bpo-1635741: In pickle module, inject module state from class methods

[koubaa]

This PR prepares for changing to heap types.
When switching to heap types, the clinic can be used to get the module state.

https://bugs.python.org/issue1635741

[koubaa]

@vstinner @corona10 @shihai1991 please review

[koubaa]

PickleState is not used here, but is added as an argument to avoid complicating the macro/switch case in

static PyObject *
load(UnpicklerObject *self, PickleState *st)

[tiran]

Most modules are using state for the state variable. Could you please replace st with state so we have a consistent naming scheme across the code base?

[koubaa]

Most modules are using state for the state variable. Could you please replace st with state so we have a consistent naming scheme across the code base?

@tiran I did that in my initial draft here. The reason I didn't do it here is because there are local variables named state in some functions (see for example line 3949 in save_reduce. Do you have a suggestion on what to change those names to? I don't understand exactly what those variables represent

[github-actions]

This PR is stale because it has been open for 30 days with no activity.

[shihai1991]

The reason I didn't do it here is because there are local variables named state in some functions (see for example line 3949 in save_reduce.

the state would be more exact. If there have other arguments have own this name, keep st unchanged is not a big probleam.

[koubaa]

@shihai1991 @vstinner I addressed these comments

[shihai1991]

This PR is huge :) . About adding parenthesis, I will update it when those code have relation with the bpo.

[iritkatriel]

https://bugs.python.org/issue1635741 is closed. What is the status of this PR?

[vstinner]

The change is still relevant, but should use a new issue number.

Moreover, the SC asked to put the conversion of static types to heap types on hold. @encukou and @erlend-aasland wrote https://peps.python.org/pep-0687/ which may unblock the situation but it's still a draft.

[erlend-aasland]

The change is still relevant, but should use a new issue number.

Moreover, the SC asked to put the conversion of static types to heap types on hold. @encukou and @erlend-aasland wrote https://peps.python.org/pep-0687/ which may unblock the situation but it's still a draft.

FYI, PEP 687 was accepted.

[encukou]

Pickle module state was isolated in #102982. koubaa is listed as co-author, so I hope some of this work made it in at the end.

[erlend-aasland]

Pickle module state was isolated in #102982. koubaa is listed as co-author, so I hope some of this work made it in at the end.

It did; PR #102982 was based off of Mohamed's work in #23188 (IIRC, I cherry-picked the commits from this that PR onto my new branch). See also #23188 (comment).