Path: csiph.com!x330-a1.tempe.blueboxinc.net!usenet.pasdenom.info!aioe.org!eternal-september.org!feeder.eternal-september.org!.POSTED!not-for-mail
From: markspace <-@.>
Newsgroups: comp.lang.java.programmer
Subject: Re: boolean to int : was char to decimal
Date: Fri, 06 May 2011 07:07:14 -0700
Organization: A noiseless patient Spider
Lines: 56
Message-ID: <iq0vat$ip6$1@dont-email.me>
References: <92ea64F3avU1@mid.individual.net> <Xns9EDBB8224D5B6vaj4088ianshef@138.125.254.103> <iptlkq$62g$2@lust.ihug.co.nz> <iq0r9c$efs$1@dont-email.me>
Mime-Version: 1.0
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Content-Transfer-Encoding: 7bit
Injection-Date: Fri, 6 May 2011 14:07:26 +0000 (UTC)
Injection-Info: mx02.eternal-september.org; posting-host="l8hxu7VxWTF0R70MQnZDgg"; logging-data="19238"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX1/pyAyktrZx25xhsmPwR6KWLgZi/WcuF3g="
User-Agent: Mozilla/5.0 (Windows; U; Windows NT 6.1; en-US; rv:1.9.2.17) Gecko/20110414 Thunderbird/3.1.10
In-Reply-To: <iq0r9c$efs$1@dont-email.me>
Cancel-Lock: sha1:cja4TNmcS9zV3sgEPTUlnDysZ1g=
Xref: x330-a1.tempe.blueboxinc.net comp.lang.java.programmer:3686

On 5/6/2011 6:00 AM, Jeff Higgins wrote:
> C++ test:
> int test = (int(d2 > e) << 1) + int(d3 > e);
>
> Java test:
> int t = d2 > e ? 1<<1 : 0;
                    ^^^^

Well, this is a 2.  The Java coding conventions say to declare this as a 
constant, I'm not sure that is valuable here though


> int test = d3 > e ? t+1 : t;

   t += d3 > e ? 1 : 0;

The whole thing could be done in one go, of course:

   int test = ((d2 > e) ? 2 : 0) + ((d3 > e) ? 1 : 0);


>
> switch(test)
> { case(0):
> case(1):
> case(2):
> case(3):
> }

Blech.  So the whole point here seems to be to set the 1 bit if d2 > e, 
and set the 0 bit if d2 > e.  Why not just do that?

if( d2 > e ) {
   if( d3 > e ) {
     // case 0
   }  else {
     // case 1
   }
} else {
   if( d3 > e ) {
     // case 2
   }  else {
     // case 3
   }
}

I don't think there's more tests for the CPU in this case.  In fact the 
jump to implement the switch might be more expensive that just a simple 
branch test.  I'd have to know what is motivating the use of switch 
instead of if-else.

Overall, I feel the last one, the if-else in Java, is the most clear and 
maintainable.  Given your apparent confusing trying to translate the C++ 
version literally, I think I'd say that promoting booleans to ints is a 
bad idea, as it produces crap.