With the development of software industry, the importance of software testing is becoming more and more important. In recent years, the intelligent industry has gradually risen, intelligent products and equipment, intelligent software testing, software automation testing more and more attention. Beijing Btestsky is tight with the software testing industry wave, in the intelligent application testing, automation testing to seek breakthrough.

Introduce the four stages of software automation test.

First stage: API automation.

1, making the API has the robustness, to normal data transmission and abnormal data transmission, server-side can respond correctly and return the correct response code.

2, for the integration of the API, make sure that each API is independent and can not be dependent on each other.

3. The correctness of the API provides a stable basis for the automatic verification of the front-end logic.

4, tools can be used: unittest, pytest (recommended).

Phase 2: automatically create test data.

The front end of some UI validation, needs some combination data, each update environment, version iteration, automation creates the required data.

This time needs to be used to ensure that each situation, including the border, is normal to the boundary. This data in each new environment needs to be verified, manually created too much waste time, through python reading excel in pre-designed. Automatically create a batch of data by using the API or the way that the database is written directly to the database. The way to write is chosen by specific business.

Phase 3: front-end operation automation.

The order of the second stage and the third stage is not too important, and it can be performed in the third stage.

Here, the front-end operation automation, the popular point is to the front-end control response to some automation verification, the foundation of the front-end test. If the text input, the button click on the response, the form is submitted after the normal display.

Depending on the demand document, the basic point of covering the requirements document is OK. There is no need for too many complex processes and operations.

Tools use appium.

Stage 4: user-real automation.

1.The user is based on the process of using the software, the real scene of the user operation, for the final finishing automation test.

If the user may be in use, stay on the page for 10 minutes, then lock the screen, then unlock it to see if the app is still alive.

If the user may be in the process of using, it is a program to return to the background. The specific operation here needs to know that different platforms are different from the definition of the program life cycle.

2.The front-end automation and interface automation.

We've been thinking about the front-end automation and interface automation.

The front-end automation focus is on the formation of the response, the data display (including length, decimal correct, etc.), the back-end focus is on the accuracy of data processing.

It is true that all the buttons in the front end of the Appium test are correct, whether or not a certain element is displayed. The effects of an action on other interface data are ignored. In fact, the front end and back-end automation focus is different, but the detection of data can be a double test. After this test, the data is more secure.

3. About data generation (preparation).

Data production (preparation) and testing in separate modules, mixed together, easy to interrupt the test code.

First data generate the data that you need and then run the test code.