refactor: Enhance markdown processing for text-to-speech conversion

- Introduced new regex patterns to remove thinking and reasoning blocks from markdown input.
- Added functionality to strip emojis from the text, improving clarity for TTS.
- Implemented HTML entity decoding to ensure proper text representation.
- Replaced the existing sanitization method with a more comprehensive markdown-to-text conversion approach, enhancing performance and maintainability.

lib/core/utils/markdown_to_text.dart

@@ -5,6 +5,25 @@
 class MarkdownToText {
   const MarkdownToText._();
 
+  static final _thinkingBlockRegex = RegExp(
+    r'<details\s+type="reasoning"[^>]*>.*?</details>',
+    multiLine: true,
+    dotAll: true,
+  );
+  static final _thinkTagRegex = RegExp(
+    r'<think>.*?</think>',
+    multiLine: true,
+    dotAll: true,
+  );
+  static final _reasoningTagRegex = RegExp(
+    r'<reasoning>.*?</reasoning>',
+    multiLine: true,
+    dotAll: true,
+  );
+  static final _emojiRegex = RegExp(
+    r'[\u{1F600}-\u{1F64F}]|[\u{1F300}-\u{1F5FF}]|[\u{1F680}-\u{1F6FF}]|[\u{1F1E0}-\u{1F1FF}]|[\u{2600}-\u{26FF}]|[\u{2700}-\u{27BF}]|[\u{1F900}-\u{1F9FF}]|[\u{1FA00}-\u{1FA6F}]|[\u{1FA70}-\u{1FAFF}]|[\u{FE00}-\u{FE0F}]|[\u{1F018}-\u{1F270}]|[\u{238C}-\u{2454}]|[\u{20D0}-\u{20FF}]',
+    unicode: true,
+  );
   static final _codeBlockRegex = RegExp(
     r'```[^\n]*\n(.*?)```',
     multiLine: true,
@@ -29,18 +48,21 @@ class MarkdownToText {
     multiLine: true,
   );
   static final _htmlTagRegex = RegExp(r'<[^>]+>');
+  static final _htmlEntityRegex = RegExp(r'&[a-z]+;|&#\d+;|&#x[0-9a-f]+;');
   static final _multipleNewlinesRegex = RegExp(r'\n{3,}');
   static final _multipleSpacesRegex = RegExp(r' {2,}');
 
   /// Converts markdown text to plain text suitable for TTS.
   ///
+  /// - Removes thinking/reasoning blocks
+  /// - Removes emojis
   /// - Removes code blocks (replaces with descriptive text)
   /// - Strips all formatting (bold, italic, strikethrough)
   /// - Converts links to just their text
   /// - Removes images (or converts to alt text)
   /// - Simplifies headings
   /// - Preserves list structure with natural pauses
-  /// - Removes HTML tags
+  /// - Removes HTML tags and entities
   /// - Normalizes whitespace
   static String convert(String markdown) {
     if (markdown.trim().isEmpty) {
@@ -49,13 +71,20 @@ class MarkdownToText {
 
     var text = markdown;
 
+    // Remove thinking/reasoning blocks (must be done before general HTML tag removal)
+    text = text.replaceAll(_thinkingBlockRegex, '');
+    text = text.replaceAll(_thinkTagRegex, '');
+    text = text.replaceAll(_reasoningTagRegex, '');
+
+    // Remove emojis
+    text = text.replaceAll(_emojiRegex, '');
+
     // Remove or replace code blocks with descriptive text
     text = text.replaceAllMapped(_codeBlockRegex, (match) {
       final code = match[1]?.trim() ?? '';
       if (code.isEmpty) {
         return '';
       }
-      // For TTS, skip code blocks or use a brief description
       return ' (code block) ';
     });
 
@@ -86,7 +115,6 @@ class MarkdownToText {
     // Simplify headings (remove # symbols)
     text = text.replaceAllMapped(_headingRegex, (match) {
       final heading = match[1] ?? '';
-      // Add a pause after headings for natural speech flow
       return '$heading.\n';
     });
 
@@ -103,6 +131,20 @@ class MarkdownToText {
     // Remove HTML tags
     text = text.replaceAll(_htmlTagRegex, '');
 
+    // Decode HTML entities
+    text = text.replaceAllMapped(_htmlEntityRegex, (match) {
+      final entity = match[0] ?? '';
+      return switch (entity) {
+        '&nbsp;' => ' ',
+        '&amp;' => '&',
+        '&lt;' => '<',
+        '&gt;' => '>',
+        '&quot;' => '"',
+        '&apos;' => "'",
+        _ => entity,
+      };
+    });
+
     // Normalize whitespace
     text = text.replaceAll(_multipleNewlinesRegex, '\n\n');
     text = text.replaceAll(_multipleSpacesRegex, ' ');