You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.
Some classes have the replace() method, which creates a modified copy of the object (modified values are provided as keyword arguments). Named tuples have the _replace() for this (to avoid conflict with attribute replace). Dataclasses provide a global function for this.
I proposed to generalize dataclasses.replace() to support all classes which need this feature. By the result of the discussion on discuss.python.org, the new function will be added in the copy module as copy.replace(). dataclasses.replace() will continue to support only dataclasses. Dataclasses, named tuples and all classes which currently have the replace() method with suitable semantic will get also the __replace__() method. Now you can add such feature in new classes without conflicting with the replace attribute, and use this feature in general code without conflicting with str.replace() and like.
For now, copy.replace() is more limited than copy.copy() and does not fall back to use the powerful pickle protocol.
It creates a modified copy of an object by calling the object's
__replace__() method.
It is a generalization of dataclasses.replace(), named tuple's _replace()
method and replace() methods in various classes, and supports all these
stdlib classes.
It creates a modified copy of an object by calling the object's
__replace__() method.
It is a generalization of dataclasses.replace(), named tuple's _replace()
method and replace() methods in various classes, and supports all these
stdlib classes.
Feature or enhancement
Has this already been discussed elsewhere?
I have already discussed this feature proposal on Discourse
Links to previous discussion of this feature:
https://discuss.python.org/t/generalize-replace-function/28511
Proposal:
Some classes have the
replace()method, which creates a modified copy of the object (modified values are provided as keyword arguments). Named tuples have the_replace()for this (to avoid conflict with attributereplace). Dataclasses provide a global function for this.I proposed to generalize
dataclasses.replace()to support all classes which need this feature. By the result of the discussion on discuss.python.org, the new function will be added in thecopymodule ascopy.replace().dataclasses.replace()will continue to support only dataclasses. Dataclasses, named tuples and all classes which currently have thereplace()method with suitable semantic will get also the__replace__()method. Now you can add such feature in new classes without conflicting with thereplaceattribute, and use this feature in general code without conflicting withstr.replace()and like.For now,
copy.replace()is more limited thancopy.copy()and does not fall back to use the powerful pickle protocol.Linked PRs
The text was updated successfully, but these errors were encountered: