Defining and Running Tests

[Charlie White]

After much trial and error and community support, I have been able to build a Docker image with Python source that includes multiple packages.  I am now trying to figure out tests.  I have defined a set of tests that I have previously executed with pytest.  I have created a unit_test target using the py_test rule and it generates an entire standalone python script app that executes that single unit test file as defined by "main" in the rule.  What I would like to do is run all *_test.py files that are in my unit_tests directory.  I typically break my tests into small pieces that test only one segment of the application in each.  I rely on pytest to find all the test files, and execute all the defined tests cases.  How do I accomplish this with Bazel?  

I apologize if this question has already been asked and answered, or if there is documentation somewhere that better defines how to use py_test.

Thanks

[Brian Silverman]

Each *_test rule defines a single executable file, so you will need a separate rule for each file with a main in it. However, you can do that automatically via list comprehensions and glob. Something like this (untested):

[py_test(

  name = src[:-3],

  srcs = [

    src,

  ],

) for src in glob(['*_test.py'])]

And then to test all of them at once, you can use a target wildcard like `bazel test //unit_tests/...`.

--
You received this message because you are subscribed to the Google Groups "bazel-discuss" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bazel-discuss+unsubscribe@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/bazel-discuss/a1f3277d-c5b4-4e98-841c-68f8dcfb9a62%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[rodr...@google.com]

There's some info in this thread: https://groups.google.com/d/msg/bazel-discuss/d1RKUmyve_Q/SEtBdFkYBQAJ

[Charlie White]

As has been my experience with Bazel thus far, I almost got it working. My application is a trivial python app.  The tree looks like this.

.
|____.env
|____env.sample
|____heartbeat
| |______init__.py
| |____heartbeat.py
| |____heartbeat.pyc
| |____tests
| | |______init__.py
| | |____unit
| | | |______init__.py
| | | |____bogus_test.py
| | | |____heartbeat_test.py
|____kubernetes
| |____configmap.yaml
| |____deployment.yaml
|____Makefile
|____README.md
|____requirements.txt
|____WORKSPACE

I added the following to my BUILD file:

SOURCES = glob(["heartbeat/*.py"])
UNIT_TESTS = glob(["heartbeat/tests/unit/*test.py"])

[
    py_test(
        name=src[:-3],
        deps=all_requirements,
        srcs=SOURCES + [src],
    ) for src in UNIT_TESTS
]

I can execute all the tests if I run in a for loop, such as this:

for i in heartbeat/tests/unit/*_test.py;

do
   bazel test //:${i%.py} --cache_test_results=no
done

Unfortunately when I try to run tests with a target wildcard, it does not work

$ bazel test //heartbeat/tests/unit/... --cache_test_results=no

ERROR: Skipping '//heartbeat/tests/unit/...': no targets found beneath 'heartbeat/tests/unit'

ERROR: no targets found beneath 'heartbeat/tests/unit'

INFO: Elapsed time: 0.059s

FAILED: Build did NOT complete successfully (0 packages loaded)

ERROR: Couldn't start the build. Unable to run tests

I'm certain that there is something about the targeting that I fail to grasp.

On Thursday, March 1, 2018 at 9:15:30 AM UTC-8, Brian Silverman wrote:

Each *_test rule defines a single executable file, so you will need a separate rule for each file with a main in it. However, you can do that automatically via list comprehensions and glob. Something like this (untested):

[py_test(

  name = src[:-3],

  srcs = [

    src,

  ],

) for src in glob(['*_test.py'])]

And then to test all of them at once, you can use a target wildcard like `bazel test //unit_tests/...`.

On Thu, Mar 1, 2018 at 12:09 PM, Charlie White <cha...@gmail.com> wrote:

After much trial and error and community support, I have been able to build a Docker image with Python source that includes multiple packages.  I am now trying to figure out tests.  I have defined a set of tests that I have previously executed with pytest.  I have created a unit_test target using the py_test rule and it generates an entire standalone python script app that executes that single unit test file as defined by "main" in the rule.  What I would like to do is run all *_test.py files that are in my unit_tests directory.  I typically break my tests into small pieces that test only one segment of the application in each.  I rely on pytest to find all the test files, and execute all the defined tests cases.  How do I accomplish this with Bazel?  

I apologize if this question has already been asked and answered, or if there is documentation somewhere that better defines how to use py_test.

Thanks

--
You received this message because you are subscribed to the Google Groups "bazel-discuss" group.

To unsubscribe from this group and stop receiving emails from it, send an email to bazel-discus...@googlegroups.com.

[Austin Schuh]

Which folder is your BUILD file in?  That deliminates what's before the : and after the :.  From what you have written, your target should be something like //:heartbeat/tests/unit/heartbeat_test

bazel test //... --cache_test_results=no

should do it as well.

Austin

To view this discussion on the web visit https://groups.google.com/d/msgid/bazel-discuss/6d2b4162-d40d-4305-8446-535b81620e95%40googlegroups.com.

[Charlie White]

Sorry, I edited out the virtualenv and git folders from the tree, and must have also removed BUILD.  My BUILD file is in the root directory with the WORKFILE.

[Charlie White]

Thanks, the target wildcard you specified worked great, it executes all the tests defined in my BUILD.

On Thursday, March 1, 2018 at 11:06:43 AM UTC-8, Austin Schuh wrote: