Test cases for extended grapheme cluster

ret.php

array(3) {
  ["all cases: "]=>
  int(348)
  ["skipped cases: "]=>
  int(54)
  ["not pass: "]=>
  int(17)
}

test_pcre_extendted_grapheme_cluster.php

<?php

// https://gist.github.com/masakielastic/c3341a74aaffa941f518
require 'GraphemeBreakTest.php';

$count = count($input);
$skipped_cases = 0;
$passed_cases = 0;

for ($i = 0; $i < $count; ++$i) {

    if (contains_surrogate($input[$i])) {
        ++$skipped_cases;
        continue;
    }

    $str = generate_str_from_array($input[$i]);
    $length = count($expected[$i]);
    $input_length = grapheme_length($str);
    if ($length !== $input_length) {
        var_dump([
            'input: ' => $input[$i],
            'expected: ' => $expected[$i],
            'counting by \X: ' =>  $input_length,
        ]);
        ++$passed_cases;
    }
}

var_dump([
    'all cases: ' => count($input) - $skipped_cases,
    'skipped cases: ' => $skipped_cases,
    'not pass: ' => $passed_cases
]);

function grapheme_length($str)
{
    return preg_match_all('/\X/u', $str);
}

function contains_surrogate($array) {

    foreach ($array as $v) {

        if (0xD800 <= $v && $v <= 0xDFFF) {
            return true;
        }

     
    }

     return false;
}

function generate_str_from_array($array)
{
    $ret = '';

    foreach ($array as $a) {
        $ret .= utf8_chr($a);
    }

    return $ret;
}

function utf8_chr($cp) {

    if (!is_int($cp)) {
        exit("$cp is not integer\n");
    }

    if ($cp < 0 || (0xD7FF < $cp && $cp < 0xE000) || 0x10FFFF < $cp) {
        exit("$cp is out of range\n");
    }

    if ($cp < 0x80) {
        return chr($cp);
    } else if ($cp < 0xA0) {
        return chr(0x1C0 | $cp >> 6).chr(0x80 | $cp & 0x3F);
    }

    return html_entity_decode('&#'.$cp.';');
}