<?php

/*
 * sample-utf8.php: UTF-8 test file for PHP.
 * Free sample from 99helpers.com/file-types/php
 * Run: php sample-utf8.php
 * If any line below looks garbled (Ã©, â€™), something read this file in the
 * wrong encoding. PHP itself doesn't decode source files: it copies the bytes
 * of strings as they are, so save PHP files as UTF-8 without a BOM.
 */

declare(strict_types=1);

$greetings = [
    'Dutch'    => 'Goedemorgen, café-eigenaar',
    'German'   => 'Grüß Gott, Straße',
    'French'   => 'Ça va, Zoë ? Très bien',
    'Greek'    => 'Καλημέρα',
    'Russian'  => 'Здравствуйте',
    'Arabic'   => 'مرحبا',
    'Hebrew'   => 'שלום',
    'Hindi'    => 'नमस्ते',
    'Japanese' => 'こんにちは',
    'Chinese'  => '你好',
    'Korean'   => '안녕하세요',
    'Symbols'  => '€ £ ¥ ₹ ≤ ≥ ∞ ✓',
    'Emoji'    => '👋🌍🐘',
];

// 1. strlen() counts bytes; mb_strlen() (the mbstring extension) counts characters.
foreach ($greetings as $language => $text) {
    printf("%-9s %s  (%d characters, %d bytes)\n", $language, $text, mb_strlen($text), strlen($text));
}

// 2. strtoupper() only changes ASCII letters; mb_strtoupper() handles the rest.
echo strtoupper('café'), ' / ', mb_strtoupper('café'), "\n";

// 3. Byte functions can cut a character in half; mb_substr() can't.
echo bin2hex(substr('é', 0, 1)), ' / ', mb_substr('Ελληνικά', 0, 4), "\n";

// 4. json_encode() escapes non-ASCII as \\uXXXX unless told not to.
echo json_encode(['city' => 'Zürich']), ' / ', json_encode(['city' => 'Zürich'], JSON_UNESCAPED_UNICODE), "\n";

// 5. Normalisation needs the intl extension, which not every host has.
if (class_exists(Normalizer::class)) {
    $decomposed = "e\u{0301}";
    var_dump($decomposed === 'é', Normalizer::normalize($decomposed) === 'é');
} else {
    echo "intl extension not installed: skipping normalisation\n";
}