next »

[perkiset]

This is a PHP adaptation of NutBalls ASP code for storing IP addresses as big ints. I have made this about as efficient as I know how... if anyone else has ideas that'd be great. Note that if you use the defined constants then you'll need to "include" this code rather than put it inline - that's because the defines will happen _after_ your code alls the functions and you'll get nasty Div by Zero errors. Also, in the IntToIP conversion I just couldn't figure out why PHP was returning the 2s compliment of the value rather than the real deal... probably something to do with the mod operator. Anyway, this works correctly. I also learned something during this exercise - the ^ is not the exponent operator in PHP  

Code:

define('_OCT0', pow(256, 3));
define('_OCT1', pow(256, 2));
define('_OCT2', 256);
function IPtoInt($inStr)
{
if (!preg_match('/([0-9]{1,3})\.([0-9]{1,3})\.([0-9]{1,3})\.([0-9]{1,3})/', $inStr, $parts))
throw new Exception("IPtoInt: Invalid IP Address ($inStr)");

return ($parts[1] * _OCT0) + ($parts[2] * _OCT1) + ($parts[3] * _OCT2) + $parts[4];
}

function InttoIP($inInt)
{
$oct4 = $inInt % 256;
$oct4 = ($oct4 < 0) ? 256 + $oct4 : $oct4;
return
(($inInt / _OCT0) % 256) . '.' .
(($inInt / _OCT1) % 256) . '.' .
(($inInt / _OCT2) % 256) . '.' .
$oct4;
}

[cdc]

I also learned something during this exercise - the ^ is not the exponent operator in PHP D'oh!  ROFLMAO

I would have guessed **, but it looks like they implemented it with a function.

<edit>was missing a / </edit>

[perkiset]

Here it is done in PERL by Dirk:

http://www.perkiset.org/forum/perl/converting_ip_addresses_to_integers_and_vice_versa-t174.0.html;new

[Fatty]

Since you're multiplying by powers of 2 why not use shifts?

IP_int = octet1 << 24 + octet2 << 16 + octect3 << 8 + octet4

If you're storing in a DB as an int you've got INET_ATON and INET_NTOA... Let the DB worry about it.

Fatty

[perkiset]

IP_int = octet1 << 24 + octet2 << 16 + octect3 << 8 + octet4

You're completely right Fatty - that'd be easier on the math processing I believe.

If you're storing in a DB as an int you've got INET_ATON and INET_NTOA... Let the DB worry about it.

Dirk covered that in a discussion about this very topic - these threads are more about just leaving code out there for someone with a specific need ie., "I need to do it in PHP." - but thanks

/p

[spazdr8cr]

Hi everybody, and thanks for a great resource, epsecially Perkiset.

I love the notice that the topic is old, as I am frequently guilty of reviving old threads. However I must do so again 

Trying to use this code I may have found an error, or I may be making an error myself. I convert the IP to an integer, and then when I try to convert the integer back to an IP, the last octet is always 255. Is this an error in the code or am I screwing it up somehow? Probably the latter. I am using the code given above by Perkiset, only modified to supply an IP and output results to the browser:

Code:

<?php

define('_OCT0', pow(256, 3));
define('_OCT1', pow(256, 2));
define('_OCT2', 256);

function IPtoInt($inStr)
{
if (!preg_match('/([0-9]{1,3})\.([0-9]{1,3})\.([0-9]{1,3})\.([0-9]{1,3})/', $inStr, $parts)) 
throw new Exception("IPtoInt: Invalid IP Address ($inStr)");

return ($parts[1] * _OCT0) + ($parts[2] * _OCT1) + ($parts[3] * _OCT2) + $parts[4];

}

function InttoIP($inInt)
{
$oct4 = $inInt % 256;
$oct4 = ($oct4 < 0) ? 256 + $oct4 : $oct4;
return 
(($inInt / _OCT0) % 256) . '.' . 
(($inInt / _OCT1) % 256) . '.' .
(($inInt / _OCT2) % 256) . '.' .
$oct4;
}

$ip1 = "206.248.172.160";

echo IPtoInt($ip1)."<br><br>";

//run script to get result of IPtoInt, then insert below

$ip2 = "3472403616";

echo InttoIP($ip2);

?>

So I am using the code as given by Perkiset, except for the echoes I added myself. I run the script once to get the integer from the first function, then insert it into the code, and then run it again to see if i get the original IP back. But the last octet is always 255. Am I doing something wrong or is the code wrong?

I calculated the result of IptoInt manually using a calculator, it seems correct.

The InttoIp I'm not sure of, I'm not familiar with that code at all, particularly "$oct4 = ($oct4 < 0) ? 256 + $oct4 : $oct4;" ... I have no idea what that means. So I'm at a loss there. Not familiar with the "?" or the ":" in this context.

Thanks!

[perkiset]

Hey Spaz -

Well, I was baffled because it works fine on my machines... but all of my production servers (where I tested it) are big ol' tweaked out 64 bit machines... I just ran it on a smaller machine and got the same answer, 255. This is clearly a PHP overflow problem, but I'm not sure how ATM... I'll do a little poking around and see what I can see.

/p

[nutballs]

if there is a MODULUS in there somewhere, i would look at that. In ASP (at least on 32 bit systems) you would overflow on mod. I assume thats what the % is.

my IP code is somewhere in here from that original post, though its asp, the concept is the same for the whacky math.

the basics are, divide the number by 2 first. then do all the math, then remember if it was an odd or even number.

[perkiset]

It is the mod operator NBs, (%) - overflows on older and smaller, works fine on bigger and newer. 

PHP docs say that in the event of an overflow, PHP reinterprets an INT as a FLOAT so that you have enough digit width ... but that does not seem to be happening. Before I concede to the math I'd like to see if I can tell PHP to do it right first...

[spazdr8cr]

Thanks for the replies guys.

This was running on my shared hosting (php 5.2). I tried running it in OSX this morning on my somewhat old (non-Intel) G5 (php 4.4) and got the same result.

Incidentally I discovered something that was interesting and a good lesson for me. Although when I wrote this tester script I felt kind of lame just typing in the result of the first function as in:

Code:

$ip2 = "3472403616";

it turns out had i done

Code:

$ip2 = IPtoInt($ip1);

it seemingly works fine, and I wouldn't have noticed the problem until much later when I tried to pull these out of the database and convert them.

[perkiset]

Actually it is my Mac that I tried it on as well and got the error, but this is an Intel 8 core 64b machine... so I'da thunk that there'd not be the problems. I'm wondering if it is a compile switch or something now...

[spazdr8cr]

I spent the last while translating nutballs' code from here: http://www.perkiset.org/forum/asp_net/classicasp_ip_to_num_and_num_to_ip_for_database_storage-t72.0.html;msg263#msg263 into php. it seems to work although i only tried it on a couple of ip's so far. Good learning experience, never looked at ASP before. Anyway this seems to work

Code:

define('_OCT0', pow(256, 3));
define('_OCT1', pow(256, 2));
define('_OCT2', 256);

function InttoIP($inInt)
{
$oct4 = $inInt % 256;
$oct4 = ($oct4 < 0) ? 256 + $oct4 : $oct4;
$a = (($inInt / _OCT0) % 256);
$b = (($inInt / _OCT1) % 256);
$c = (($inInt / _OCT2) % 256);

// this is the nutballs stuff, or at least my version of it

$tempnum = intval($inInt/2);
$tempnum = ($tempnum*2);
$adjust = $inInt - $tempnum;
$d = ((intval(($inInt/2)) % 128)*2)+$adjust;

//

$newip = $a . "." . $b . "." . $c . "." . $d;
return $newip;

}

Thanks guys.

[perkiset]

Try this Spaz - I compressed the maths to one line and spit out the return value without any string concatenations, which is way faster.

Code:

<?php
function InttoIP($inInt)
{
$oct4 = $inInt % 256;
$oct4 = ($oct4 < 0) ? 256 + $oct4 : $oct4;

$a = (($inInt / _OCT0) % 256); 
$b = (($inInt / _OCT1) % 256);
$c = (($inInt / _OCT2) % 256);
$d = ((intval(($inInt/2)) % 128) * 2) + ($inInt - (intval($inInt/2)) * 2);

return "$a.$b.$c.$d";
}
?>

[spazdr8cr]

You rock Perk, thanks!