add failing test for filterBy #12538
Closed
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Prior to this change, a setter that exists on the prototype of the object was not detected and was effectively clobbered. This change updates `handleMandatorySetter` to check both the objects own descriptors and its prototypes descriptors. Also removed the guard for `Object.getOwnPropertyDescriptor` since all supported platforms are ES5.
[DOC beta] Add link to `Ember.Component#defaultLayout` deprecation guide.
Contextual components show a message when passed null or unknown component paths.
QUnit 1.19 was released and changed the way the npm package exports its value (it no longer sets up `global.QUnit` and all the assertion methods automatically). This updates to a better pattern of importing `QUnit` and using the `assert` argument for assertions.
This flag is no longer needed now that this feature is out of beta.
Fix node tests.
…ponents [BUGFIX] Fix contextual components handling invalid component paths
allow @Protected in JSCS
Remove `ember-registry-container-reform` feature flag.
When specifying `postionalParams` like this:
```javascript
App.FooBarComponent = Ember.Component.extend();
App.FooBarComponent.reopenClass({
positionalParams: 'allTheThings'
});
```
You should be able to avoid using positional params by specifying
`allTheThings` as a hash argument:
```hbs
{{foo-bar allTheThings=blah}}
```
Unfortunately, we were triggering an assertion that you used a positional
param that conflicted with a hash param. This change fixes that
assertion by avoiding doing any work when no params are specified.
…al-params [BUGFIX release] Ensure rest arg positionalParams can be omitted.
the build updates per feedback
…erating [BUGFIX Beta] Move warn api doc to a place where it gets picked up by the build
Due to the nature of `positionalParams`, it is trivial to supply
`params` as a hash argument to `{{link-to}}` instead of ordered args
(this allows truly dynamic link's). However, if `{{link-to params=foo}}`
is used `params` will be a `MutableCell` and the current code throws an
error in that circumstance.
*tldr;* `this.attrs` should be avoided. It is quite a 🧌 in
non-`GlimmerComponent`'s.
Cleanup meta
Due to values in `this.attrs` being either a `MutableCell` or the actual value, many flags in `link-to` would not function properly when provided a bound property.
remove extra comma
When compute is called it is expected that object is an actual Object, sometimes the object is a Bool or String and yet we continue down the stack look values up on it even though the result will be undefined. This change bails out early if `object` is a Bool or String.
[BUGFIX beta] Allow usage of bound properties in {{link-to}}.
[PERF] Bail on get if type does not have accessors
We check to see if a function contains `_super` / `.call` / `.apply` before attempting to super wrap a given method (this saves quite a bit of extra super wrapping for functions that do not need it). Unfortunately, a number of platforms that we support do not support calling `func.toString()` so we attempt to detect this and fall back to always super wrap everything mode. We have roughly this code (from [here](https://github.com/emberjs/ember.js/blob/b4718218dbe5ffe7736c485a594248b20977c621/packages/ember-metal/lib/utils.js#L257-L271)) to detect if we can call `toString()` and get the original source: ```javascript let sourceAvailable = (function() { return this; }).toString().indexOf('return this;') > -1; ``` This works perfectly for development builds, but unfortunately not when minified. Take a look at the minified source and you will see why: ```javascript var e=function(){return this}.toString().indexOf("return this;")>-1; ``` Note that minifier has stripped the trailing semicolon from the function body we are attempting to check against, but sadly our `indexOf` check still contains the semicolon! **tldr;** We super wrap every function in uglified builds. --- This commit removes the trailing semicolon from the source check, and adds a test that should hopefully allow us to detect if this is accidentally reintroduced.
…thing [BUGFIX release] Fix uglification introduced bug with super wrapping.
A similar feature is being proposed for ES. The syntax is still up in the air, and it will be async. But we can make use of this today
…extendable factories. The fake container is required to provide access to the `container` during instantiation of objects by non-extendable factories. The fake container provides access to the following methods: * `lookup` * `lookupFactory` Deprecation warnings are provided that direct the user to call `getOwner` followed by the equivalent `ContainerProxy` methods (`lookup` and `_lookupFactory`) on the owner.
native-array -> mutable-array -> enumerable -> native-array To construct native-array enumerable must have been evaluated. By ensuring enumerable EmberA is lazy, all this works without leaking the cycle out.
* move Ember.computed.collect back to runtime (its test are already there)
[FEATURE ember-container-inject-owner] Inject fake container with deprecations.
Remove internal references to the Ember Global
[PERF] Fix Mixin shape issues Fix emberjs#12606 Make Mixin monomorphic and use debugSeal in development builds to flag issues.
…ple, r=rwjblue [DOC release] Add example and clear up wording for CP get/set This PR makes the following changes to `Ember.computed` API documentation: - Changes the `fullName` getter function to match [the example in the latest guides](http://guides.emberjs.com/v2.1.0/object-model/computed-properties/). This involved changing to a template string from a string concatenation. - Switches class property default values to be defined in `init()` instead ([suggested](emberjs#12619 (comment))) - Minor changes to wording and sentence structure, mostly adding articles. - Adds an example that uses the hash format with `get` and `set` - Removes confusing/incorrect phrase: > The `get` function should accept two parameters, `key` and `value`. My one concern is that the two examples are very similar except for a few lines. Would it be more helpful if the second example only included the computed property (rather than the whole class and usage example)?
|
☔ The latest upstream changes (presumably 0c3b904) made this pull request unmergeable. Please resolve the merge conflicts. |
|
93 commits, is a bit big for us to review. Can you clean this up? |
|
@stefanpenner - Likely my fault. Since this targeted release branch, when I rolled beta to release a whole bunch more commits dumped in there messing up this diff. |
csantero
force-pushed
the
filterBy-each
branch
from
7c14e31
to
4a098e8
Compare
|
@rwjblue I'll submit a new PR targeting master |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
DO NOT MERGE
I believe this test demonstrates behavior that used to work. The first
deepEqualassertion passes but the second one fails.This is related to #12475.