[Feature] Add test.todo()

[mxschmitt]

Discussed in https://github.com/microsoft/playwright/discussions/10914

^{Originally posted by zomars December 14, 2021}
This is pretty helpful for planning future tests and keep track of them.

Jest has it: https://jestjs.io/docs/api#testtodoname

[pavelfeldman]

We call it test.fixme!

[zomars]

We call it test.fixme!

I think they should be semantically different IMO.

[unlikelyzero]

I had a definite need for this when migrating a 5 year old angular app from nightwatch.js to Cypress.io. We almost went with jest-puppeteer over Cypress at the time specifically for this feature.

We needed to keep track of all of the tests which would exist if the app functionality existed separate from the tests which were broken.

Today, this would drastically improve our ability to manage test stubs for a 5 year old angular app which has just completed it's move to vue2.

nasa/openmct#4303

test.fixme() will work for us, but todo() would make it clearer what the open source testing community could contribute vs what has been broken

[orenmizr]

i really hoped this would be available in 2022. for semantics 👀
fixme says it's broken, we know, no time at this moment to fix the broken test.
todo says, there is nothing here yet. but we do need this test in the near future.

[pavelfeldman]

If you are looking for the parity with Jest, please use test.fixme and think of it as of test.todo. It has exact same semantics, but is slightly more flexible in that it allows putting some draft code into the test.

[zomars]

How does it has the same semantics? It clearly doesn't. Just declare that is out of scope and be done with it. But don't ignore the main reason this issue got made in the first place.

[orenmizr]

true. i can understand if it is a very low priority and i respect that completely. but it is not the same semantics. it is like saying that skip and fixme can by used for "skipping" broken tests.

[beorn]

A few reasons why .todo would be preferable as a way to use tests-as-documentation / specification:

1. test.todo doesn't require a code callback, but test.fixme does, which is annoying and adds clutter.
2. test.todo has different semantics than test.fixme.
3. test.todo is supported in other test systems, so it'll be easier to switch over to playwright from those systems.