Danielle Madeley

I agree. I've got a branch that refractors a lot of the runner code and I'm soon going to refactor the step definition lookups to try and make them faster...

Could it be caching?

Can you replicate this with the subunit output? I'm planning on unifying all the output types into one codepath based on testtools (using the subunit codepath) in the near future....

The flags are the analogues of the xunit ones, i.e. --with-subunit. You can then convert from subunit to junitxml using a utility that comes with python-subunit, although it's not as...

Interesting to know if this is fixed with the new parser (#429). Can you turn this into a test case and put it in a pull request?

Yep, this is fixed in the new-parser branch (#429). Needs some tests though.

I've added subunit support to Lettuce via python-subunit/testtools. It's quite a bit better. I have plans to replace the xUnit support via the same API or perhaps using subunit2junitxml.

I've actually wondered about something more REST-looking like int:id returns str:value and then having the parameters exposed as kwargs with the correct types.

So I don't think Python 3 is going to happen in Lettuce unless someone takes it over. I had been doing it, but I've now moved all of my energy...

Aloe attempts to pick up where Lettuce leaves off. Uses the official Gherkin parser. On Wed, 4 Jan 2017 at 02:54, geokala wrote: > behave works reasonably well in my...