Message 330572 - Python tracker

Message330572

Author	methane
Recipients	Dan Rose, brett.cannon, mark.dickinson, methane, miss-islington, serhiy.storchaka, terry.reedy, tim.peters, vstinner
Date	2018-11-28.03:13:59
SpamBayes Score	-1.0
Marked as misclassified	Yes
Message-id	<1543374840.38.0.788709270274.issue34100@psf.upfronthosting.co.za>
In-reply-to

Content
I agree that modifying tuple is bad idea in general case. But in case of constants, in-place modify is the easiest approach. Our purpose is "merging" constant. On the other hand, "build new tuple and replace old tuple" approach makes two same constant tuples. If old tuple is referenced from somewhere, we fail to "merging". So I don't think we need to follow the general principle unless it reduces code. FYI, we have similar in-place editing for interning already. https://github.com/python/cpython/blob/master/Objects/codeobject.c#L47-L95

Content

I agree that modifying tuple is bad idea in general case.

But in case of constants, in-place modify is the easiest approach.

Our purpose is "merging" constant.  On the other hand, "build new tuple and
replace old tuple" approach makes two same constant tuples.
If old tuple is referenced from somewhere, we fail to "merging".

So I don't think we need to follow the general principle unless it reduces code.

FYI, we have similar in-place editing for interning already.
https://github.com/python/cpython/blob/master/Objects/codeobject.c#L47-L95

History
Date	User	Action	Args
2018-11-28 03:14:00	methane	set	recipients: + methane, tim.peters, brett.cannon, terry.reedy, mark.dickinson, vstinner, serhiy.storchaka, miss-islington, Dan Rose
2018-11-28 03:14:00	methane	set	messageid: <1543374840.38.0.788709270274.issue34100@psf.upfronthosting.co.za>
2018-11-28 03:14:00	methane	link	issue34100 messages
2018-11-28 03:13:59	methane	create