Reactive Programming: The Steps To Integrate It Into Your Application
Reactive programming is an amalgamation of observables, observers, and schedulers. Learn more about implementing reactive programming in this post.
Join the DZone community and get the full member experience.
Join For FreeAccording to Wikipedia, reactive programming (Rx) "is a declarative programming approach in computers that deals with data streams and change propagation. This paradigm allows users to easily specify static (e.g., arrays) or dynamic (e.g., event emitters) data streams, as well as indicate that an inferred dependence exists inside the related execution model, allowing for automatic propagation of the modified data flow." The word was first used in the IT industry in the 1960s, and since then, much has been written about it.
Unfortunately, as is so frequently the case, the new notion rapidly sparked a slew of misinterpretations that stick to this day. The 2014 Reactive Manifesto, which introduced "reactive systems" and their four sacred principles further twisted things up. In Rx programming, data flows are produced by one component, and the underlying structure supplied by the Rx libraries propagates those changes to other components that have been registered to accept those data changes.
Reactive programming may be described in the following manner to grasp it in simple terms.
About Reactive Programming
Reactive programming is an amalgamation of observables, observers, and schedulers. We'll go through each in-depth one after the other.
Observable
Simply put, observables are just data streams. The data that may be transmitted through one thread to another is stored in an observable. They essentially send data regularly or only once over their life cycle, depending on their setups.
Several operators can assist the observer in emitting specific data in response to certain events, but we'll look at them in more detail in the next sections. For the time being, consider observables to be suppliers. They process the data and send it to the rest of the system.
Observers
They are, in essence, the consumers. Observers consume the data stream emitted by the observable, which it has previously registered to.
Schedulers
This is thread management, in a nutshell. Because we're talking about asynchronous programming, thread management becomes a no-brainer. As a result, schedulers instruct observables and observers which threads to use.
When Is It Appropriate To Employ Reactive Programming?
We now understand why reactive programming was created and what it is, but when do we use it? When dealing with asynchronous data streams, it's a popular choice. Even a minor variation in the use case could become a determining factor in our decision.
Below are some examples of the use of reactive programming in the real world:
Using Reactive Programming to Develop Mobile Applications
Because mobile devices are not powerful enough to handle the heavy lifting, we frequently need to update the User Interface on the main thread from the background thread amidst the activity or after the task. As a result, we execute heavy work and complicated computations on our servers. As a result, we require asynchronous work for network activities, which is where reactive programming comes into play.
Use of Reactive Programming Alongside RxJava in the Netflix API
To successfully decrease network regular interactions, reactive programming in the Netflix API using RxJava server-side concurrency is required. Since each network request from a device automatically operates in parallel with other network requests, a single "heavy" client request may not be substantially better than several "light" ones if the server does not support concurrent execution. Even when network delay is taken into consideration, if the server-side processing of a compressed "heavy" request does not attain the same degree of simultaneous execution, it may be slower than numerous "light" requests.
Visit Netflix's highly recommended article for a more complete explanation.
External Call Service
The underlying protocol is blocking and synchronous calls to outside services because many backend services nowadays are RESTful (i.e., they use HTTP). Perhaps not apparent FRP terrain, but it's a pretty rich ground because the development of such services frequently entails contacting other services, and then calling even additional services based on the outcomes of the first calls. With so much IO going on, waiting for one call to finish before making the next request might cause your poor customer to give up in disgust before you could put together a response.
Therefore, optimizing external service calls, particularly complicated orchestrations of dependencies across calls, is a smart idea. FRP promises "composability" of the logic underlying such activities, making it easier for the calling service's developer to write.
Highly Concurrent Message Consumers
A consumer of a high number of concurrent message processing is a typical corporate use case, especially when it is highly synchronized. Reactive frameworks like to brag about how many messages per second they can process on the JVM by measuring microbenchmarks.
The statistics are fantastic (tens of millions of messages per second are not difficult to produce), but they may be a little skewed: you wouldn't be as thrilled if they stated they were benchmarking a basic "for" loop. However, we should not dismiss such effort too quickly, and it's clear that when it comes to performance, all contributions should be gladly received.
When it comes to particular sorts of high-load or multi-user applications, going reactive is an elegant solution. Games, social media, and chat rooms are apps for audio and video (mostly in streaming), but it should be remembered that the use of reactive technique when there is no true need for it will cause havoc on an application by adding unneeded complexity.
Integrating Rx
To integrate Rx into your application, follow these three easy steps.
Step 1: Make a Data-Emitting Observable
The database is observable that emits data: in this case, it emits the strings. just()
is a function, which simply emits one by one the data supplied in the argument.
Step 2: Make a Data-Consuming Observer
The observer in the preceding code snippet consumes the data generated by the database observable. It takes in the data and processes it, as well as handling errors.
Step 3: Regulate Concurrency
In the final step, we define our concurrency schedulers. Database observable is told to execute in the background thread via subscribeOn(Schedulers.newThread())
. The observer is told to execute on the main thread via observeOn(AndroidSchedulers.mainThread())
. This is the most basic reactive programming code.
To Conclude
We've taken an in-depth look at the reactive pattern trends in this post, putting them in perspective in the modern workplace. For the Java Virtual Machine, there are several reactive libraries or frameworks under active development. They have a lot of comparable capabilities but owing to reactive programming, they are becoming increasingly compatible.
Opinions expressed by DZone contributors are their own.
Comments