I’ve been starting to use TestItemRunner
’s new @testsnippet
and @testmodule
mechanisms to help organize testing code, and I’ve found it very helpful.
One thing that’s a setback, though, is that if a test fails and I need to do some troubleshooting, it’s not clear how to proceed.
- Setting a VS Code breakpoint generally doesn’t work - it did work for a little while after the feature was released a couple months ago, but these days all my tests either crash when I hit the debugger button[1], or they just don’t stop at the breakpoint.
- I tried
@run_package_tests filter = ti -> contains(ti.filename, "my_test")
from the REPL, but this generally crashes too[2]. - Copying & pasting the test code into my REPL, which was my previous go-to workaround, doesn’t work anymore because it now requires the setup module/snippet to function.
Does anyone have any recommendations for this?
[1] error generally looks like The test process crashed while running this test
, and stack trace shows @ Main ~/.vscode/extensions/julialang.language-julia-1.127.2/scripts/error_handler.jl:15 ERROR: Endpoint is not running, the current state is closed
.
[2] error generally looks like LoadError: UndefVarError: 'Foo' not defined
, where ‘Foo’ is the name of the package being tested. Even if I put using Foo
in the test.