List of files required to run the test. The filenames are relative to the test WORKING_DIRECTORY unless an absolute path is specified.

If set to a list of files, the test will not be run unless all of the files exist.


Suppose that test.txt is created by test baseTest and none.txt does not exist:

add_test(NAME baseTest ...)   # Assumed to create test.txt
add_test(NAME fileTest ...)

# The following ensures that if baseTest is successful, test.txt will
# have been created before fileTest is run
set_tests_properties(fileTest PROPERTIES
  DEPENDS baseTest

add_test(NAME notRunTest ...)

# The following makes notRunTest depend on two files. Nothing creates
# the none.txt file, so notRunTest will fail with status "Not Run".
set_tests_properties(notRunTest PROPERTIES
  REQUIRED_FILES "test.txt;none.txt"

The above example demonstrates how REQUIRED_FILES works, but it is not the most robust way to implement test ordering with failure detection. For that, test fixtures are a better alternative (see FIXTURES_REQUIRED).