Nice work. I'll look more closely a bit later but for the most part it seems reasonable. I'm uncertain on how best to untangle the wsgi-intercept aspect of the test failures. That will require a bit of thought.

I notice that your editor, or some other part of your process, has reformatted a lot of code that didn't otherwise need to change. That makes it hard to distinguish your actual changes from "cosmetic" changes.

If it is easy for you to do, could you update the PR so it is just the main changes? At some point we should probablyblack the entire codebase, but for a change like this it would be good to have a commit that is clearly only about the main change. Thanks.

If it's a pain, don't worry, we'll work through it.

Yeah, my IDE has calledblack, etc.

Code updated.

Looks good. I'll see if I can untangle the intercept stuff in a branch below this one.

Once we get that working we should be good to move this forward.

Thanksvery much for pushing this along. It is great that gabbi continues to evolve and improve.

I'm very very close. Looking good. More tomorrow.

I've now got it down to everything working except for the tests inhttps://github.com/cdent/gabbi/blob/main/gabbi/runner.py . These used wsgi-intercept in a way that can't be replicated with httpx's transport so I'm switching it to use external processes. This is taking longer than expected because there's some complex stuff going on with stdin/stdout/stderr handling that has buffering issues. It's feeling easier to switch to managing subprocesses that untangling the buffers. Will be a few more days.

Down to just two failing tests

All passing. Will see what needs to be done to target this branch.