Fix uk-to-en transliteration

fre5h · Apr 12, 2022 · 7d88fa6 · 7d88fa6
1 parent 734d89a
commit 7d88fa6
Show file tree

Hide file tree

Showing 7 changed files with 226 additions and 173 deletions.
diff --git a/README.md b/README.md
@@ -36,7 +36,7 @@ PHP library for transliteration. 🇺🇦 🇬🇧 🔡 🐘
             <td>
                 Resolution of the Cabinet of Ministers of Ukraine №55 dated January 27, 2010
                 <br />
-                http://zakon1.rada.gov.ua/laws/show/55-2010-%D0%BF
+                https://zakon.rada.gov.ua/laws/show/55-2010-%D0%BF#Text
             </td>
         </tr>
     </tbody>
@@ -54,7 +54,7 @@ use Fresh\Transliteration\UkrainianToEnglish;
 
 class Foo
 {
-    public function bar($text)
+    public function bar(string $text): void
     {
         // You can use in this way
         $transliterator = new Transliterator();

diff --git a/composer.json b/composer.json
@@ -20,13 +20,13 @@
         "php": ">=8.1"
     },
     "require-dev": {
-        "friendsofphp/php-cs-fixer": "^3.3",
-        "phpstan/phpstan": "^1.2.0",
-        "phpstan/phpstan-phpunit": "^1.0.0",
+        "friendsofphp/php-cs-fixer": "^3.8",
+        "phpstan/phpstan": "^1.5",
+        "phpstan/phpstan-phpunit": "^1.1",
         "phpunit/phpunit": "^9.5",
         "slam/phpstan-extensions": "^6.0",
         "squizlabs/php_codesniffer": "^3.6",
-        "thecodingmachine/phpstan-strict-rules": "^1.0.0"
+        "thecodingmachine/phpstan-strict-rules": "^1.0"
     },
     "autoload": {
         "psr-4": {

diff --git a/phpunit.xml.dist b/phpunit.xml.dist
@@ -1,7 +1,7 @@
 <?xml version="1.0" encoding="UTF-8"?>
 
 <phpunit xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
-         xsi:noNamespaceSchemaLocation="http://schema.phpunit.de/9.5/phpunit.xsd"
+         xsi:noNamespaceSchemaLocation="https://schema.phpunit.de/9.5/phpunit.xsd"
          colors="true"
          bootstrap="vendor/autoload.php"
 >

diff --git a/src/RussianToEnglish.php b/src/RussianToEnglish.php
@@ -94,19 +94,19 @@ class RussianToEnglish implements TransliteratorInterface
     ];
 
     /**
-     * @param string $russianText
+     * @param string $textToTransliterate
      *
      * @return string
      */
-    public static function transliterate(string $russianText): string
+    public static function transliterate(string $textToTransliterate): string
     {
         $transliteratedText = '';
 
-        if ('' !== $russianText) {
+        if ('' !== $textToTransliterate) {
             $transliteratedText = \str_replace(
                 \array_keys(self::RUSSIAN_TO_ENGLISH_RULES),
                 \array_values(self::RUSSIAN_TO_ENGLISH_RULES),
-                $russianText
+                $textToTransliterate
             );
         }
 

diff --git a/src/UkrainianToEnglish.php b/src/UkrainianToEnglish.php
@@ -66,13 +66,13 @@ class UkrainianToEnglish implements TransliteratorInterface
         'ґ' => 'g',
         'д' => 'd',
         'е' => 'e',
-        'є' => 'ie',
+        'є' => 'ye',
         'ж' => 'zh',
         'з' => 'z',
         'и' => 'y',
         'і' => 'i',
-        'ї' => 'i',
-        'й' => 'i',
+        'ї' => 'yi',
+        'й' => 'y',
         'к' => 'k',
         'л' => 'l',
         'м' => 'm',
@@ -90,41 +90,76 @@ class UkrainianToEnglish implements TransliteratorInterface
         'ш' => 'sh',
         'щ' => 'shch',
         'ь' => '',
+        'ю' => 'yu',
+        'я' => 'ya',
+        '\'' => '',
+    ];
+
+    // only inside words
+    private const VOWEL_EXCEPTIONS = [
+        'є' => 'ie',
+        'ї' => 'i',
+        'й' => 'i',
         'ю' => 'iu',
         'я' => 'ia',
-        '\'' => '',
     ];
 
     /**
-     * @param string $ukrainianText
+     * @param string $textToTransliterate
      *
      * @return string
      */
-    public static function transliterate(string $ukrainianText): string
+    public static function transliterate(string $textToTransliterate): string
     {
         $transliteratedText = '';
 
-        if ('' !== $ukrainianText) {
-            if (self::checkForZghException($ukrainianText)) {
-                $ukrainianText = \str_replace(['Зг', 'зг'], ['Zgh', 'zgh'], $ukrainianText);
+        if ('' !== $textToTransliterate) {
+            if (self::checkForZghException($textToTransliterate)) {
+                $textToTransliterate = \str_replace(['Зг', 'зг'], ['Zgh', 'zgh'], $textToTransliterate);
+            }
+
+            if (1 === \preg_match('/[єюїйя]/u', $textToTransliterate)) {
+                $textToTransliterate = self::processExceptionsForVowelsInsideWords($textToTransliterate);
             }
+
             $transliteratedText = \str_replace(
                 \array_keys(self::UKRAINIAN_TO_ENGLISH_RULES),
                 \array_values(self::UKRAINIAN_TO_ENGLISH_RULES),
-                $ukrainianText
+                $textToTransliterate
             );
         }
 
         return $transliteratedText;
     }
 
     /**
-     * @param string $ukrainianText
+     * @param string $textToTransliterate
+     *
+     * @return string
+     */
+    private static function processExceptionsForVowelsInsideWords(string $textToTransliterate): string
+    {
+        $characters = mb_str_split($textToTransliterate);
+        $vowelsWithExceptions = \array_keys(self::VOWEL_EXCEPTIONS);
+        $ukrainianLetters = \array_keys(self::UKRAINIAN_TO_ENGLISH_RULES);
+
+        foreach ($characters as $i => $character) {
+            // Check that vowel inside word
+            if (0 !== $i && \in_array($character, $vowelsWithExceptions, true) && \in_array($characters[$i - 1], $ukrainianLetters, true)) {
+                $characters[$i] = self::VOWEL_EXCEPTIONS[$character];
+            }
+        }
+
+        return \implode('', $characters);
+    }
+
+    /**
+     * @param string $textToTransliterate
      *
      * @return bool
      */
-    private static function checkForZghException(string $ukrainianText): bool
+    private static function checkForZghException(string $textToTransliterate): bool
     {
-        return (bool) \mb_substr_count($ukrainianText, 'Зг') || (bool) \mb_substr_count($ukrainianText, 'зг');
+        return (bool) \mb_substr_count($textToTransliterate, 'Зг') || (bool) \mb_substr_count($textToTransliterate, 'зг');
     }
 }
diff --git a/tests/RussianToEnglishTest.php b/tests/RussianToEnglishTest.php
@@ -31,6 +31,11 @@ protected function setUp(): void
         $this->transliterator = new Transliterator();
     }
 
+    protected function tearDown(): void
+    {
+        unset($this->transliterator);
+    }
+
     /**
      * @dataProvider alphabetProvider
      */
@@ -39,7 +44,7 @@ public function testTransliterationFromRussianToEnglish(string $russianText, str
         self::assertEquals($transliteratedText, $this->transliterator->ruToEn($russianText));
     }
 
-    public function alphabetProvider(): iterable
+    public static function alphabetProvider(): iterable
     {
         yield ['А', 'A'];
         yield ['Б', 'B'];