norse-blue/php-scalar-objects

View on GitHub
src/Support/Character.php

Summary

Maintainability
C
1 day
Test Coverage
<?php

declare(strict_types=1);

namespace NorseBlue\ScalarObjects\Support;

/**
 * Class Character
 *
 * @package NorseBlue\Prim\Characters
 */
abstract class Character
{
    /**
     * Returns the replacements for the ascii method.
     *
     * Note: Adapted from Stringy\Stringy.
     *
     * @see https://github.com/danielstjules/Stringy/blob/3.1.0/LICENSE.txt
     *
     * @return array<int|string, array<string>>
     */
    public static function charsArray(): array
    {
        static $charsArray;

        if (isset($charsArray)) {
            return $charsArray;
        }

        return $charsArray = [
            '0' => ['°', '₀', '۰', '0'],
            '1' => ['¹', '₁', '۱', '1'],
            '2' => ['²', '₂', '۲', '2'],
            '3' => ['³', '₃', '۳', '3'],
            '4' => ['⁴', '₄', '۴', '٤', '4'],
            '5' => ['⁵', '₅', '۵', '٥', '5'],
            '6' => ['⁶', '₆', '۶', '٦', '6'],
            '7' => ['⁷', '₇', '۷', '7'],
            '8' => ['⁸', '₈', '۸', '8'],
            '9' => ['⁹', '₉', '۹', '9'],
            'a' => [
                'à',
                'á',
                'ả',
                'ã',
                'ạ',
                'ă',
                'ắ',
                'ằ',
                'ẳ',
                'ẵ',
                'ặ',
                'â',
                'ấ',
                'ầ',
                'ẩ',
                'ẫ',
                'ậ',
                'ā',
                'ą',
                'å',
                'α',
                'ά',
                'ἀ',
                'ἁ',
                'ἂ',
                'ἃ',
                'ἄ',
                'ἅ',
                'ἆ',
                'ἇ',
                'ᾀ',
                'ᾁ',
                'ᾂ',
                'ᾃ',
                'ᾄ',
                'ᾅ',
                'ᾆ',
                'ᾇ',
                'ὰ',
                'ά',
                'ᾰ',
                'ᾱ',
                'ᾲ',
                'ᾳ',
                'ᾴ',
                'ᾶ',
                'ᾷ',
                'а',
                'أ',
                'အ',
                'ာ',
                'ါ',
                'ǻ',
                'ǎ',
                'ª',
                'ა',
                'अ',
                'ا',
                'a',
                'ä',
            ],
            'b' => ['б', 'β', 'ب', 'ဗ', 'ბ', 'b'],
            'c' => ['ç', 'ć', 'č', 'ĉ', 'ċ', 'c'],
            'd' => ['ď', 'ð', 'đ', 'ƌ', 'ȡ', 'ɖ', 'ɗ', 'ᵭ', 'ᶁ', 'ᶑ', 'д', 'δ', 'د', 'ض', 'ဍ', 'ဒ', 'დ', 'd'],
            'e' => [
                'é',
                'è',
                'ẻ',
                'ẽ',
                'ẹ',
                'ê',
                'ế',
                'ề',
                'ể',
                'ễ',
                'ệ',
                'ë',
                'ē',
                'ę',
                'ě',
                'ĕ',
                'ė',
                'ε',
                'έ',
                'ἐ',
                'ἑ',
                'ἒ',
                'ἓ',
                'ἔ',
                'ἕ',
                'ὲ',
                'έ',
                'е',
                'ё',
                'э',
                'є',
                'ə',
                'ဧ',
                'ေ',
                'ဲ',
                'ე',
                'ए',
                'إ',
                'ئ',
                'e',
            ],
            'f' => ['ф', 'φ', 'ف', 'ƒ', 'ფ', 'f'],
            'g' => ['ĝ', 'ğ', 'ġ', 'ģ', 'г', 'ґ', 'γ', 'ဂ', 'გ', 'گ', 'g'],
            'h' => ['ĥ', 'ħ', 'η', 'ή', 'ح', 'ه', 'ဟ', 'ှ', 'ჰ', 'h'],
            'i' => [
                'í',
                'ì',
                'ỉ',
                'ĩ',
                'ị',
                'î',
                'ï',
                'ī',
                'ĭ',
                'į',
                'ı',
                'ι',
                'ί',
                'ϊ',
                'ΐ',
                'ἰ',
                'ἱ',
                'ἲ',
                'ἳ',
                'ἴ',
                'ἵ',
                'ἶ',
                'ἷ',
                'ὶ',
                'ί',
                'ῐ',
                'ῑ',
                'ῒ',
                'ΐ',
                'ῖ',
                'ῗ',
                'і',
                'ї',
                'и',
                'ဣ',
                'ိ',
                'ီ',
                'ည်',
                'ǐ',
                'ი',
                'इ',
                'ی',
                'i',
            ],
            'j' => ['ĵ', 'ј', 'Ј', 'ჯ', 'ج', 'j'],
            'k' => ['ķ', 'ĸ', 'к', 'κ', 'Ķ', 'ق', 'ك', 'က', 'კ', 'ქ', 'ک', 'k'],
            'l' => ['ł', 'ľ', 'ĺ', 'ļ', 'ŀ', 'л', 'λ', 'ل', 'လ', 'ლ', 'l'],
            'm' => ['м', 'μ', 'م', 'မ', 'მ', 'm'],
            'n' => ['ñ', 'ń', 'ň', 'ņ', 'ʼn', 'ŋ', 'ν', 'н', 'ن', 'န', 'ნ', 'n'],
            'o' => [
                'ó',
                'ò',
                'ỏ',
                'õ',
                'ọ',
                'ô',
                'ố',
                'ồ',
                'ổ',
                'ỗ',
                'ộ',
                'ơ',
                'ớ',
                'ờ',
                'ở',
                'ỡ',
                'ợ',
                'ø',
                'ō',
                'ő',
                'ŏ',
                'ο',
                'ὀ',
                'ὁ',
                'ὂ',
                'ὃ',
                'ὄ',
                'ὅ',
                'ὸ',
                'ό',
                'о',
                'و',
                'θ',
                'ို',
                'ǒ',
                'ǿ',
                'º',
                'ო',
                'ओ',
                'o',
                'ö',
            ],
            'p' => ['п', 'π', 'ပ', 'პ', 'پ', 'p'],
            'q' => ['ყ', 'q'],
            'r' => ['ŕ', 'ř', 'ŗ', 'р', 'ρ', 'ر', 'რ', 'r'],
            's' => ['ś', 'š', 'ş', 'с', 'σ', 'ș', 'ς', 'س', 'ص', 'စ', 'ſ', 'ს', 's'],
            't' => ['ť', 'ţ', 'т', 'τ', 'ț', 'ت', 'ط', 'ဋ', 'တ', 'ŧ', 'თ', 'ტ', 't'],
            'u' => [
                'ú',
                'ù',
                'ủ',
                'ũ',
                'ụ',
                'ư',
                'ứ',
                'ừ',
                'ử',
                'ữ',
                'ự',
                'û',
                'ū',
                'ů',
                'ű',
                'ŭ',
                'ų',
                'µ',
                'у',
                'ဉ',
                'ု',
                'ူ',
                'ǔ',
                'ǖ',
                'ǘ',
                'ǚ',
                'ǜ',
                'უ',
                'उ',
                'u',
                'ў',
                'ü',
            ],
            'v' => ['в', 'ვ', 'ϐ', 'v'],
            'w' => ['ŵ', 'ω', 'ώ', 'ဝ', 'ွ', 'w'],
            'x' => ['χ', 'ξ', 'x'],
            'y' => ['ý', 'ỳ', 'ỷ', 'ỹ', 'ỵ', 'ÿ', 'ŷ', 'й', 'ы', 'υ', 'ϋ', 'ύ', 'ΰ', 'ي', 'ယ', 'y'],
            'z' => ['ź', 'ž', 'ż', 'з', 'ζ', 'ز', 'ဇ', 'ზ', 'z'],
            'aa' => ['ع', 'आ', 'آ'],
            'ae' => ['æ', 'ǽ'],
            'ai' => ['ऐ'],
            'ch' => ['ч', 'ჩ', 'ჭ', 'چ'],
            'dj' => ['ђ', 'đ'],
            'dz' => ['џ', 'ძ'],
            'ei' => ['ऍ'],
            'gh' => ['غ', 'ღ'],
            'ii' => ['ई'],
            'ij' => ['ij'],
            'kh' => ['х', 'خ', 'ხ'],
            'lj' => ['љ'],
            'nj' => ['њ'],
            'oe' => ['ö', 'œ', 'ؤ'],
            'oi' => ['ऑ'],
            'oii' => ['ऒ'],
            'ps' => ['ψ'],
            'sh' => ['ш', 'შ', 'ش'],
            'shch' => ['щ'],
            'ss' => ['ß'],
            'sx' => ['ŝ'],
            'th' => ['þ', 'ϑ', 'ث', 'ذ', 'ظ'],
            'ts' => ['ц', 'ც', 'წ'],
            'ue' => ['ü'],
            'uu' => ['ऊ'],
            'ya' => ['я'],
            'yu' => ['ю'],
            'zh' => ['ж', 'ჟ', 'ژ'],
            '(c)' => ['©'],
            'A' => [
                'Á',
                'À',
                'Ả',
                'Ã',
                'Ạ',
                'Ă',
                'Ắ',
                'Ằ',
                'Ẳ',
                'Ẵ',
                'Ặ',
                'Â',
                'Ấ',
                'Ầ',
                'Ẩ',
                'Ẫ',
                'Ậ',
                'Å',
                'Ā',
                'Ą',
                'Α',
                'Ά',
                'Ἀ',
                'Ἁ',
                'Ἂ',
                'Ἃ',
                'Ἄ',
                'Ἅ',
                'Ἆ',
                'Ἇ',
                'ᾈ',
                'ᾉ',
                'ᾊ',
                'ᾋ',
                'ᾌ',
                'ᾍ',
                'ᾎ',
                'ᾏ',
                'Ᾰ',
                'Ᾱ',
                'Ὰ',
                'Ά',
                'ᾼ',
                'А',
                'Ǻ',
                'Ǎ',
                'A',
                'Ä',
            ],
            'B' => ['Б', 'Β', 'ब', 'B'],
            'C' => ['Ç', 'Ć', 'Č', 'Ĉ', 'Ċ', 'C'],
            'D' => ['Ď', 'Ð', 'Đ', 'Ɖ', 'Ɗ', 'Ƌ', 'ᴅ', 'ᴆ', 'Д', 'Δ', 'D'],
            'E' => [
                'É',
                'È',
                'Ẻ',
                'Ẽ',
                'Ẹ',
                'Ê',
                'Ế',
                'Ề',
                'Ể',
                'Ễ',
                'Ệ',
                'Ë',
                'Ē',
                'Ę',
                'Ě',
                'Ĕ',
                'Ė',
                'Ε',
                'Έ',
                'Ἐ',
                'Ἑ',
                'Ἒ',
                'Ἓ',
                'Ἔ',
                'Ἕ',
                'Έ',
                'Ὲ',
                'Е',
                'Ё',
                'Э',
                'Є',
                'Ə',
                'E',
            ],
            'F' => ['Ф', 'Φ', 'F'],
            'G' => ['Ğ', 'Ġ', 'Ģ', 'Г', 'Ґ', 'Γ', 'G'],
            'H' => ['Η', 'Ή', 'Ħ', 'H'],
            'I' => [
                'Í',
                'Ì',
                'Ỉ',
                'Ĩ',
                'Ị',
                'Î',
                'Ï',
                'Ī',
                'Ĭ',
                'Į',
                'İ',
                'Ι',
                'Ί',
                'Ϊ',
                'Ἰ',
                'Ἱ',
                'Ἳ',
                'Ἴ',
                'Ἵ',
                'Ἶ',
                'Ἷ',
                'Ῐ',
                'Ῑ',
                'Ὶ',
                'Ί',
                'И',
                'І',
                'Ї',
                'Ǐ',
                'ϒ',
                'I',
            ],
            'J' => ['J'],
            'K' => ['К', 'Κ', 'K'],
            'L' => ['Ĺ', 'Ł', 'Л', 'Λ', 'Ļ', 'Ľ', 'Ŀ', 'ल', 'L'],
            'M' => ['М', 'Μ', 'M'],
            'N' => ['Ń', 'Ñ', 'Ň', 'Ņ', 'Ŋ', 'Н', 'Ν', 'N'],
            'O' => [
                'Ó',
                'Ò',
                'Ỏ',
                'Õ',
                'Ọ',
                'Ô',
                'Ố',
                'Ồ',
                'Ổ',
                'Ỗ',
                'Ộ',
                'Ơ',
                'Ớ',
                'Ờ',
                'Ở',
                'Ỡ',
                'Ợ',
                'Ø',
                'Ō',
                'Ő',
                'Ŏ',
                'Ο',
                'Ό',
                'Ὀ',
                'Ὁ',
                'Ὂ',
                'Ὃ',
                'Ὄ',
                'Ὅ',
                'Ὸ',
                'Ό',
                'О',
                'Θ',
                'Ө',
                'Ǒ',
                'Ǿ',
                'O',
                'Ö',
            ],
            'P' => ['П', 'Π', 'P'],
            'Q' => ['Q'],
            'R' => ['Ř', 'Ŕ', 'Р', 'Ρ', 'Ŗ', 'R'],
            'S' => ['Ş', 'Ŝ', 'Ș', 'Š', 'Ś', 'С', 'Σ', 'S'],
            'T' => ['Ť', 'Ţ', 'Ŧ', 'Ț', 'Т', 'Τ', 'T'],
            'U' => [
                'Ú',
                'Ù',
                'Ủ',
                'Ũ',
                'Ụ',
                'Ư',
                'Ứ',
                'Ừ',
                'Ử',
                'Ữ',
                'Ự',
                'Û',
                'Ū',
                'Ů',
                'Ű',
                'Ŭ',
                'Ų',
                'У',
                'Ǔ',
                'Ǖ',
                'Ǘ',
                'Ǚ',
                'Ǜ',
                'U',
                'Ў',
                'Ü',
            ],
            'V' => ['В', 'V'],
            'W' => ['Ω', 'Ώ', 'Ŵ', 'W'],
            'X' => ['Χ', 'Ξ', 'X'],
            'Y' => ['Ý', 'Ỳ', 'Ỷ', 'Ỹ', 'Ỵ', 'Ÿ', 'Ῠ', 'Ῡ', 'Ὺ', 'Ύ', 'Ы', 'Й', 'Υ', 'Ϋ', 'Ŷ', 'Y'],
            'Z' => ['Ź', 'Ž', 'Ż', 'З', 'Ζ', 'Z'],
            'AE' => ['Æ', 'Ǽ'],
            'Ch' => ['Ч'],
            'Dj' => ['Ђ'],
            'Dz' => ['Џ'],
            'Gx' => ['Ĝ'],
            'Hx' => ['Ĥ'],
            'Ij' => ['IJ'],
            'Jx' => ['Ĵ'],
            'Kh' => ['Х'],
            'Lj' => ['Љ'],
            'Nj' => ['Њ'],
            'Oe' => ['Œ'],
            'Ps' => ['Ψ'],
            'Sh' => ['Ш'],
            'Shch' => ['Щ'],
            'Ss' => ['ẞ'],
            'Th' => ['Þ'],
            'Ts' => ['Ц'],
            'Ya' => ['Я'],
            'Yu' => ['Ю'],
            'Zh' => ['Ж'],
            ' ' => [
                "\xC2\xA0",
                "\xE2\x80\x80",
                "\xE2\x80\x81",
                "\xE2\x80\x82",
                "\xE2\x80\x83",
                "\xE2\x80\x84",
                "\xE2\x80\x85",
                "\xE2\x80\x86",
                "\xE2\x80\x87",
                "\xE2\x80\x88",
                "\xE2\x80\x89",
                "\xE2\x80\x8A",
                "\xE2\x80\xAF",
                "\xE2\x81\x9F",
                "\xE3\x80\x80",
                "\xEF\xBE\xA0",
            ],
        ];
    }

    /**
     * Returns the language specific replacements for the ascii method.
     *
     * Note: Adapted from Stringy\Stringy.
     *
     * @see https://github.com/danielstjules/Stringy/blob/3.1.0/LICENSE.txt
     *
     * @return array<int, string>|null
     */
    public static function languageSpecificCharsArray(?string $language): ?array
    {
        static $languageSpecific;

        if (! isset($languageSpecific)) {
            $languageSpecific = [
                'bg' => [
                    ['х', 'Х', 'щ', 'Щ', 'ъ', 'Ъ', 'ь', 'Ь'],
                    ['h', 'H', 'sht', 'SHT', 'a', 'А', 'y', 'Y'],
                ],
                'da' => [
                    ['æ', 'ø', 'å', 'Æ', 'Ø', 'Å'],
                    ['ae', 'oe', 'aa', 'Ae', 'Oe', 'Aa'],
                ],
                'de' => [
                    ['ä', 'ö', 'ü', 'Ä', 'Ö', 'Ü'],
                    ['ae', 'oe', 'ue', 'AE', 'OE', 'UE'],
                ],
                'ro' => [
                    ['ă', 'â', 'î', 'ș', 'ț', 'Ă', 'Â', 'Î', 'Ș', 'Ț'],
                    ['a', 'a', 'i', 's', 't', 'A', 'A', 'I', 'S', 'T'],
                ],
            ];
        }

        return $languageSpecific[$language] ?? null;
    }
}