A report is the result of the comparison of two states with one or multiple steps and contains all differences between these two states. It is a static artifact and can be viewed without the corresponding Golden Master. However, in order to approve changes, a Golden Master must be present.
The report is bound to the recheck version used during comparison. We currently offer no compatibility support between versions. If you need a report in a newer version, simply execute the test again with the new recheck version.
A report is structured to comply with the lifecycle of
Report +-- Suite (1..n) +-- Test (1..n) +-- Check (1..n) +-- Difference (0..n)
Per default, reports are located under
target/test-classes/retest/recheck/, which can be changed with some configuration.
If you execute your tests on a remote system such as a CI server, you can enable the recheck's report upload. This will upload your report to rehub, so that it can be accessed and downloaded to view it.
There are several types of differences that can be encountered. Note that the example output is based of the recheck.cli.
An attribute difference indicates that an element's attribute has changed. This is the most common difference and may include every attribute that is extracted through a extension. For example, this could include
text changes or changes from/to the default value of the respective attribute.
p [recheck] at 'html/body/header/nav/div': text: expected="recheck", actual="Recheck" font-weight: expected="(default or absent)", actual="700"
An inserted difference indicates that a new element has been added.
p [recheck] at 'html/body/header/nav/div': was inseted
A deleted difference indicates that an element has been removed.
p [recheck] at 'html/body/header/nav/div': was deleted
A metadata difference indicates changes in the metadata of a state. This is most likely due to the state being created on a different system or at a different time. These differences have no effect on the report and will be applied automatically. For a list of metadata, please refer to the state page.
os.name: expected="Windows 10", actual="Ubuntu"
We capture every difference that occurs. This is to retain the changes history. However, too many changes can be overwhelming, so you can apply filters to ignore them.
Filtering does not get rid of the differences, we still capture them. However, filtered difference don't cause tests to fail and will be ignored by the user interfaces. This approach allows us to changed the filters even after test execution.
A Golden Master must be present to approve changes from a report. If the Golden Master cannot be found (as indicated by the user interface), please make sure that the report is within the same project root as the Golden Master.
Differences can be maintained in two ways. Please refer to the respective documentation for the user interface on how to do so.
- Accept: The difference is expected or can be explained by changes done to the object under test. This will update the Golden Master with the new value(s).
- Ignore: If the element or attribute is dynamic or you do not care about these differences, they can be ignored and will be saved in the projects