Testing edalize


To run the tests, call pytest.


Mocks for commands

We provide mocks (stand-ins) for all tools that we want to exercise in tests (located in tests/mock_commands/). These mocks are very simplified “models” of the actual tool, and are called instead of the actual tool. They are prepended to PATH by the setup code in the edalize_common.make_edalize_test() pytest fixture.

In the easiest case, these mocks just write out the commandline they were called with (into a file <tool>.cmd).

In a more complex test setup (e.g. for vcs),

  • if a tool is creating files, we create it too (the file to create is taken from a tool option given)

  • we make the file executable

  • we set the access and modified times of generated files to the current time


To define a testcase, use the edalize_common.make_edalize_test() pytest factory fixture. This defines a factory that you can call to set up a mocked-up backend appropriately. See the documentation for edalize_common.TestFixture for details of the supported keywords.

The backend attribute of the returned fixture has configure(), build() and run() methods. The testcase should call these in order, setting up files as necessary between calls and checking whether the results match by calling the fixture’s compare_files() method.

If the environment variable GOLDEN_RUN is set, the compare_files() method copies the generated files are copied to become the new reference files, rather than checking their contents.

Helper Module