ReactiveX

We will understand what is Reactive Streams and how helpful it is to deal with Async Task

Transcript

1. ReactiveX solves complex problems easy by thinking differently Srikrishna S, Arch Team
2. for a long time nobody cared me. Now you are also here.
3. If silicon valley is moving forward, then we have to move forward.
4. else you are so called obsolete / outdated
5. Jokes Apart. Why nobody cared ? - Rx loves functional style - Lot of operators - Mind shift in thinking. Async way - Not easy to explain
6. What is ReactiveX ? lets take look at official definition
9. Lets analyze keywords we found there: Reactive Asynchronous Functional Streams Iterator Observer Pattern Observable
10. Asynchronous Programing - Subscribing to a JMS queue - Application life cycle hooks. - Event listeners $.get(“localhost:8080", (response)=>{}); $("button").click(function(){log("Click!"}); simplest program you could ever written in Async.
11. But wait. Async Programing is not limited to Events Async
12. Async is non blocking. Data arrive over time UI Events, JMS, PUB/SUB Async
13. Which means It’s a pain to chain Async operations Async
14. JS callback hell Sequential Async
15. Parallel Async Race Conditions (auto complete) Memory Leaks (event listener) Complex State Machines (auto complete) Uncaught Async Errors (callback hell) Async
16. Each platform implements Async as different Async
17. ReactiveX is here to solve complex Async problems easily Async
18. But Why do you need Async programing? Use cases Async
19. Big Companies are pushing it beyond the limits Async
20. Sometimes its really necessary Modern Web Apps (complex js app) Async
21. Functional Programing - list.forEach(System.out::println) $("button").click(()=>alert("Yes! you did it.")); If you ever written a program like this, $.get(“localhost:8080", (response)=>{});
22. Streams are flow of data from src to dst I/O, Events, Collections
23. Let’s talk about iterators In computer programming, an iterator is an object that enables a programmer to traverse a container
24. Facebook: Get all the friends of a given user who is working in the same company and sort by names ? Lets understand by a problem
25. public List<User> getFriendsByWork(String userId){ List<User> friends = new LinkedList<>(); List<String> friendIds = userService.getFriends(userId); for(int ii = 0;ii<friendIds.size();ii++){ String friendId = friendIds.get(ii); User user = userService.getUser(friendId); if(user.company == "sella"){ friends.add(user); } } Collections.sort(friends); return friends; }
26. Set gives some serious performance. But remember it is not ordered.
27. This is where iterator comes into play. It allows you iterate through container
28. interface Iterator<E> { boolean hasNext(); E next(); }
29. interface Iterator<E> { boolean hasNext(); E next(); remove(); // Java }
30. public List<User> getFriendsByWork(String userId){ Set<User> friends = new HashSet<>(); Set<String> friendIds = userService.getFriends(userId); Iterator<String> iterator = friendIds.iterator(); while(iterator.hasNext()){ String id = iterator.next(); User user = userService.getUser(friendId); if(user.company == "sella"){ friends.add(user); } } List<User> temp = toList(friends); Collections.sort(temp); return temp; } {
31. public List<User> getFriendsByWork(String userId){ Set<User> friends = new HashSet<>(); Set<String> friendIds = userService.getFriends(userId); for(String friendId: friendIds){ User user = userService.getUser(friendId); if(user.company == "sella"){ friends.add(user); } } List<User> temp = toList(friends); Collections.sort(temp); return temp; } {
32. These are so called External Iterators. You control them.
33. Internal iterator? that simplifies even more common iteration problems
34. List<User> users; user.forEach(System.out::println); void forEach(Consumer<? super T> action) { Objects.requireNonNull(action); for (T t : this) { action.accept(t); } } public interface Consumer<T> { void accept(T t); }
35. Map > [1, 2, 3].map(x => x + 1) > [2, 3, 4] >
36. Filter > [1, 2, 3].filter(x => x > 1) > [2, 3] >
37. Sort > [3, 2, 1].sort() > [1, 2, 3] >
38. public List<User> getFriendsByWork(String userId){ Set<User> friends = new HashSet<>(); Set<String> friendIds = userService.getFriends(userId); Iterator<String> iterator = friendIds.iterator(); while(iterator.hasNext()){ String id = iterator.next(); User user = userService.getUser(friendId); if(user.company == "sella"){ friends.add(user); } } List<User> temp = toList(friends); return temp; } Before Imperative Programing
39. public List<User> getFriendsByWork(String userId){ return userService.getFriends(userId) .stream() .map((friendId)->userService.getUser(friendId)) .filter((user)->user.company == "sella") .sorted() .collect(Collectors.toList()) } After Declarative Programing
40. Stream API is nothing more then a Internal Iteration operators to solve iteration problems in functional style
41. This is why functional style is loved by Streams and ReactiveX API
42. This is where Kotlin shines over Java
43. Observer Pattern The observer pattern is a software design pattern in which an object, called the subject, maintains a list of its dependents, called observers, and notifies them automatically of any state changes, usually by calling one of their methods. It is mainly used to implement distributed event handling systems, in "event driven" software. - wikipedia
44. Observer Pattern Ex: - Subscribing to a JMS queue in Java - Application life cycle hooks. - Event listeners $.get(“localhost:8080", (response)=>{}); $("button").click(function(){log("Click!"});
45. Iterator Observer Arrives On Time or Static Arrives Over time Collection or Container Events Pull Push next(), hasNext(), throw error(); Only next() Diff: Iterator vs. Observer Finite Infinite
46. But what is similarities between these two
47. Average Programmer won’t notice because we followed this book.
48. They designed… - Iterators pattern for collections - Observer pattern for events
49. - collection that arrives over time ? - stream of data that can be finite? - has onComplete() notification ? - has onError() notification? - can have operators like Stream API What if we consider Observer as…
50. That is where an Observable Born var source$ = Observable.create( (observer) => { observer.next(data); observer.complete(); observer.error(error); }); source$ .map().filter().sort() .subscribe(onNext, onError, onComplete);
51. So Many Push/Async APIs can be unified DOM Events Websockets Pub/Sub Database API Rest Calls I/O Streams
52. Db .table("user") .filter(row -> row .get(“company”) .eq("sella")) .orderBy("name") .run(conn); RethinkDB
53. No surprise if you are wondering how this works. Lets see some more operators before jumping into it
54. interval() Rx.Observable .interval(1000) .subscribe(console.log); // output: 0,1,2,3,4,5....
55. fromEvent() Rx.Observable .fromEvent(document, 'click') .subscribe(console.log); // output: [event], [event], [event]
56. debounceTime() https://codepen.io/dcorb/pen/KVxGqN
57. distinctUntilChanged() Rx.Observable .from([1,1,3,1,1]) .distinctUntilChanged() .subscribe(console.log); // output: 1,3,1
58. Lets get hands dirty with auto complete
59. We love to hear you Feel free to say hi! If you have anything to ask srikrishna.s@sella.it