isInteger, isFloat functions?

[FredOz]

Re: isInteger, isFloat functions?

In article <BD75213E.21826 %phil.ronanzzz@ virgin.net>,
phil.ronanzzz@v irgin.net says...[color=blue]
> On 20/9/04 9:29 pm, Fred Oz wrote:
>[color=green]
> > Philip Ronan wrote:[/color]
>
> Sorry, I shouldn't have mentioned division by zero in the first place.
>[/color]

Don't be sorry, the discussion was very useful (to me, anyway).
[color=blue]
> What I meant to refer to is the Javascript (ECMA-262) entity called
> "Infinity". Any number greater than (roughly?) 10^308 equates to "Infinity"
> in Javascript.
>[/color]

Spot on.
[color=blue]
> Let's get back to the original question:
>
> If I entered this into a text box:
> "9999999---99999.9" (replace "---" with 300 "9"s)
> then it would fail this test for integers:
> alert(!/\./.test(+t) && +t==t)
>
> OK now?
>[/color]

Yes. Incidentally, the OP has not followed up ... hopefully something
here was of use?

[Andrew Thompson]

Re: isInteger, isFloat functions?

On Mon, 20 Sep 2004 22:39:22 +0200, Lasse Reichstein Nielsen wrote:
[color=blue]
> /L 'and 1/-0 == -Infinity, even though 0 == -0 ... just don't think about it:)'[/color]

What I was about to add in my 1st post was that
the 'simple'situati ons are obvious (to me), but
the others make my head hurt, so I add checks to
avoid them! ;-)

--
Andrew Thompson
http://www.PhySci.org/codes/ Web & IT Help
http://www.PhySci.org/ Open-source software suite
http://www.1point1C.org/ Science & Technology
http://www.lensescapes.com/ Images that escape the mundane

[Grant Wagner]

Re: isInteger, isFloat functions?

Fred Oz wrote:
[color=blue]
> Philip Ronan wrote:
>[color=green]
> > We are discussing JAVASCRIPT here, right?[/color]
>
> Yes. And like all good programmers, if there is *any* chance
> that your algorithm may result in division by zero, you
> absolutely must test for it and avoid it.[/color]

Agreed.
[color=blue]
> As Andrew pointed out, 3 browsers gave 3 different (likely fully
> ECMA-262 compliant) results. So if you don't test for division
> by zero in the first place, you had better have done it afterward
> because now you must deal with a bigger problem - infinity![/color]

Andrew pointed out that 3 browsers gave consistent results. His
original message contained:

---

All three of IE 6.0026, Moz. 1.7 and Opera 6.54 (editor: probably
meant 7.54, but I tested 6.05 as well and it was the same) give
consistent results for me..

javascript:aler t(1/0) -> 'Infinity'
javascript:aler t(0/0) -> 'NaN'
javascript:aler t(-1/0) -> '-Infinity'

---

I have to disagree that "Infinity" is a bigger problem. The behaviour
of Number.POSITIVE _INFINITY (and NEGATIVE_INFINI TY) are clearly
documented and if the ECMAScript is following the implementation, the
results will be the same regardless of the environment:

<url:

http://devedge.netscape.com/library/manuals/2000/javascript/1.3/reference/number.html#1193380

/>


So you can catch division by zero with:

var result = (nominator + Number.MIN_VALU E) / denominator;
if (Math.abs(resul t) == Number.POSITIVE _INFINITY) { ....


Number.MIN_VALU E should not affect the result too terribly much, and
it prevents 0/0 == NaN, 0/0 now becomes Infinity as well. Of course,
identifying and dealing with demoninator == 0 would be better, but
allowing division by zero in JavaScript is no where near as fatal as
it can be in other languages.

--
Grant Wagner <gwagner@agrico reunited.com>
comp.lang.javas cript FAQ - http://jibbering.com/faq

[Andrew Thompson]

Re: isInteger, isFloat functions?

On Tue, 21 Sep 2004 13:34:05 GMT, Grant Wagner wrote:

(Browser tesing - versions)[color=blue]
> Opera 6.54 (editor: probably
> meant 7.54, ..[/color]

Correct, thank you. ( And I *was* doing so
well in being precise in that answer.. ;-)

[Dr John Stockton]

Re: isInteger, isFloat functions?

JRS: In article <isa8oeo9.fsf@h otpop.com>, dated Mon, 20 Sep 2004
22:42:30, seen in news:comp.lang. javascript, Lasse Reichstein Nielsen
<lrn@hotpop.com > posted :[color=blue]
>"Evertjan." <exjxw.hannivoo rt@interxnl.net > writes:
>[color=green]
>> I don't think infinity is an integer but I wouldn't mind if it is.[/color]
>
>It is a number in Javascript. Whether it is an integer, I won't
>begin to think about. Hmm. Math.floor(Infi nity) == Infinity
>Nah.
>[color=green]
>> (+t+1)!=+t ?[/color]
>
>You don't need infinity for that in IEEE-754 floating point numbers.
>As soon as the numbers becomes so large, that the precission is less
>than 1, then t+1 cannot be represented and the result is the same as
>t.
>
>Example:
>---
> var t = Math.pow(2,53);
> if (t+1 == t) { alert("surprice "); }
>---[/color]

But try it with var t = - Math.pow(2,53)
var t = - Math.pow(2,53) - 66666

***

The OP needs to be aware that, in current javascript, all Numbers are
floating-point binary IEEE-754 doubles.

--
© John Stockton, Surrey, UK. ?@merlyn.demon. co.uk Turnpike v4.00 MIME. ©
Web <URL:http://www.merlyn.demo n.co.uk/> - FAQqish topics, acronyms & links;
some Astro stuff via astro.htm, gravity0.htm; quotes.htm; pascal.htm; &c, &c.
No Encoding. Quotes before replies. Snip well. Write clearly. Don't Mail News.

[Michael Winter]

[Correction] Re: isInteger, isFloat functions?

On Mon, 20 Sep 2004 14:05:10 GMT, Michael Winter
<M.Winter@bluey onder.co.invali d> wrote:

[snip]
[color=blue]
> /^(0|[1-9])\d*$/[/color]

This should have actually been

/^(0|[1-9]\d*)$/

Where this same pattern, (0|[1-9])\d*, occurs in other expressions (almost
all), it should be replaced with (0|[1-9]\d*).

[snip]

Mike

--
Michael Winter
Replace ".invalid" with ".uk" to reply by e-mail.