Incorrect docs: Making session-scoped fixtures execute only once

[brandon-leapyear]

https://github.com/pytest-dev/pytest-xdist/blob/master/README.rst#making-session-scoped-fixtures-execute-only-once

worker_id will never be a falsey value to satisfy the if not worker_id check. If tests are not running in distributed workers, worker_id will be set to master.

https://github.com/pytest-dev/pytest-xdist/blob/master/src/xdist/plugin.py#L221-L233

I'm not sure what the correct solution is here.

• If worker_id == 'master' guarantees that tests are not running in distributed workers, then one should just change not worker_id to worker_id == 'master'.
• Otherwise, the worker_id fixture should be changed (or another fixture should be added) to distinguish between a non-distributed master node and a distributed master node.
• Alternatively, we can not special case the master node and just do something like

 @pytest.fixture(scope="session")
 def session_data(tmp_path_factory, worker_id):
-    if not worker_id:
-        # not executing in with multiple workers, just produce the data and let
-        # pytest's fixture caching do its job
-        return produce_expensive_data()
-
-    # get the temp directory shared for by all workers
-    root_tmp_dir = tmp_path_factory.getbasetemp().parent
+    tmp_dir = tmp_path_factory.getbasetemp()
+    root_tmp_dir = tmp_dir if worker_id == 'master' else tmp_dir.parent

     fn = root_tmp_dir / "data.json"
     with FileLock(str(fn) + ".lock"):
         if fn.is_file():
             data = json.loads(fn.read_text())
         else:
             data = produce_expensive_data()
             fn.write_text(json.dumps(data))
     return data

[nicoddemus]

If worker_id == 'master' guarantees that tests are not running in distributed workers, then one should just change not worker_id to worker_id == 'master'.

This is the correct one: when distributing tests, the master node will never execute any tests, so it is not possible for worker_id be "master" when distributing tests. I'm fixing that in #582.

Thanks for the report!