Home » Php » security – PHP API Key Generator

security – PHP API Key Generator

Posted by: admin April 23, 2020 Leave a comment

Questions:

Does anyone know of any API key generator script/class for PHP? The class should have method generate, that would generate a key and isValid() method, to check if the key is valid.

How to&Answers:

There are multiple ways to generate API keys. I’ve used following 3 methods depending on the applications,

  1. Base62(random). Generate a large secure random number and Base-62 encode it. The key looks like “w5vt2bjzf8ryp63t”. This is good for self-provisioned system. You don’t have to worry about collision and inappropriate keys. You have to check database to know if the key is valid.

  2. Base62(MD5-HMAC(key, Normalize(referer))). This works great if the API is only allowed from one site. Just check the HMAC of the normalized referer and you know if the key is valid, no database access. You need to keep the HMAC key secret to do this.

  3. Human assigned friendly names like “example.com”. This works great if API users are required to own a domain or they are your corporate partners.

Please keep in mind that there is no security in API keys. It’s just a name assigned to your API application. More and more people are using terms like “App ID” or “Dev ID” to reflect what it really is. You have to assign another secret key if you want secure your protocol, like consumer_key/consumer_secret in OAuth.

Answer:

Here is my simple answer to this question:

$key = implode('-', str_split(substr(strtolower(md5(microtime().rand(1000, 9999))), 0, 30), 6));

Answer:

just use something like this (pseudo code) sha1(salt + time + mac-addr + another salt + some other random data) crc32 or md5 would also work inestead of sha1 and store it in a database and then isValid() checks the db if the key exists?

Answer:

Well as it has been mentioned, it is all dependant on the situation. One method that I needed to use was to authenticate a referer url with a specifically assigned API key. So with the API key all that was really needed was (pseudo) key = md5(referer url + name + salt) which you then can have a checksum for.
I know it has been mentioned similar to this before, but it is just that way. As for the isValid() function, all you need to do with this is compare it against the checksum and URL.

Edit: Just realised the age of the original question :S

Answer:

Depending on what you want, you can also use something like 3scale to create keys and manage access to the API. It generates keys, but also tracks rate limits, analytics etc. and allows devs on the API to create new keys.

There’s a PHP library as one of the connectors: https://support.3scale.net/reference/libraries

Answer:

This is an old question, but i stumbled about the same problem yesterday, and found this class which is RFC4122 compliant:

/*- * Copyright (c) 2008 Fredrik Lindberg - http://www.shapeshifter.se * All rights reserved. * * Redistribution and use in source and binary forms, with or without * modification, are permitted provided that the following conditions * are met: * 1. Redistributions of source code must retain the above copyright * notice, this list of conditions and the following disclaimer. * 2. Redistributions in binary form must reproduce the above copyright * notice, this list of conditions and the following disclaimer in the * documentation and/or other materials provided with the distribution. * * THIS SOFTWARE IS PROVIDED BY THE AUTHOR ``AS IS'' AND ANY EXPRESS OR * IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED WARRANTIES * OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE DISCLAIMED. * IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR ANY DIRECT, INDIRECT, * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT * NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, * DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY * THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF * THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. * */ /* * UUID (RFC4122) Generator * http://tools.ietf.org/html/rfc4122 * * Implements version 1, 3, 4 and 5 */ class GEN_UUID { /* UUID versions */ const UUID_TIME = 1; /* Time based UUID */ const UUID_NAME_MD5 = 3; /* Name based (MD5) UUID */ const UUID_RANDOM = 4; /* Random UUID */ const UUID_NAME_SHA1 = 5; /* Name based (SHA1) UUID */ /* UUID formats */ const FMT_FIELD = 100; const FMT_STRING = 101; const FMT_BINARY = 102; const FMT_QWORD = 1; /* Quad-word, 128-bit (not impl.) */ const FMT_DWORD = 2; /* Double-word, 64-bit (not impl.) */ const FMT_WORD = 4; /* Word, 32-bit (not impl.) */ const FMT_SHORT = 8; /* Short (not impl.) */ const FMT_BYTE = 16; /* Byte */ const FMT_DEFAULT = 16; public function __construct() { } /* Field UUID representation */ static private $m_uuid_field = array( 'time_low' => 0, /* 32-bit */ 'time_mid' => 0, /* 16-bit */ 'time_hi' => 0, /* 16-bit */ 'clock_seq_hi' => 0, /* 8-bit */ 'clock_seq_low' => 0, /* 8-bit */ 'node' => array() /* 48-bit */ ); static private $m_generate = array( self::UUID_TIME => "generateTime", self::UUID_RANDOM => "generateRandom", self::UUID_NAME_MD5 => "generateNameMD5", self::UUID_NAME_SHA1 => "generateNameSHA1" ); static private $m_convert = array( self::FMT_FIELD => array( self::FMT_BYTE => "conv_field2byte", self::FMT_STRING => "conv_field2string", self::FMT_BINARY => "conv_field2binary" ), self::FMT_BYTE => array( self::FMT_FIELD => "conv_byte2field", self::FMT_STRING => "conv_byte2string", self::FMT_BINARY => "conv_byte2binary" ), self::FMT_STRING => array( self::FMT_BYTE => "conv_string2byte", self::FMT_FIELD => "conv_string2field", self::FMT_BINARY => "conv_string2binary" ), ); /* Swap byte order of a 32-bit number */ static private function swap32($x) { return (($x & 0x000000ff) << 24) | (($x & 0x0000ff00) << 8) | (($x & 0x00ff0000) >> 8) | (($x & 0xff000000) >> 24); } /* Swap byte order of a 16-bit number */ static private function swap16($x) { return (($x & 0x00ff) << 8) | (($x & 0xff00) >> 8); } /* Auto-detect UUID format */ static private function detectFormat($src) { if (is_string($src)) return self::FMT_STRING; else if (is_array($src)) { $len = count($src); if ($len == 1 || ($len % 2) == 0) return $len; else return (-1); } else return self::FMT_BINARY; } /* * Public API, generate a UUID of 'type' in format 'fmt' for * the given namespace 'ns' and node 'node' */ static public function generate($type=self::UUID_RANDOM, $fmt = self::FMT_STRING, $node = "", $ns = "") { $func = self::$m_generate[$type]; if (!isset($func)) return null; $conv = self::$m_convert[self::FMT_FIELD][$fmt]; $uuid = self::$func($ns, $node); return self::$conv($uuid); } /* * Public API, convert a UUID from one format to another */ static public function convert($uuid, $from, $to) { $conv = self::$m_convert[$from][$to]; if (!isset($conv)) return ($uuid); return (self::$conv($uuid)); } /* * Generate an UUID version 4 (pseudo random) */ static private function generateRandom($ns, $node) { $uuid = self::$m_uuid_field; $uuid['time_hi'] = (4 << 12) | (mt_rand(0, 0x1000)); $uuid['clock_seq_hi'] = (1 << 7) | mt_rand(0, 128); $uuid['time_low'] = mt_rand(0, 0xffff) + (mt_rand(0, 0xffff) << 16); $uuid['time_mid'] = mt_rand(0, 0xffff); $uuid['clock_seq_low'] = mt_rand(0, 255); for ($i = 0; $i < 6; $i++) $uuid['node'][$i] = mt_rand(0, 255); return ($uuid); } /* * Generate UUID version 3 and 5 (name based) */ static private function generateName($ns, $node, $hash, $version) { $ns_fmt = self::detectFormat($ns); $field = self::convert($ns, $ns_fmt, self::FMT_FIELD); /* Swap byte order to keep it in big endian on all platforms */ $field['time_low'] = self::swap32($field['time_low']); $field['time_mid'] = self::swap16($field['time_mid']); $field['time_hi'] = self::swap16($field['time_hi']); /* Convert the namespace to binary and concatenate node */ $raw = self::convert($field, self::FMT_FIELD, self::FMT_BINARY); $raw .= $node; /* Hash the namespace and node and convert to a byte array */ $val = $hash($raw, true); $tmp = unpack('C16', $val); foreach (array_keys($tmp) as $key) $byte[$key - 1] = $tmp[$key]; /* Convert byte array to a field array */ $field = self::conv_byte2field($byte); $field['time_low'] = self::swap32($field['time_low']); $field['time_mid'] = self::swap16($field['time_mid']); $field['time_hi'] = self::swap16($field['time_hi']); /* Apply version and constants */ $field['clock_seq_hi'] &= 0x3f; $field['clock_seq_hi'] |= (1 << 7); $field['time_hi'] &= 0x0fff; $field['time_hi'] |= ($version << 12); return ($field); } static private function generateNameMD5($ns, $node) { return self::generateName($ns, $node, "md5", self::UUID_NAME_MD5); } static private function generateNameSHA1($ns, $node) { return self::generateName($ns, $node, "sha1", self::UUID_NAME_SHA1); } /* * Generate UUID version 1 (time based) */ static private function generateTime($ns, $node) { $uuid = self::$m_uuid_field; /* * Get current time in 100 ns intervals. The magic value * is the offset between UNIX epoch and the UUID UTC * time base October 15, 1582. */ $tp = gettimeofday(); $time = ($tp['sec'] * 10000000) + ($tp['usec'] * 10) + 0x01B21DD213814000; $uuid['time_low'] = $time & 0xffffffff; /* Work around PHP 32-bit bit-operation limits */ $high = intval($time / 0xffffffff); $uuid['time_mid'] = $high & 0xffff; $uuid['time_hi'] = (($high >> 16) & 0xfff) | (self::UUID_TIME << 12); /* * We don't support saved state information and generate * a random clock sequence each time. */ $uuid['clock_seq_hi'] = 0x80 | mt_rand(0, 64); $uuid['clock_seq_low'] = mt_rand(0, 255); /* * Node should be set to the 48-bit IEEE node identifier, but * we leave it for the user to supply the node. */ for ($i = 0; $i < 6; $i++) $uuid['node'][$i] = ord(substr($node, $i, 1)); return ($uuid); } /* Assumes correct byte order */ static private function conv_field2byte($src) { $uuid[0] = ($src['time_low'] & 0xff000000) >> 24; $uuid[1] = ($src['time_low'] & 0x00ff0000) >> 16; $uuid[2] = ($src['time_low'] & 0x0000ff00) >> 8; $uuid[3] = ($src['time_low'] & 0x000000ff); $uuid[4] = ($src['time_mid'] & 0xff00) >> 8; $uuid[5] = ($src['time_mid'] & 0x00ff); $uuid[6] = ($src['time_hi'] & 0xff00) >> 8; $uuid[7] = ($src['time_hi'] & 0x00ff); $uuid[8] = $src['clock_seq_hi']; $uuid[9] = $src['clock_seq_low']; for ($i = 0; $i < 6; $i++) $uuid[10+$i] = $src['node'][$i]; return ($uuid); } static private function conv_field2string($src) { $str = sprintf( '%08x-%04x-%04x-%02x%02x-%02x%02x%02x%02x%02x%02x', ($src['time_low']), ($src['time_mid']), ($src['time_hi']), $src['clock_seq_hi'], $src['clock_seq_low'], $src['node'][0], $src['node'][1], $src['node'][2], $src['node'][3], $src['node'][4], $src['node'][5]); return ($str); } static private function conv_field2binary($src) { $byte = self::conv_field2byte($src); return self::conv_byte2binary($byte); } static private function conv_byte2field($uuid) { $field = self::$m_uuid_field; $field['time_low'] = ($uuid[0] << 24) | ($uuid[1] << 16) | ($uuid[2] << 8) | $uuid[3]; $field['time_mid'] = ($uuid[4] << 8) | $uuid[5]; $field['time_hi'] = ($uuid[6] << 8) | $uuid[7]; $field['clock_seq_hi'] = $uuid[8]; $field['clock_seq_low'] = $uuid[9]; for ($i = 0; $i < 6; $i++) $field['node'][$i] = $uuid[10+$i]; return ($field); } static public function conv_byte2string($src) { $field = self::conv_byte2field($src); return self::conv_field2string($field); } static private function conv_byte2binary($src) { $raw = pack('C16', $src[0], $src[1], $src[2], $src[3], $src[4], $src[5], $src[6], $src[7], $src[8], $src[9], $src[10], $src[11], $src[12], $src[13], $src[14], $src[15]); return ($raw); } static private function conv_string2field($src) { $parts = sscanf($src, '%x-%x-%x-%x-%02x%02x%02x%02x%02x%02x'); $field = self::$m_uuid_field; $field['time_low'] = ($parts[0]); $field['time_mid'] = ($parts[1]); $field['time_hi'] = ($parts[2]); $field['clock_seq_hi'] = ($parts[3] & 0xff00) >> 8; $field['clock_seq_low'] = $parts[3] & 0x00ff; for ($i = 0; $i < 6; $i++) $field['node'][$i] = $parts[4+$i]; return ($field); } static private function conv_string2byte($src) { $field = self::conv_string2field($src); return self::conv_field2byte($field); } static private function conv_string2binary($src) { $byte = self::conv_string2byte($src); return self::conv_byte2binary($byte); } } 

Hope that helps

Answer:

See uniqid:

string uniqid ([ string $prefix = "" [, bool $more_entropy = false ]] ) 

Answer:

You can just use md5(uniqid()) and divide it into parts or format in other preferable way.

Answer:

GUID would work but is not cryptographically secure.

Server answers use md5 or sha1 hashing methods on microtime() or mt_rand.

Hashing a uniqid, uuid or timestamp would not necessarily create unique results!
Actually hashing increases the chance of collisions so I would strongly advise against this.