Impact of Model Constructs on Test Generation
- malaya.panigrahi (Unlicensed)
- Suhas Deshpande
Conformiq offers different mechanisms to generate, import, or cover specific test data which can easily increase the computational complexity of a model
Combinatorial action (used with alternative values)
Combinatorial & exploratory data coverage in input actions
Specification of value lists in external spreadsheet files (like Excel)
Conformiq uses and at same time validates all test data specified as part of input action constraints during test generation. By default, every individual data element or generated combination is a coverage target. The more data elements in the model (e.g., Excel rows, data combinations), the longer the test generation.
Recommendation: Use these constructs with care and check your project for Livecheck warnings regarding computational complexity. |
Recommendation: Please check if adding data into the model adds any value to the model (for e.g., as variables to control the flows, to create optimized set of data). If not, the data can be added to the test during execution (either manually or automatically through a tool, with appropriate change to framework). |
Recommendation: Even though Creator’s capability can be used for various end goals (for e.g., as test data generator or pre-requisite setup through automatic flow execution via Creator’s integration with such tools), please analyze the impact on the core purpose and Creator’s core capability – modelling and/for automatic and optimized test case generation, with data. |
See section Model Combinatorial Data Only After Completing Modelling of Targeted Functionality and section Limit Use of Combinatorial Data to Avoid Test Case Explosion
See also “Combinatorial Testing with Conformiq Creator ” for further read.
Expect longer test generation times for Excel files with more than 100 rows.
Recommendation: Please check and, if possible, use Data Tables modelling construct available within Creator. |
It is important to understand that
Combinations that are invalid or not necessary to be covered can be “Blocked”. Review test target settings for combinations.
Conformiq can generate data from pure logic, i.e., it is not required to specify test data in a model to get tests with test data
More test data does not imply better tests or coverage – it just means more tests.
Find more advice in section Test Generation of this document.
Related content
Copyright © 2023 Conformiq and its subsidiaries. All rights reserved.