IPLib is a modern, PSR-compliant, test-driven IP addresses and subnets manipulation library. It implements primitives to handle IPv4 and IPv6 addresses, as well as IP ranges (subnets), in CIDR format (like ::1/128
or 127.0.0.1/32
) and in pattern format (like ::*:*
or 127.0.*.*
).
IPLib has very basic requirements as:
- Works with any PHP version greater than 5.3.3 (PHP 5.3.x, 5.4.x, 5.5.x, 5.6.x, 7.x, and 8.x are fully supported).
- No external dependencies
- No special PHP configuration needed (yes, it will always work even if PHP has not been built with IPv6 support!).
Download the latest version, unzip it and add these lines in our PHP files:
require_once 'path/to/iplib/ip-lib.php';
Simply run
composer require mlocati/ip-lib
or add these lines to your composer.json
file:
"require": {
"mlocati/ip-lib": "^1"
}
To parse an IPv4 address:
$address = \IPLib\Address\IPv4::parseString('127.0.0.1');
To parse an IPv6 address:
$address = \IPLib\Address\IPv6::parseString('::1');
To parse an address in any format (IPv4 or IPv6):
$address = \IPLib\Factory::parseAddressString('::1');
$address = \IPLib\Factory::parseAddressString('127.0.0.1');
$address = \IPLib\Factory::parseAddressString('::1');
// This will print ::
echo (string) $address->getPreviousAddress();
// This will print ::2
echo (string) $address->getNextAddress();
For addresses:
$address = \IPLib\Factory::parseAddressString('::1');
// This will print ::1
echo (string) $address->getAddressAtOffset(0);
// This will print ::2
echo (string) $address->getAddressAtOffset(1);
// This will print ::3
echo (string) $address->getAddressAtOffset(2);
// This will print ::3e9
echo (string) $address->getAddressAtOffset(1000);
// This will print ::
echo (string) $address->getAddressAtOffset(-1);
// This will print NULL
echo var_dump($address->getAddressAtOffset(-2));
For ranges:
$range = \IPLib\Factory::parseRangeString('::ff00/120');
// This will print ::ff00
echo (string) $range->getAddressAtOffset(0);
// This will print ::ff10
echo (string) $range->getAddressAtOffset(16);
// This will print ::ff64
echo (string) $range->getAddressAtOffset(100);
// This will print NULL because the address ::1:0 is out of the range
var_dump($range->getAddressAtOffset(256));
// This will print ::ffff
echo (string) $range->getAddressAtOffset(-1);
// This will print ::fff0
echo (string) $range->getAddressAtOffset(-16);
// This will print ::ff00
echo (string) $range->getAddressAtOffset(-256);
// This will print NULL because the address ::feff is out of the range
var_dump($range->getAddressAtOffset(-257));
To parse a subnet (CIDR) range:
$range = \IPLib\Range\Subnet::parseString('127.0.0.1/24');
$range = \IPLib\Range\Subnet::parseString('::1/128');
To parse a pattern (asterisk notation) range:
$range = \IPLib\Range\Pattern::parseString('127.0.0.*');
$range = \IPLib\Range\Pattern::parseString('::*');
To parse an andress as a range:
$range = \IPLib\Range\Single::parseString('127.0.0.1');
$range = \IPLib\Range\Single::parseString('::1');
To parse a range in any format:
$range = \IPLib\Factory::parseRangeString('127.0.0.*');
$range = \IPLib\Factory::parseRangeString('::1/128');
$range = \IPLib\Factory::parseRangeString('::');
You can calculate the smallest range that comprises two addresses:
$range = \IPLib\Factory::getRangeFromBoundaries('192.168.0.1', '192.168.255.255');
// This will print 192.168.0.0/16
echo (string) $range;
You can also calculate a list of ranges that exactly describes all the addresses between two addresses:
$ranges = \IPLib\Factory::getRangesFromBoundaries('192.168.0.0', '192.168.0.5');
// This will print 192.168.0.0/30 192.168.0.4/31
echo implode(' ', $ranges);
$range = \IPLib\Factory::parseRangeString('127.0.0.*');
// This will print 127.0.0.0
echo (string) $range->getStartAddress();
// This will print 127.0.0.255
echo (string) $range->getEndAddress();
Both IP addresses and ranges have a toString
method that you can use to retrieve a textual representation:
// This will print 127.0.0.1
echo \IPLib\Factory::parseAddressString('127.0.0.1')->toString();
// This will print 127.0.0.1
echo \IPLib\Factory::parseAddressString('127.000.000.001')->toString();
// This will print ::1
echo \IPLib\Factory::parseAddressString('::1')->toString();
// This will print ::1
echo \IPLib\Factory::parseAddressString('0:0::1')->toString();
// This will print ::1/64
echo \IPLib\Factory::parseRangeString('0:0::1/64')->toString();
When working with IPv6, you may want the full (expanded) representation of the addresses. In this case, simply use a true
parameter for the toString
method:
// This will print 0000:0000:0000:0000:0000:0000:0000:0000
echo \IPLib\Factory::parseAddressString('::')->toString(true);
// This will print 0000:0000:0000:0000:0000:0000:0000:0001
echo \IPLib\Factory::parseAddressString('::1')->toString(true);
// This will print 0fff:0000:0000:0000:0000:0000:0000:0000
echo \IPLib\Factory::parseAddressString('fff::')->toString(true);
// This will print 0000:0000:0000:0000:0000:0000:0000:0000
echo \IPLib\Factory::parseAddressString('::0:0')->toString(true);
// This will print 0001:0002:0003:0004:0005:0006:0007:0008
echo \IPLib\Factory::parseAddressString('1:2:3:4:5:6:7:8')->toString(true);
// This will print 0000:0000:0000:0000:0000:0000:0000:0001/64
echo \IPLib\Factory::parseRangeString('0:0::1/64')->toString();
The address and range objects implements the __toString()
method, which call the toString()
method.
So, if you want the string (short) representation of an object, you can do any of the following:
$address = \IPLib\Address\IPv6::parseString('::1');
// All these will print ::1
echo $address->toString();
echo $address->toString(false);
echo (string) $address;
All the range types offer a contains
method, and all the IP address types offer a matches
method: you can call them to check if an address is contained in a range:
$address = \IPLib\Factory::parseAddressString('1:2:3:4:5:6:7:8');
$range = \IPLib\Factory::parseRangeString('0:0::1/64');
$contained = $address->matches($range);
// that's equivalent to
$contained = $range->contains($address);
Please remark that if the address is IPv4 and the range is IPv6 (or vice-versa), the result will always be false
.
All the range types offer a containsRange
method: you can call them to check if an address range fully contains another range:
$range1 = \IPLib\Factory::parseRangeString('0:0::1/64');
$range2 = \IPLib\Factory::parseRangeString('0:0::1/65');
$contained = $range1->containsRange($range2);
If you want to know if an address is within a private network, or if it's a public IP, or whatever you want, you can use the getRangeType
method:
$address = \IPLib\Factory::parseAddressString('::');
$type = $address->getRangeType();
$typeName = \IPLib\Range\Type::getName($type);
The most notable values of the range type are:
\IPLib\Range\Type::T_UNSPECIFIED
if the address is all zeros (0.0.0.0
or::
)\IPLib\Range\Type::T_LOOPBACK
if the address is the localhost (usually127.0.0.1
or::1
)\IPLib\Range\Type::T_PRIVATENETWORK
if the address is in the local network (for instance192.168.0.1
orfc00::1
)\IPLib\Range\Type::T_PUBLIC
if the address is for public usage (for instance104.25.25.33
or2001:503:ba3e::2:30
)
If you want to know the type of an address range, you can use the getRangeType
method:
$range = \IPLib\Factory::parseRangeString('2000:0::1/64');
// $type will contain the value of \IPLib\Range\Type::T_PUBLIC
$type = $range->getRangeType();
// This will print Public address
echo \IPLib\Range\Type::getName($type);
Please note that if a range spans across multiple range types, you'll get NULL as the range type:
$range = \IPLib\Factory::parseRangeString('::/127');
// $type will contain null
$type = $range->getRangeType();
// This will print Unknown type
echo \IPLib\Range\Type::getName($type);
This library supports converting IPv4 to/from IPv6 addresses using the 6to4 notation or the IPv4-mapped notation:
$ipv4 = \IPLib\Factory::parseAddressString('1.2.3.4');
// 6to4 notation
$ipv6 = $ipv4->toIPv6();
// This will print 2002:102:304::
echo (string) $ipv6;
// This will print 1.2.3.4
echo $ipv6->toIPv4();
// IPv4-mapped notation
$ipv6 = $ipv4->toIPv6IPv4Mapped();
// This will print ::ffff:1.2.3.4
echo (string) $ipv6;
// This will print 1.2.3.4
echo $ipv6_6to4->toIPv4();
This library supports IPv4/IPv6 ranges in pattern format (eg. 192.168.*.*
) and in CIDR/subnet format (eg. 192.168.0.0/16
), and it offers a way to convert between the two formats:
// This will print ::*:*:*:*
echo \IPLib\Factory::parseRangeString('::/64')->asPattern()->toString();
// This will print 1:2::/96
echo \IPLib\Factory::parseRangeString('1:2::*:*')->asSubnet()->toString();
// This will print 192.168.0.0/24
echo \IPLib\Factory::parseRangeString('192.168.0.*')->asSubnet()->toString();
// This will print 10.*.*.*
echo \IPLib\Factory::parseRangeString('10.0.0.0/8')->asPattern()->toString();
Please remark that all the range types implement the asPattern()
and asSubnet()
methods.
You can use the getSubnetMask()
to get the subnet mask for IPv4 ranges:
// This will print 255.255.255.0
echo \IPLib\Factory::parseRangeString('192.168.0.*')->getSubnetMask()->toString();
// This will print 255.255.255.252
echo \IPLib\Factory::parseRangeString('192.168.0.12/30')->getSubnetMask()->toString();
You can use the getSize()
to get the count of addresses this IP range contains:
// This will print 256
echo \IPLib\Factory::parseRangeString('192.168.0.*')->getSize();
// This will print 4
echo \IPLib\Factory::parseRangeString('192.168.0.12/30')->getSize();
// This will print 1
echo \IPLib\Factory::parseRangeString('192.168.0.1')->getSize();
To perform reverse DNS queries, you need to use a special format of the IP addresses.
You can use the getReverseDNSLookupName()
method of the IP address instances to retrieve it easily:
$ipv4 = \IPLib\Factory::parseAddressString('1.2.3.255');
$ipv6 = \IPLib\Factory::parseAddressString('1234:abcd::cafe:babe');
// This will print 255.3.2.1.in-addr.arpa
echo $ipv4->getReverseDNSLookupName();
// This will print e.b.a.b.e.f.a.c.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.d.c.b.a.4.3.2.1.ip6.arpa
echo $ipv6->getReverseDNSLookupName();
To parse addresses in reverse DNS lookup format you can use the IPLib\ParseStringFlag::ADDRESS_MAYBE_RDNS
flag when parsing a string:
$ipv4 = \IPLib\Factory::parseAddressString('255.3.2.1.in-addr.arpa', \IPLib\ParseStringFlag::ADDRESS_MAYBE_RDNS);
$ipv6 = \IPLib\Factory::parseAddressString('e.b.a.b.e.f.a.c.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.0.d.c.b.a.4.3.2.1.ip6.arpa', \IPLib\ParseStringFlag::ADDRESS_MAYBE_RDNS);
// This will print 1.2.3.255
echo $ipv4->toString();
// This will print 1234:abcd::cafe:babe
echo $ipv6->toString();
You can also use getReverseDNSLookupName()
for IP ranges.
In this case, the result is an array of strings:
$range = \IPLib\Factory::parseRangeString('10.155.16.0/22');
/*
* This will print:
* array (
* 0 => '16.155.10.in-addr.arpa',
* 1 => '17.155.10.in-addr.arpa',
* 2 => '18.155.10.in-addr.arpa',
* 3 => '19.155.10.in-addr.arpa',
* )
*/
var_export($range->getReverseDNSLookupName());
This package offers a great feature: you can store address ranges in a database table, and check if an address is contained in one of the saved ranges with a simple query.
To save a range, you need to store the address type (for IPv4 it's 4
, for IPv6 it's 6
), as well as two values representing the start and the end of the range.
These methods are:
$range->getAddressType();
$range->getComparableStartString();
$range->getComparableEndString();
Let's assume that you saved the type in a field called addressType
, and the range boundaries in two fields called rangeFrom
and rangeTo
.
When you want to check if an address is within a stored range, simply use the getComparableString
method of the address and check if it's between the fields rangeFrom
and rangeTo
, and check if the stored addressType
is the same as the one of the address instance you want to check.
Here's a sample code:
/*
* Let's assume that:
* - $pdo is a PDO instance
* - $range is a range object
* - $address is an address object
*/
// Save the $range object
$insertQuery = $pdo->prepare('
insert into ranges (addressType, rangeFrom, rangeTo)
values (:addressType, :rangeFrom, :rangeTo)
');
$insertQuery->execute(array(
':addressType' => $range->getAddressType(),
':rangeFrom' => $range->getComparableStartString(),
':rangeTo' => $range->getComparableEndString(),
));
// Retrieve the saved ranges where an address $address falls:
$searchQuery = $pdo->prepare('
select * from ranges
where addressType = :addressType
and :address between rangeFrom and rangeTo
');
$searchQuery->execute(array(
':addressType' => $address->getAddressType(),
':address' => $address->getComparableString(),
));
$rows = $searchQuery->fetchAll();
$searchQuery->closeCursor();
If you want to accept addresses that may include ports, you can specify the IPLib\ParseStringFlag::MAY_INCLUDE_PORT
flag:
use IPLib\Factory;
use IPLib\ParseStringFlag;
require_once __DIR__ . '/../ip-lib.php';
// These will print NULL
var_export(Factory::parseAddressString('127.0.0.1:80'));
var_export(Factory::parseAddressString('[::]:80'));
// This will print 127.0.0.1
echo (string) Factory::parseAddressString('127.0.0.1:80', ParseStringFlag::MAY_INCLUDE_PORT);
// This will print ::
echo (string) Factory::parseAddressString('[::]:80', ParseStringFlag::MAY_INCLUDE_PORT);
If you want to accept IPv6 addresses that may include a zone ID, you can specify the IPLib\ParseStringFlag::MAY_INCLUDE_ZONEID
flag:
use IPLib\Factory;
use IPLib\ParseStringFlag;
// This will print NULL
var_export(Factory::parseAddressString('::%11'));
// This will print ::
echo (string) Factory::parseAddressString('::%11', ParseStringFlag::MAY_INCLUDE_ZONEID);
IPv4 addresses are usually expressed in decimal notation, for example as 192.168.0.1
.
By the way, the GNU (used in many Linux distros), BSD (used in Mac) and Windows implementations of inet_aton
and inet_addr
accept IPv4 addresses with numbers in octal and/or hexadecimal format.
Please remark that this does not apply to the inet_pton
and ip2long
functions, as well as to the Musl implementation (used in Alpine Linux) of inet_aton
and inet_addr
.
So, for example, these addresses are all equivalent to 192.168.0.1
:
0xC0.0xA8.0x0.0x01
(only hexadecimal)0300.0250.00.01
(only octal)192.0250.0.0x01
(decimal, octal and hexadecimal numbers)
(try it: if you browse to http://0177.0.0.0x1
, your browser will try to browse http://127.0.0.1
).
If you want to accept this non-decimal syntax, you may use the IPLib\ParseStringFlag::IPV4_MAYBE_NON_DECIMAL
flag:
use IPLib\Factory;
use IPLib\ParseStringFlag;
// This will print NULL
var_export(Factory::parseAddressString('0177.0.0.0x1'));
// This will print 127.0.0.1
var_export((string) Factory::parseAddressString('0177.0.0.0x1', ParseStringFlag::IPV4_MAYBE_NON_DECIMAL));
// This will print NULL
var_export(Factory::parseRangeString('0177.0.0.0x1/32'));
// This will print 127.0.0.1/32
var_export((string) Factory::parseRangeString('0177.0.0.0x1/32', ParseStringFlag::IPV4_MAYBE_NON_DECIMAL));
Please be aware that the IPV4_MAYBE_NON_DECIMAL
flag may also affect parsing decimal numbers:
use IPLib\Factory;
use IPLib\ParseStringFlag;
// This will print 127.0.0.10 since the last digit is assumed to be decimal
var_export((string) Factory::parseAddressString('127.0.0.010'));
// This will print 127.0.0.8 since the last digit is assumed to be octal
var_export((string) Factory::parseAddressString('127.0.0.010', ParseStringFlag::IPV4_MAYBE_NON_DECIMAL));
IPv4 addresses are usually expressed with 4 numbers, for example as 192.168.0.1
.
By the way, the GNU (used in many Linux distros), BSD (used in Mac) and Windows implementations of inet_aton
and inet_addr
accept IPv4 addresses with 1 to 4 numbers.
Please remark that this does not apply to the inet_pton
and ip2long
functions, as well as to the Musl implementation (used in Alpine Linux) of inet_aton
and inet_addr
.
If you want to accept this non-decimal syntax, you may use the IPLib\ParseStringFlag::IPV4ADDRESS_MAYBE_NON_QUAD_DOTTED
flag:
use IPLib\Factory;
use IPLib\ParseStringFlag;
// This will print NULL
var_export(Factory::parseAddressString('1.2.500'));
// This will print 0.0.0.0
var_export((string) Factory::parseAddressString('0', ParseStringFlag::IPV4ADDRESS_MAYBE_NON_QUAD_DOTTED));
// This will print 0.0.0.1
var_export((string) Factory::parseAddressString('1', ParseStringFlag::IPV4ADDRESS_MAYBE_NON_QUAD_DOTTED));
// This will print 0.0.1.244
var_export((string) Factory::parseAddressString('0.0.500', ParseStringFlag::IPV4ADDRESS_MAYBE_NON_QUAD_DOTTED));
// This will print 255.255.255.255
var_export((string) Factory::parseAddressString('4294967295', ParseStringFlag::IPV4ADDRESS_MAYBE_NON_QUAD_DOTTED));
Even if there isn't an RFC that describe it, IPv4 subnet notation may also be written in a compact form, omitting extra digits (for example, 127.0.0.0/24
may be written as 127/24
).
If you want to accept such format, you can specify the IPLib\ParseStringFlag::IPV4SUBNET_MAYBE_COMPACT
flag:
use IPLib\Factory;
use IPLib\ParseStringFlag;
// This will print NULL
var_export(Factory::parseRangeString('127/24'));
// This will print 127.0.0.0/24
echo (string) Factory::parseRangeString('127/24', ParseStringFlag::IPV4SUBNET_MAYBE_COMPACT);
Of course, you may use more than one IPLib\ParseStringFlag
flag at once:
use IPLib\Factory;
use IPLib\ParseStringFlag;
// This will print 127.0.0.255
var_export((string) Factory::parseAddressString('127.0.0.0xff:80', ParseStringFlag::MAY_INCLUDE_PORT | ParseStringFlag::IPV4_MAYBE_NON_DECIMAL));
// This will print ::
var_export((string) Factory::parseAddressString('[::%11]:80', ParseStringFlag::MAY_INCLUDE_PORT | ParseStringFlag::MAY_INCLUDE_ZONEID));
You can offer me a monthly coffee or a one-time coffee 😉