],r=e[2],o=e[3],i=e[4],a=e[5],s=e[6],u=e[7],c=e[8],l=e[9],p=e[10],d=e[11],h=e[12],f=e[13],m=e[14],g=e[15];}`
while using latest iOS:
`function (e,t,n,r,o,i,a,s,u,c,l,p,d,h,f,m,g){t=e[0],n=e[1],r=e[2],o=e[3],i=e[4],a=e[5],s=e[6],u=e[7],c=e[8],l=e[9]
Closes https://github.com/facebook/react-native/pull/1672
Github Author: Artem Yarulin <artem.yarulin@fessguid.com>
Test Plan: Imported from GitHub, without a `Test Plan:` line.
Summary:
@public
For some weird reason sometimes displayName is undefined on the OSS tests, manually
set it for now to unbreak the tests.
Test Plan: Run the tests on OSS
Summary:
resizeMode is a native prop, but it is also in the propTypes, so this causes an incorrect warning:
```
Prop resizeMode = `contain` should not be set directly on Image.
```
@public
Test Plan: No warnings on image example in UIExplorer
Summary:
@public
I've added an alert to notify when the profile has been generated, but it was
being created out of the main thread.
Test Plan:
Launch the UIExplorer, start and stop profiling, an alert should show up, and
everything should just keep working as expected.
Summary:
@public
If something changes in the list view that should trigger more loads, it
wouldn't. Example case is tap to load more - only the first new row would load,
but it wouldn't trigger a re-measure and subsequent layout of additional new
rows.
Test Plan: View More in Events works.
Summary:
@public
When the profile is initialized, it automatically hooks into every method of
the bridge modules, that was causing `+initialize` to be called twice.
Also add a popup to notify the user that the profile has been created.
Test Plan:
Run the UIExplorer, start the profiler, try to rage shake to open the dev menu
again. It should now work, and show an alertview with some information once the
profile is stopped.
Summary:
@public
The current getter for `navigationContext` always return a static
context, and it should return an instance-based one, instead.
Test Plan:
Use console.log() in inspect that two different navigators do
have their own `navigationContext` created.
Summary:
Live reload is disabled when an error has occurred. This requires the developer to fix the error and then switch to the simulator to reload the device manually; impacting developer flow and increasing alt tabbing. This pull request fixes that by allowing live reload to work even on errors.
This fixes issue: #1343.
Closes https://github.com/facebook/react-native/pull/1549
Github Author: Ruben Niculcea <ruben.niculcea@gmail.com>
Test Plan: Imported from GitHub, without a `Test Plan:` line.
Summary:
@public
That was eventually being released before all the queues had been cleared.
Update it so the each modules' queue is immediately invalidated after sending
the `-invalidate` message to it, and introduce an intentional retain cycle so
the bridge is only released together with all modules, when all the messages
have been dispatched.
Test Plan: Launch the UIExplorer, and reload it, like, a lot.
Summary:
Similarly to npm-installed react, this change makes changes to the packager so that it understands that it's been installed via Cocoapods and determines the project and asset roots properly (from the main application directory).
Closes https://github.com/facebook/react-native/pull/1568
Github Author: Jarek Potiuk <jarek@potiuk.com>
Test Plan: Imported from GitHub, without a `Test Plan:` line.
Summary:
This PR adds support for UIImagePickerController to allow selecting a photo / video from the users camera roll.
Example:
Selecting something from camera roll
```
ImagePickerIOS.openSelectDialog(<config>, <successCallback>, <cancelCallback>);
ImagePickerIOS.openSelectDialog({
showImages: true, // defaults to true
showVideos: false // defaults to false
}, function (data) {
console.info("Got a callback!");
console.info(data); // file URL as in assets-library://asset/asset.JPG?id=E2741A73-D185-44B6-A2E6-2D55F69CD088&ext=JPG
}, function() {
console.info("Cancelled");
});
```
Using camera
```
ImagePickerIOS.openCameraDialog(<config>, <successCallback>, <cancelCallback>);
ImagePickerIOS.openSelectDialog({
videoMode: false, // defaults to true, whether to record videos instead
}, function (data) {
console.info("Got
Closes https://github.com/facebook/react-native/pull/1620
Github Author: David Mohl <me@dave.cx>
Test Plan: Imported from GitHub, without a `Test Plan:` line.
Summary:
@public
UIExplrer's test targets were being built on normal builds as well, remove it.
Test Plan: Insert an `#error` on a test file, and run the project. It shouldn't fail anymore.
Summary:
Fixed#879 - I'm not too familiar with this aspect of XCode so a sanity check here would be great @tadeuzagallo 😉
Closes https://github.com/facebook/react-native/pull/1466
Github Author: Brent Vatne <brent.vatne@madriska.com>
Test Plan: Imported from GitHub, without a `Test Plan:` line.
Summary:
@public
corrected small typo in the 500 response from the packager server
Test Plan: add throw to promise function prior to error handler, run packager, cache a bundle with bundle extension URI, open /debug/packages, see clean 500 error
Summary:
@public
Fixes#773, #1055
The resolver was getting a bit unwieldy because a lot has changed since the initial writing (porting node-haste).
This also splits up a large complex file into the following:
* Makes use of classes: Module, AssetModule, Package, and AssetModule_DEPRECATED (`image!` modules)
* DependencyGraph is lazy for everything that isn't haste modules and packages (need to read ahead of time)
* Lazy makes it fast, easier to reason about, and easier to add new loaders
* Has a centralized filesystem wrapper: fast-fs (ffs)
* ffs is async and lazy for any read operation and sync for directory/file lookup which makes it fast
* we can easily drop in different adapters for ffs to be able to build up the tree: watchman, git ls-files, etc
* use es6 for classes and easier to read promise-based code
Follow up diffs will include:
* Using new types (Module, AssetModule etc) in the rest of the codebase (currently we convert to plain object which is a bit of a hack)
* using watchman to build up the fs
* some caching at the object creation level (we are recreating Modules and Packages many times, we can cache them)
* A plugin system for loaders (e.g. @tadeuzagallo wants to add a native module loader)
Test Plan:
* ./runJestTests.sh react-packager
* ./runJestTests.sh PackagerIntegration
* Export open source and run the e2e test
* reset cache
* ./fbrnios.sh run and click around
Summary:
@public
We cached based on url, which wasn't unique becuase some options would be defaulted. This was obvious when starting the server via fbrnios which tries to warmup the bundle.
And then when the device woke up it will send a request (that is identical in reality) but would miss the cache.
This changes the cache key into a JSON stringification of the options.
Test Plan:
* ./runJestTests.sh
* ./fbrnios.sh run
Summary:
@public
Add PerformanceLogger to keep track of JS download, initial script execution and
full TTI.
Test Plan:
The Native side currently calls `addTimespans` when it's finish initializing
with the six values (start and end for the three events), so I just checked it
with a `PerformanceLogger.logTimespans()` at the end of the function.
```
2015-06-18 16:47:19.096 [info][tid:com.facebook.React.JavaScript] "ScriptDownload: 48ms"
2015-06-18 16:47:19.096 [info][tid:com.facebook.React.JavaScript] "ScriptExecution: 106ms"
2015-06-18 16:47:19.096 [info][tid:com.facebook.React.JavaScript] "TTI: 293ms"
```
Summary:
Before this diff every time you reload in debug mode Chrome window
is actiavated. Looks like that behaviour is pretty annoying.
Fixes#689
@public
Test Plan:
```
$ ./packager/launchChromeDevTools.applescript 'https://www.facebook.com/'
```
First time it opens a new tab and activates Chrome, running this again does
not activate Chrome if the tab already exists.
Summary:
@public
I was using UIKeyCommand as a key in a dictionary, but it seems iOS wasn't treating identical commands as equal, so it was possible to register the same key command twice, resulting in the command triggering the action multiple times.
I've now created a container object for the key commands, and not relying on undocumented hashing behavior of UIKeyCommand for deduplication any more.
Test Plan: Reload bridge multiple times, then check that the number of registered keys in the command set inside RCTKeyCommands doesn't keep increasing.
Summary:
@public
There was an iVar being directly referenced from inside a block on RCTDevMenu
that was causing a retain cycle and the dev menu wasn't being released.
Test Plan: Put a break point on dealloc, it should be called now.
Tadeu Zagallo
Jing Chen
Amjad Masad
Daryl Rowland
Christopher Chedeau
Alex Kotliarskyi
Spencer Ahrens
Pieter De Baets
Christopher Dro
Artem Yarulin
Hannes Verlinde
Nick Lockwood
Eric Vicenti
Philipp von Weitershausen
Hedger Wang
Gabe Levi
Forbes Lindesay
Ruben Niculcea
Felix Oghină
Jarek Potiuk
David Mohl
Brent Vatne
Bill Fisher
Andrei Coman
Alex Akers
Andy Street
