"test_plug_and_play.cc" is an optional source file for catch2 tests. If
the source file is available, the "make plug-and-play-tests" command
will use it to create a catch2 tests binary. If the source file is not
available, the "make plug-and-play-tests" command will fail. This source
file is in the .gitignore and should never be added to the git project.

The main use of test_plug_and_play.cc is for the file to be used in
conjunction with the "git bisect" command. If you don't know how the git
bisect command works, reference "git bisect --help" because otherwise
none of the following will make sense.

test_plug_and_play.cc is a file which:

1) Compilations of versions of crawl older than master can search for
catch2 tests

2) Will not be overwritten by git history changes, such as those which
occur while using git bisect.

This combination of features allows for you to write tests once, then
use them for any version of crawl which has previously existed, all the
way back to the original implementation of catch2 support for crawl.

While you are hunting down a bug introduced in newer versions of crawl,
see if you can create a catch2 test which checks for the bug in
test_plug_and_play.cc. Verify this test works with "make
plug-and-play-tests". Then, see if this test works correctly when you
"git checkout" a version of crawl which does not have the bug. If both
are true, you can use test_plug_and_play.cc to easily find the bug.

Simply use git bisect as normally, but compile and run
plug-and-play-tests instead of normal crawl. Whenever the bug is
present, your test will fail and you can mark that commit as bad.
Whenever the bug is absent, your test will pass and you can mark that
commit as good.

This test process is so easy git bisect allows you to automate it. Look
into "git bisect run" for more details.

When you've finished solving the bug, remember to move the new test from
test_plug_and_play.cc to a regular test file included in the git
project! This will help prevent regressions.