04:18 <catern> is there a way to list all the structs that are defined in an ffi/lib object? this header has a lot of different structs and I want to do some introspection to make them easier to deal with

06:10 <mattip> dir(lib) ?

11:17 <cfbolz> arigo: help!

11:17 <cfbolz> arigo: I have an extremely weird JIT bug that looks like a finally is skipped in a relatively complicated program (running pypy tests on pypy)

11:18 <cfbolz> arigo: I tried to describe everything I found out in this commit message: https://foss.heptapod.net/pypy/pypy/-/commit/0d762269b18aeb63613198bb8704fea4a197c2e3

13:00 <catern> mattip: lib only lists defines, not structs/types

13:03 <catern> (afaict)

13:09 <mattip> maybe this can help https://cffi.readthedocs.io/en/latest/ref.html#ffi-getctype-ffi-list-types

13:09 <mattip> but I couldn't get ffi.getctype(ffi.list_types()[1][0]) to work

13:35 <catern> mattip: wow I completely missed that method list_types, thanks! that's perfect!

13:43 <mattip> catern: if you have a complete snippet maybe you could make a PR to the docs

13:44 <mattip> I think it is a common desire

13:44 <mattip> to be able to print some definition of structs

14:12 <catern> well - list_types does exactly what I want - there's not much more to PR (it worked fine with ffi.new for me)

22:02 <mjacob> arigo: so far i saw the "cannot be rendered as a plain raw address" error only with bytearray. at some point i planned to make a raw memory-backed version of bytearray, but for that i would need some help (maybe at the next sprint, whenever that will be?)

22:02 <mjacob> arigo: but a general solution would be very nice