Espresso is a testing framework for Android which automatically synchronizes your test actions with the UI of your app. It also lets the test wait until all the background activities have finished.
The pCloudy utility uploads your Application APK, test APK and other APK files to pCloudy. It will then acquire a device to run your test on, execute those tests and report back the status and test report.
In pCloudy you don’t have to set up the environment. Just follow these simple steps.
After the test cases are run you can see the detailed reports. Once automation ends, the device under test will be released automatically so that you or other users can use it later.
There are two ways you can execute espresso test cases. Here are the steps for both
1. pCloudy UI Dashboard a. In pCloudy UI there is a section called Automation where you will find multiple automation frameworks integrated to the platform. b.Select the Espresso automation and to fill in the required details.
c. pCloudy Support two TestRunners.
i. AndroidJUnitRunner (android.support.test.runner.AndroidJUnitRunner)
ii. AndroidXJUnitRunner (androidx.test.runner.AndroidJUnitRunner)
d. Select your Application and Test Application.
e. Select the required devices.
f. Click on Next button and Schedule the automation by clicking on the Schedule button.
g. You will get a dialog box. Click on Ok.
h. You will get the active session details on the Dashboard and you will also be able to see the Live view.
i. Liveview will help you to see the live execution on your test cases. Click on the Liveview icon
2.pCloudy Utility
a. Download the espresso starter pack from here and Unzip it b. You will find two files,
i. pCloudy_Espresso.jar
ii.Config.properties c. Place the config.properties file and pCloudy_Espresso.js file into the same directory. d. Update the config.properties file as guided in the comments in the file. e. In your workspace, run the command by typing the java -jar pCloudy_Espresso.jar
f. This will upload the required files which you have mentioned inside the config.properties files.
g. You will get the status of the Automation in pCloudy Dashboard ActiveSession.
h. Even on the pCloudy dashboard, you can use the Liveview by which you can see the running test cases.
i. Once the test execution is over, You will get a URL of the test execution report on the terminal.
j. You can open the report from the dashboard Report section also. In the report, all the test details will be there as shown in the screenshots below.
k. You will also get individual device reports. Inside the test case level details, you will find performance details, test case logs, device logs, and video files.
Studies have indicated that after installing a new app, only a quarter of the users return to the app on the second day. The retention rate is eminently low and the majority of the users uninstall the app after the first use. You might have guessed the reason for this. If you think the reason is scanty testing then you are correct. A lot of times testers show little interest when testing is repetitive. Although regression testing is crucial to the success of the project, it is often overlooked to some extent. So let’s understand the concept of regression testing.
What is Regression Testing
In simple terms, regression testing can be defined as retesting a computer program after some changes are made to it to ensure that the changes do not adversely affect the existing code. Regression testing increases the chance of detecting bugs caused by changes to the application. It can help catch defects early and thus reduce the cost to resolve them.
Regression testing ensures the proper functioning of the software so that the best version of the product is released to the market. Although, creating and maintaining a near-infinite set of regression tests is not feasible at all. This is why enterprises are focusing on automating most of the regression tests to save time and effort.
Types of Regression Testing
There are different types of regression testing for different phases of testing. Let’s understand the types of regression testing.
Unit Testing: It is done when coding changes are made for a single unit and a tester re-runs all previously-passed unit tests. Often automated unit tests are built into the code to make unit testing efficient.
Progressive Testing: This type of testing works effectively when there are changes done in the software/application specifications as well as new test cases are designed.
Selective Testing: In this to cut down the retesting cost and effort, testers use a subset of the current test cases. A test unit must be rerun when there are changes in the program entities it covers.
Retest-All Testing: In this, we reuse all test cases even where the changes have not been made. It shouldn’t be done where there is a small change in the application as it is very time-consuming.
Complete Testing: Complete testing is effective when multiple changes have been done in the existing code. It is done to identify unexpected bugs and the final system can be made available to the user once complete testing is done.
How to Derive a Regression Testing Plan
Whenever there is a change in the app or a new version is released the developer carries out these tests as a part of the regression testing process.
First, the developer executes unit-level regression tests to validate the code that they have modified along with any new test that is created to cover any new functionality. Then the changed code is merged and integrated to create a new build of AUT. After that smoke tests are performed to assure that the build that we have created in the previous step is good before any additional testing is performed.
These tests can be executed automatically by continuous integration services such as Jenkins. Once we get the assurance that the build is perfect then we do the sanity testing to confirm that the functionality works as expected and known defects are resolved before conducting integration testing.
Integration testing is performed to verify that the units of the apps interact with each other and with backend services such as databases. The next step is to schedule regression tests depending on the size and the scope of the code either a partial or a full regression test may take place.
Defects are reported back to the development team and many require additional runs of regression testing to confirm the resolution. So based on the test reports you can analyze and figure out what test cases can be added for the next checking process as well and you create your report on that.
Regression Testing Challenges
Automation is always a smart move but there will be some challenges that would need your attention. Let’s have a look at the challenges that you might face in automation testing.
Cost is High
In terms of business expenses, the company will have to invest a lot of time and money in repetitive testing. Also, there’s no specific return on investment that is attached to regression testing. It is a complicated process and to explain the importance of regression testing in terms of business benefits to the management is really difficult. But this has to be done to draw the budget for the testing even though it is expensive from a management point of view.
Unstructured Information
If new testers are recruited during the ongoing software testing project, either to speed up the process or to release some of the tasks. The existing testers in the organization might not have time to give complete details about the functionalities and features of the app, existing bugs, delivery time, etc. The new testers might have to initiate testing with inadequate data about the ongoing testing project. Challenges like these need an appropriate test strategy for efficient collaboration.
Time Restrictions
The focus of enterprises is on developing quality apps and delivering them faster. This is why regression testing is always time-bound. Testers always get a deadline to complete the process of regression testing. To keep pace with time, The testers might have to take chances to give more importance to the critical regression tests and skip a few tests. It is a serious challenge as it is critical to decide which tests can be skipped.
Maintenance and Optimization
A major challenge is to maintain and optimize the existing regression test suites. The existing regression test suites might need modification like adding, removing or editing the existing test cases, whenever there are new changes. All this need to be done before the deadline fixed for the regression testing. Therefore, optimization and maintenance is another challenge that needs attention.
Best Practices to Perform Regression Testing
Now that we are aware of the challenges in regression testing. Let’s have a look at the key changes we can do to enhance it.
Focus on the Frequently Used Paths
Frequently used paths also make the most frequent use cases for your app. They must include the most popular features and the basic functionality of your app. You should be aware of the core group of users and their usual interaction with the app features. The regression test cases must ensure that this core functionality is working as expected.
Update the Regression Pack Regularly
A regression pack is a collection of test cases that are performed when a new app version is released or if there are any new updates. The tests in a regression pack are created keeping in mind the app’s older version requirement specifications. Ad-hoc tests should be included in the regression pack to keep it up to date. You wouldn’t want to waste your time on tests to validate an old feature that is not included in the latest version of the app.
Create an Entry/Exit Criteria
The entry/exit criteria that we usually follow in the software development lifecycle can be implemented in regression testing as well.
As entry criteria, a fixed set of conditions need to be met like reviewing bugs and performing regression testing to fix those bugs. Exit criteria can be applied with a fixed set of conditions like making sure that all the tests are executed and no bugs remain unfixed.
Automate Regression Tests
Automating regression tests makes the process much more efficient and saves time and effort. Repetitive testing can make testers lethargic and testers might lose motivation. It also affects the quality of their work as the testing is repetitive. Automated regression testing will free up resources to work on the more complicated tests. When we use automation, the test cases along with the software can also be reused which increases the return on investment.
In a Nutshell
Change is hardest at the beginning, messiest in the middle and best at the end. This quote by Robin Sharma suits the importance of regression testing in delivering a smooth functioning app. As mentioned earlier in this blog, different types of regression testing can be performed at all stages of the testing life cycle. And there are ways to overcome the challenges in regression testing too.
One option is to use a cloud-based app testing platform which will enhance the effectiveness of regression testing by many folds. There will be hundreds of devices to perform testing in parallel and features which complement regression testing automation.
In the past couple of years, there has been a tremendous increase in voice-enabled apps in the market. This is because business is looking to make its app more user-friendly and easy to use. If the same function can be performed using just your voice then why would you waste your time entering text? Voice recognition is being used by many apps from all sectors like food delivery, eCommerce, gaming, weather, navigation, social media, etc.
The Rise of Voice-Enabled Apps and Intelligent Virtual Assistant
Voice-enabled apps are quite common these days as enterprises are trying to make it easier for users to control the app. Google was the first widely used app that utilized voice reorganization to make it simple for users to look for anything on the search engine. But when it comes to Intelligent Virtual Assistant (IVA), Apple does lead the way with Siri.
The idea of IVAs is not new and we have seen it in many Sci-Fi movies too. Take the example of Weebo from the movie Flubber. It was a flying robot assistant who recognized voice commands and helps perform daily tasks. It’s quite possible that we will see something like that soon in the market leveraging AI and drone technology.
Graphical User Interface Vs Voice User Interface
Apart from GUI, a new trend is shaping up and it’s called VUI. In the world of voice apps, the primary interaction between human and machine is done through the Voice User Interface (VUI). It is a natural form of interaction where the user can speak to the device and the device will reply or perform the task that it is asked to do. For instance, If you ask the Google assistant “How is the weather in Palo Alto?” the app will convert the analog command into binary and using AI it will reply with an audio output. Here the testers need to make sure that the app is generating the correct reply to the questions. Exploratory testing plays a key role in identifying the bugs in audio-based testing.
When it comes to GUI, it is designed in a way to make it easier for the user to navigate through the business process. The graphical interface has controls such as checkboxes, dropdown lists, radio buttons, etc, through which all the features are controlled. Therefore the tester needs to test if all the control elements are working fine in all the pages.
The fundamental difference between testing the GUI and the VUI is the input type. There is limited data input in GUI testing but in VUI testing, it can create an indefinite number of test cases. When it comes to voice apps, pairwise testing can help to get better test coverage with fewer test cases.
The Easiest way to do Audio Interaction Testing
The best way to test your app for audio functions is by using pCloudy. To perform audio testing, users need to upload their app on the platform and install it on any device of their choice. Once the device is booked and the app is installed on the device, click on the “More” option at the bottom left corner of the screen. Then you will get the audio in and audio out options in the device interaction page to do audio testing. Users can also use their voice as an input for testing using the “Audio In” feature.
This feature can be used for audio out testing as well and it is available only for devices on which audio icon is present. To perform audio output testing, just book a device on the pCloudy platform, click on the “Audio Out” icon and play any audio/video file to get the sound output on your system.
The audio testing feature is not available on any other cloud-based testing platform and this gives a big advantage to pCloudy. Testers will be able to use this feature to test their apps with voice commands. They will also be able to check if the app is responding with sound output for a particular function.
Conclusion
Looking at the popularity of voice apps, It is estimated that in the US, 111.8 million people will use a voice assistant at least once in a month in 2019. This makes up about 39.4% of internet users and 33.8% of the total population. Voice apps and assistants are gaining popularity due to its ease of use.
A lot of organizations are working on using AI and ML to make customer support automated. This means that there will be no more humans answering the phones or replying to chats. A basic version is already in practice which is the IVR. In the near future, there will be smart voice support applications that will listen to the customers and resolve their issues without human interference. This will also reduce the waiting time to talk to a support executive from the organizations. Again a great example of efficiency and speed.
In our previous chapter on Android, we learned about UI Automator Viewer, Which is available on Android SDK, to get the properties of the application object. In the case of iOS, Appium itself provides an Inspector which helps users to locate those elements in the application.
First, open the simulator by clicking on the dock option.
Now in the Device/Simulators window, select the simulator. Open the Appium Desktop and keep the simulator side by side.
Once the inspector is started, select any of the objects on the screen. It will show you the complete hierarchy and properties of that object.
At the top of the window, you can see the Record button which is used to record all the actions taken and record the script.
To select any object, click on the Select Element button and then you can use Tap button to click on an object, Send Keys to enter text and clear to undo the action.
As soon as you perform an action on an object, it is recorded in the form of a script.
Once you are done with the recording you can copy the script and paste in eclipse editor.
In the next blog, we will learn how to write the first appium script for iOS.
