Re: Casting Integer to Boolean in assignment

From: Ron <ronljohnsonjr(at)gmail(dot)com>
To: pgsql-general(at)lists(dot)postgresql(dot)org
Subject: Re: Casting Integer to Boolean in assignment
Date: 2019-01-24 14:09:18
Message-ID: 3c598971-fa25-81f8-13d8-6e4ba0d6995b@gmail.com
Views: Raw Message | Whole Thread | Download mbox | Resend email
Thread:
Lists: pgsql-general

On 1/24/19 5:04 AM, Alexandre GRAIL wrote:
> Hello the list,
>
> Maybe this question has been debated before (I didn't find anything
> helpful) but :
>
> Why the default is to throw an error when casting Integer to Boolean in
> assignment, and accepting it everywhere else ?
>
> So you can type :
>
> postgres=# select 1::boolean;
>  bool
> ------
>  t
> (1 row)
>
> or
>
> postgres=# select 0::boolean;
>  bool
> ------
>  f
> (1 row)
>
>
> But you *cannot* use 1 or 0 as valid input for boolean type when inserting
> or updating :
>
> test=# CREATE TABLE test1 (a boolean);
> CREATE TABLE
> test=# INSERT INTO test1 VALUES (1);
> ERROR:  column "a" is of type boolean but expression is of type integer
> LINE 1: INSERT INTO test1 VALUES (1);
>                                   ^
> HINT:  You will need to rewrite or cast the expression.
>
>
> This behavior cannot be changed, as this cast is hard coded with
> "Implicit?=no".
>
> And added to this weirdness is the fact that '1' or '0' (with quote) is OK.
>
> So is there a reason to forbid 0 and 1 as valid boolean, without explicit
> cast ?

I'm confused at the question.  You are happily casting 1 and 0 to boolean in
the SELECT statements, and then grumbling when *not* casting them in the
INSERT statements.  Thus, why aren't you casting during the INSERT statements?

test=# INSERT INTO test1 VALUES (1::boolean);

--
Angular momentum makes the world go 'round.

In response to

Browse pgsql-general by date

  From Date Subject
Next Message Stephen Frost 2019-01-24 14:11:53 Re: [pgbackrest] Expiring the last backup?
Previous Message Ron 2019-01-24 14:01:23 Re: [pgbackrest] Expiring the last backup?