ircmaxell

<?php

declare(ticks=1);

$__stack = [];

function getArrayKey(array $haystack, $needle) {
    return isset($haystack[$needle]) ? $haystack[$needle] : '';
}

ircmaxell

<?php

class HashCipher {
    protected $hash = 'sha512';
    protected $hashLen = 64;
    protected $rounds = 16;

    public function __construct($hash, $rounds = 16) {
        $this->hash = $hash;
        // run a test hash, to get the proper hash length

ircmaxell

Keybase proof

I hereby claim:

• I am ircmaxell on github.
• I am ircmaxell (https://keybase.io/ircmaxell) on keybase.
• I have a public key whose fingerprint is D620 2127 D88F 66C2 10A3 BFEB 6A7C A4E4 A190 DC6B

To claim this, I am signing this object:

ircmaxell

<?hh

class B {
    public function doB() : string {
        return "b";
    }
}

class C extends B {
    public function doC() : string {

ircmaxell

<?php

class a {
    const a = b::c;
    const b = b::d;
}

ircmaxell

interface Foo1 {
    public function bar()
}

interface Foo2 {
    public function bar()
}

interface Foo3 {
    public function bar()

ircmaxell

<?php

function ackersum($a, $b) {
    if ($a == 0) {
        return $b;
    } elseif ($b == 0) {
        return $a;
    } elseif ($a == 1) {
        return $b + 1;
    }

ircmaxell

<?php

define('CRYPT_SALT_LEN_MAX', 16);
define('CRYPT_ROUNDS_DEFAULT', 5000);

/**
 * A PHP shim for crypt sha256 and sha512
 * to workaround bug 62372
 * @see https://bugs.php.net/bug.php?id=62372
 */

ircmaxell

13:11 <ajf> I’m mostly happy with it as-is, but given that I made int/float/numeric/string not accept NULL, I think booleans shouldn’t be accepted for them either.
13:19  → arty joined  ⇐ DaveRand quit  
13:49  → sonotos and lonnylot joined  ⇐ dmitry and RemiFedor quit  
14:41 <ircmaxell> ajf: that sounds sane (consistency wise). So it means that `int/float/numeric` only accept int/float/valid-number-string, and string only accepts string/int/float
14:43 <ajf> ircmaxell: Yeah
14:43 <ajf> Also, we don’t accept int/float/string for bool, so we shouldn’t accept bool for int/float/string
14:43 <ircmaxell> yeah, makes perfect sense
14:44 <ajf> Another point someone raised was that FALSE is a common error value alongside NULL.
14:44 <ajf> So, should I go ahead and amend the patch and RFC to do that?
14:44 <ircmaxell> I think that's the direction we're going, so that's valid

ircmaxell

<?php

/**
 * @param int
 * @param int
 * @return int
 */
function foo($a, $b) {
    $c = 0;
    if ($a) {