Software testing method and platform

Abstract

The invention provides a software testing method and platform. The software testing method comprises the following steps: a. injecting a test case through a test case file, and parsing and caching thetest case; b. establishing a static association between the input data in the test case and the input interface of the software component to be tested; c. packaging the software component under testby using a predefined software component under test container; d. Controlling the periodic execution of the test; e. The input and output interfaces of the tested software components are statically associated with the test result output buffer through a data pointer, and the test result is stored in the test result output buffer; f. reading a test result from that output buff, and recording and saving the test result. The software testing method and platform have the characteristics of being able to cross-software reuse, being able to correctly read a variety of different types of data in testcases, being able to ensure event synchronization between test input and output data, and the like.

Description

technical field [0001] The invention mainly relates to the field of software testing, in particular to a software testing method and a platform thereof. Background technique [0002] Software testing platform usually consists of software testing platform framework, software components under test (such as a closed-loop system, the controlled object model is regarded as a software component for scheduling but not as a tested piece, which is not clearly distinguished in the present invention), Test input and test result interpretation are composed of four parts, such as figure 1 shown. [0003] The software test platform solution architecture usually consists of 1) an input module for test case injection, parsing and caching, and 2) an interface association module for statically associating the input data (buffer) in the use case file with the external input interface of the software component , 3) test execution control (including the access of software components and contro...

AI Technical Summary

Problems solved by technology

[0006] 1. Existing software testing platforms are all aimed at a specific software and build a dedicated testing platform. The testing platform is tightly coupled with the tested components and cannot be separated, lacking a general solution

Therefore, each software test platform must be built separately, and it is difficult to realize the transplantation and reuse of functional logic such as cycle scheduling and timing control between different software test platforms, resulting in repeated construction and increasing the overall cost of software development

[0007] 2. Regarding the injection of test cases, there are mainly two types of solutions in the existing software testing platforms: 1) One kind of solution is to receive and process different types of input data through the controls of the graphical user interface, and each control is directly connected to the input interface of the software component. Static binding, because the data input of most controls usually overwrites the old value with the new value, it is impossible to save multiple sets of input data at the same time, it is difficult to ensure that the input data of the control is refreshed on time, and it cannot support the precise control of the test input beat

2) Another solution is to organize all external controllable inputs and their input tempo control information into use case input files in digital form, and provide them to the software under test uniformly, which solves the problem of precise control of the former test input tempo; about Analysis of use case files. Existing software testing platforms require all use case data to adopt the same data type. When the interface is associated with the software under test, it is matched through mandatory type conversion, which loses the accuracy of some test input data and affects the accuracy of test results. accuracy

[0008] 3. In the closed-loop control system, part of the external input comes from the complex controlled object (model) instead of the test case, the test output is not uniquely determined by the use case input, and the correctness of the test result cannot be automatically judged only based on the use case input

However, since the input and output of test cases are not associated in the test results, it is difficult to maintain the synchronization relationship between input and output, which affects the efficiency and accuracy of test result interpretation

Images