Double-encoded UTF-8**: Riesterförderung instead of Riesterförderung
Latin-1 Mojibake**: ä instead of ä, ü instead of ü
CP1252 artifacts**: “ instead of ", â€" instead of —
MySQL latin1 quirks**: Ãœ (Ã+œ) instead of Ü, ß (Ã+Ÿ) instead of ß

This component scans your entire database, identifies these problems, and repairs them safely.
Features:

Scans all text columns (VARCHAR, TEXT, etc.) across any table
Detects four encoding problem types:

  - Double-encoded UTF-8
  - Latin-1 misread as UTF-8 (Mojibake)
  - Windows-1252 misread as UTF-8
  - Invalid UTF-8 byte sequences

Handles MySQL's latin1 = CP1252 mapping (bytes 0x80-0x9F)
Regex-based repair that fixes intact patterns without destroying already-broken bytes

       Dry-run preview before any changes are made

Transaction-based repairs (atomic rollback on error)
Configurable row limit per table
Works with Joomla and non-Joomla tables

    - No external dependencies — pure PHP with mbstring
Requirements:

Joomla 5.0+ or 6.0+
PHP 8.1+ with mbstring extension
MySQL or MariaDB