Now formatting is kept through sentences when breaking a paragraph.

Author: parpalak

src/S2/Rose/Helper/StringHelper.php

@@ -45,15 +45,19 @@ public static function sentencesFromText(string $text, bool $hasFormatting): arr
         $substrings = str_replace("�", ' ', $substrings);
 
         if ($hasFormatting) {
-            // We keep the formatting scope within a single sentence.
+            // We keep the formatting scope through several sentences.
             //
             // For example, consider the input: 'Sentence <i>1. Sentence 2. Sentence</i> 3.'
-            // After processing, it becomes ['Sentence <i>1.</i>', 'Sentence 2.', '<i>Sentence</i> 3.'].
-            //
-            // This approach is reasonable because individual sentences are typically joined into snippets,
-            // and preserving formatting across multiple sentences may not be meaningful.
-            array_walk($substrings, static function (string &$text) {
-                $text = self::fixUnbalancedInternalFormatting($text);
+            // After processing, it becomes ['Sentence <i>1.</i>', '<i>Sentence 2.</i>', '<i>Sentence</i> 3.'].
+            $tagsFromPrevSentence = [];
+            array_walk($substrings, static function (string &$text) use (&$tagsFromPrevSentence) {
+                foreach (array_reverse($tagsFromPrevSentence) as $possibleTag => $num) {
+                    if ($num > 0) {
+                        $text = str_repeat('\\' . $possibleTag, $num) . $text;
+                        $tagsFromPrevSentence[$possibleTag] = 0;
+                    }
+                }
+                $text = self::fixUnbalancedInternalFormatting($text, $tagsFromPrevSentence);
             });
         }
 
@@ -101,11 +105,11 @@ public static function clearInternalFormatting(string $text): string
         ]);
     }
 
-    public static function fixUnbalancedInternalFormatting(string $text): string
+    public static function fixUnbalancedInternalFormatting(string $text, array &$tagsNum): string
     {
         preg_match_all('#\\\\([' . self::FORMATTING_SYMBOLS . '])#i', $text, $matches);
 
-        $tagsNum = [];
+//        $tagsNum = [];
         foreach ($matches[1] as $match) {
             $lowerMatch           = strtolower($match);
             $tagsNum[$lowerMatch] = ($tagsNum[$lowerMatch] ?? 0) + ($match === $lowerMatch ? 1 : -1);
@@ -118,8 +122,7 @@ public static function fixUnbalancedInternalFormatting(string $text): string
                 $result = str_repeat('\\' . $possibleTag, -$num) . $result;
             }
         }
-        $tagsNum = array_reverse($tagsNum);
-        foreach ($tagsNum as $possibleTag => $num) {
+        foreach (array_reverse($tagsNum) as $possibleTag => $num) {
             if ($num > 0) {
                 $result .= str_repeat('\\' . strtoupper($possibleTag), $num);
             }

tests/unit/Rose/Helper/StringHelperTest.php

@@ -1,6 +1,6 @@
 <?php declare(strict_types=1);
 /**
- * @copyright 2023 Roman Parpalak
+ * @copyright 2023-2024 Roman Parpalak
  * @license   MIT
  */
 
@@ -17,9 +17,9 @@ class StringHelperTest extends Unit
     /**
      * @dataProvider sentenceDataProvider
      */
-    public function testSentences(string $text, array $sentences): void
+    public function testSentences(string $text, array $sentences, bool $hasFormatting = false): void
     {
-        foreach (StringHelper::sentencesFromText($text, false) as $i => $str) {
+        foreach (StringHelper::sentencesFromText($text, $hasFormatting) as $i => $str) {
             $this->assertEquals($sentences[$i], $str);
         }
     }
@@ -36,6 +36,9 @@ public function sentenceDataProvider(): array
                 '1, 2, 3 и т. д.',
                 'Цифры, буквы, и т. п., могут встретиться.',
             ]],
+            ['Sentence \i1. Sentence 2. Sentence\I 3.', ['Sentence \i1.\I', '\iSentence 2.\I', '\iSentence\I 3.'], true],
+            ['Sentence \i1. Sentence 2. Sentence\B 3.', ['Sentence \i1.\I', '\iSentence 2.\I', '\b\iSentence\B 3.\I'], true],
+            ['\i\uSentence \b1\B. Sentence 2. Sentence 3.\U\I', ['\i\uSentence \b1\B.\U\I', '\i\uSentence 2.\U\I', '\i\uSentence 3.\U\I'], true],
             [
                 'Поезд отправился из пункта А в пункт Б. Затем вернулся назад.',
                 [
@@ -97,13 +100,49 @@ public function sentenceDataProvider(): array
         ];
     }
 
-    public function testFixUnbalancedInternalFormatting(): void
+    /**
+     * @dataProvider unbalancedInternalFormattingDataProvider
+     */
+    public function testFixUnbalancedInternalFormatting(string $text, string $expected, array $expectedTags): void
+    {
+        $tags = [];
+        $this->assertEquals($expected, StringHelper::fixUnbalancedInternalFormatting($text, $tags));
+        $this->assertEquals($expectedTags, $tags);
+    }
+
+    public function unbalancedInternalFormattingDataProvider(): array
     {
-        $this->assertEquals('\\iThis is \\bformatted text\\I with \\Bspecial characters\\i.\\I', StringHelper::fixUnbalancedInternalFormatting('\\iThis is \\bformatted text\\I with \\Bspecial characters\\i.'));
-        $this->assertEquals('', StringHelper::fixUnbalancedInternalFormatting(''));
-        $this->assertEquals('456789i', StringHelper::fixUnbalancedInternalFormatting('456789i'));
-        $this->assertEquals('\\i456789\\I', StringHelper::fixUnbalancedInternalFormatting('456789\\I'));
-        $this->assertEquals('\\u456789\\U', StringHelper::fixUnbalancedInternalFormatting('\\u456789'));
-        $this->assertEquals('\\i\\d\\u\\D\\\\I\\b\\B\\U', StringHelper::fixUnbalancedInternalFormatting('\\u\\D\\\\I\\b'));
+        return [
+            [
+                '\\iThis is \\bformatted text\\I with \\Bspecial characters\\i.',
+                '\\iThis is \\bformatted text\\I with \\Bspecial characters\\i.\\I',
+                ['i' => 1, 'b' => 0],
+            ],
+            [
+                '',
+                '',
+                [],
+            ],
+            [
+                '456789i',
+                '456789i',
+                [],
+            ],
+            [
+                '456789\\I',
+                '\\i456789\\I',
+                ['i' => -1],
+            ],
+            [
+                '\\u456789',
+                '\\u456789\\U',
+                ['u' => 1],
+            ],
+            [
+                '\\u\\D\\\\I\\b',
+                '\\i\\d\\u\\D\\\\I\\b\\B\\U',
+                ['i' => -1, 'd' => -1, 'u' => 1, 'b' => 1],
+            ],
+        ];
     }
 }