Re: strange case of "if ((a & b))" - Mailing list pgsql-hackers

From Justin Pryzby
Subject Re: strange case of "if ((a & b))"
Date
Msg-id 20210818192318.GS10479@telsasoft.com
Whole thread Raw
In response to Re: strange case of "if ((a & b))"  (Bruce Momjian <bruce@momjian.us>)
List pgsql-hackers
On Wed, Aug 18, 2021 at 03:15:21PM -0400, Bruce Momjian wrote:
> On Wed, Aug 18, 2021 at 01:28:56PM -0500, Justin Pryzby wrote:
> > > > 0002 is a bonus patch I found in my typos branch.  I will hold onto it for
> > > > later if nobody wants to deal with it.
> > > 
> > > I am ready to deal with this patch.  Should I apply it to master soon?
> > 
> > Thanks for looking at it.  I suggest not to apply 0002 - I'll resend it on
> > another thread with other, similar cleanups.
> 
> OK.
> 
> > However, I have another patch to clean up stuff like "? true : false", which
> > seems related to this patch (but maybe it should be applied separately).
> 
> Yes, that is odd.  I think it is related to the confusion that if ()
> compares non-zero(true) and zero(false), while booleans return only 1/0
> (no other values).  This explores that:
> 
>     https://stackoverflow.com/questions/22489517/c-language-boolean-expression-return-value
> 
> Do you want me to consider this patch now?

Yes, please.
It may be helpful to dispose of the first patch first.

Thanks,
Justin



pgsql-hackers by date:

Previous
From: Bruce Momjian
Date:
Subject: Re: strange case of "if ((a & b))"
Next
From: Peter Geoghegan
Date:
Subject: Re: The Free Space Map: Problems and Opportunities