Determining how much code your tests cover
Use code coverage to focus new test development on areas that lack adequate testing.
Overview
By implementing code coverage, you can visualize and measure how much of your code you are actually testing. Use it during development to identify areas that your tests are missing. Although achieving a high level of coverage is an excellent goal, code coverage alone doesn’t ensure that your tests are doing their job and are robust enough for unexpected behaviors. Be sure to pair high code coverage with well-written tests.
Code coverage can answer:
What code is actually running when you perform your tests?
What parts of your code aren’t you testing?
Have you written enough tests to ensure that you’re checking all of your code?
Enable code coverage in your test plan
Code coverage is a testing option you can configure for your test plans. When you enable code coverage, the build system instructs the code to gather coverage data based on the frequency that it calls methods and functions. The code coverage option can collect data to report on tests of correctness and of performance, whether unit tests or UI tests.
You enable code coverage in the Configurations pane of your test plan using the following steps:
Open a test plan from the Project navigator or the scheme editor. For more information on creating a test plan, see Improving code assessment by organizing tests into test plans.
Click the Configurations tab.
Select a specific configuration, or select Shared Settings to enable testing across all configurations.
Scroll down to the Code Coverage section.
Click the value for Code Coverage and select the “Gather coverage for” checkbox on the popover.
Use the options from the pop-up menu to select targets to collect the information from.
[Image]
Examine code coverage results
After completing a test run, Xcode uses the coverage data to create a report in the Coverage pane of the Report navigator. The coverage report shows summary information about the test run, a listing of source files and functions within the files, and the coverage percentage for each.
[Image]
The source editor shows counts for each line of code in the file, and highlights code that didn’t execute. It highlights areas of code that need coverage rather than areas that have coverage.
For example, positioning the pointer over the Calculator.input(_:) method in the coverage report above shows a button that takes you to the annotated source code.
[Image]
The coverage annotation appears on the right and shows the number of times that the test executed a particular part of the code. You can hover over areas with red highlights to identify code your test didn’t cover.
[Image]
According to the counts in the screenshot above, the test frequently called the Calculator.input(_:) method. However, there are sections of the method that the test didn’t call. This report data indicates an opportunity to write a test for a missing condition to ensure that the error handling works the way you intend.