[paugier]

I'd like to work on the support for OpSpec(impl_str, [*m_args]) (parser and interpreter).

I think it is doable with current SPy and that it will be needed for "enough metaprogramming capabilities to implement e.g. dataclasses" (see https://fluiddyn.pages.heptapod.net/spy-book/metaprog-dataclass/ for the long term goal).

Note that this PR is not about *args in red functions.

[antocuni]

I'd like to work on the support for OpSpec(impl_str, [*m_args]) (parser and interpreter).

sounds like a good idea to me, thanks!
But, isn't [*x] the same as list(x)?
Maybe it might be easier to just implement list(x)?

[paugier]

I agree that for this case, list should be much simpler!

My plan was to work on the parser to support Starred syntax that I think is more generally useful.

[paugier]

Also, in SPy, tuple and list are blue.generic so they cannot be called without the type(s).

Should they become something else than just a generic to allow list(m_args)?

I seems to me that this is a different work that what I planned with this PR. But it can be interesting!

[antocuni]

Also, in SPy, tuple and list are blue.generic so they cannot be called without the type(s).

Should they become something else than just a generic to allow list(m_args)?

oh right, didn't think about it. Yes, we need to invent something to support list(something), not sure what is the best way to achieve it.

I seems to me that this is a different work that what I planned with this PR. But it can be interesting!

agreed. If your goal is to implement Starred, go ahead! The title of the PR is a bit misleading though :)

[paugier]

I think this is ready for a first review.

I don't know how to organize the tests and where to put them.

I would also need feedback on the implementation.