Implemented "First" and "FirstOrDefault" operations #357
Conversation
|
RxJava-pull-requests #248 SUCCESS |
| * source Observable completes without emitting a single item. | ||
| * @see <a href="http://msdn.microsoft.com/en-us/library/hh229177%28v=vs.103%29.aspx">MSDN: Observable.First</a> | ||
| */ | ||
| public Observable<T> first() { |
There was a problem hiding this comment.
The naming convention here is different than last and takeLast so wondering if we should change to match.
The last operator is blocking so on BlockingObservable. This first operator seems to be similar to single and blocking last.
Thus I wonder if we should have takeFirst instead of first? This is different than Rx.Net but would clean up the same difference in convention that it has.
There was a problem hiding this comment.
I'd expect the names to be first and last whether blocking or not. Also, should BlockingObservable really extend Observable? Seems like a potential source of confusion.
There was a problem hiding this comment.
It doesn't extend from Observable any longer.
There was a problem hiding this comment.
Cool. I'm behind the times :)
There was a problem hiding this comment.
So takeFirst/takeLast/ or first/last ... or both aliased?
Until we split off BlockingObservable last and takeLast meant different things, we could now have last exist on both but with different return types.
|
In Scala, I'm okay with both naming schemes, with a slight preference for leaving the I was also playing around with |
|
I need to look at what Java 8 is calling these things, as that is the more important naming convention for the core lib to match. |
|
Java 8 uses |
…into operator-first-merge Conflicts: rxjava-core/src/test/java/rx/ObservableTests.java This merges pull request ReactiveX#357 Also aliased first with takeFirst to match takeLast.
…into operator-first-merge Conflicts: rxjava-core/src/test/java/rx/ObservableTests.java This merges pull request ReactiveX#357 Also aliased first with takeFirst to match takeLast.
…into operator-first-merge Conflicts: rxjava-core/src/test/java/rx/ObservableTests.java This merges pull request ReactiveX/RxJava#357 Also aliased first with takeFirst to match takeLast.
…into operator-first-merge Conflicts: rxjava-core/src/test/java/rx/ObservableTests.java This merges pull request ReactiveX/RxJava#357 Also aliased first with takeFirst to match takeLast.
…bscriber is is not permitted to subscribe, the subscription must be canceled and the ResilienceBaseSubscriber must call onSubscribe on the target Subscriber with an EmptySubscription followed by a call to onError with the supplied error. (ReactiveX#452) Issue ReactiveX#357: Fixed a bug in ResilienceBaseSubscriber. If a subscriber is is not permitted to subscribe, the subscription must be canceled and the ResilienceBaseSubscriber must call onSubscribe on the target Subscriber with an EmptySubscription followed by a call to onError with the supplied error. The bug was that the ResilienceBaseSubscriber cancelled it's own subscription as well.
This PR builds upon the
skipWhilePR (#355) which makes implementing the twofirstvariants extremely easy.I changed
firstOrDefaultslightly from what Rx.NET does by explicitly requiring a default value as parameter of the function call. I don't know what default value I should return otherwise, exceptnull(and imho this would be pretty useless).This PR addresses issue #44.