Re: const qualifier for list APIs - Mailing list pgsql-hackers

From Tom Lane
Subject Re: const qualifier for list APIs
Date
Msg-id 1340488.1668609948@sss.pgh.pa.us
Whole thread Raw
In response to const qualifier for list APIs  (Ashutosh Bapat <ashutosh.bapat.oss@gmail.com>)
List pgsql-hackers
Ashutosh Bapat <ashutosh.bapat.oss@gmail.com> writes:
> Functions like lappend_*() in list.c do not modify the second
> argument. So it can be qualified as const. Any reason why we don't do
> that? Is it because the target pointer ptr_value is not const
> qualified?

It would be a lie in many (most?) cases, wherever somebody later pulls the
pointer out of the list without applying "const" to it.  So I can't see
that adding "const" there would be an improvement.

> So the coding practice though questionable, is
> safe and avoids unnecessary pallocs. But SonarQube does complain about
> it.

Maybe an explicit cast to (void *) would shut it up.

            regards, tom lane



pgsql-hackers by date:

Previous
From: "Anton A. Melnikov"
Date:
Subject: Make a 100_bugs.pl test more faster.
Next
From: "Anton A. Melnikov"
Date:
Subject: Re: [BUG] Logical replica crash if there was an error in a function.