Samuel Colvin

Sorry for the delay in getting back to this 🙏.

String parsing in JSON is extremely complex. We might be able to recover from some end of string situations, but it will require a lot of work. PR welcome to...

This should be fixed in #101, @willbakst can you take try it out and let me know if it solves your problem.

Yes, we'll probably avoid changing the name of the kwarg in pydantic-core to avoid breaking changes. In retrospect maybe it was a mistake therefore to change them in jiter.

closing until someone wants to take this on.

Yup, I think `pydantic_core.from_json` is pretty redundant now, it's also about 30% slower than jiter due to the PGO test cases used.

This makes a substantial difference locally, but codspeed seems to disagree ???

Locally was macos.

Not clear there's an improvement on x86, might even be worse (although I have no idea how it could be):  So I'll leave this, we can either reconsider or...

closing as this was superceded in aarch64 with #65 and x86 will need to start gain.