[issue41901] Added some explaining to pickle errors.

[Marco Sulla]

Marco Sulla  added the comment:

I closed it for this reason:

https://github.com/python/cpython/pull/22438#issuecomment-702794261

--
stage:  -> resolved
status: open -> closed

___
Python tracker 

___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue41901] Added some explaining to pickle errors.

[Marco Sulla]

Marco Sulla  added the comment:

I do not remember the problem I had, but when I experimented with frozendict I 
get one of these errors. I failed to understand the problem so I added the 
additional info.

Maybe adding an assert in debug mode? It will be visible only to devs.

--

___
Python tracker 

___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue41901] Added some explaining to pickle errors.

[Serhiy Storchaka]

Serhiy Storchaka  added the comment:

It was intentionally left a generic "cannot pickle 'type' object". Because 
there are many ways to make an object pickleable, and it is non-pickleable in 
all other cases.

* "missing `__getstate__()` on variable-length object"

It creates false impression that implementing __getstate__() is the only way to 
solve the problem. Actually you can make that object pickleable by:

- implementing __reduce__() or __reduce_ex__()
- implementing __getnewargs__() or __getnewargs_ex__()
- implementing __getstate__()
- make it a subclass of list or dict

So more correct error message should contain "missing __reduce__(), 
__reduce_ex__(), __getnewargs__(), __getnewargs_ex__() and __getstate__() on 
variable-length object which is not a list or dict". And even this is not 
completely correct, because it has __reduce__() and __reduce_ex__() inherited 
from class object, but it is not enough. Completely correct error message can 
be hundreds characters long.

And this still may be misleading error message, because the root cause is not 
that some object is not pickleable, but that you try to pickle an object which 
has an unexpected non-pickleable component.

* "basic size greater than expected"

Basic size is an implementation detail, and the end user can have no guess what 
is it. Actually it means that an object implemented in C likely has some inner 
state and the pickle module does not have access to it because the object does 
not implement __reduce__(), __reduce_ex__(), __getnewargs__(), 
__getnewargs_ex__() or __getstate__().

* "object has no `__new__()`"

Every object has __new__ at Python level, even if its tp_new slot is NULL. 
tp_new is the C implementation detail and should not be mentioned in the error 
message. And the object without tp_new still can be pickleable, if it 
implements __reduce__() or __reduce_ex__().

It is all too complicated. It is better to omit details in error messages than 
provide not correct and direct user in wrong way.

--
nosy: +serhiy.storchaka

___
Python tracker 

___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com

[issue41901] Added some explaining to pickle errors.

[Marco Sulla]

New submission from Marco Sulla :

All pickle error messages in typeobject.c was a generic "cannot pickle 'type' 
object". Added some explaining for every individual error.

--
components: Interpreter Core
messages: 377747
nosy: Marco Sulla
priority: normal
pull_requests: 21494
severity: normal
status: open
title: Added some explaining to pickle errors.
type: enhancement
versions: Python 3.10

___
Python tracker 

___
___
Python-bugs-list mailing list
Unsubscribe: 
https://mail.python.org/mailman/options/python-bugs-list/archive%40mail-archive.com