Fix UTF-8 BOM file type detection for first-line syntax patterns

[krikera]

This fix strips the UTF-8 BOM character from the first line before performing syntax detection using strip_prefix('\u{feff}'), ensuring that files with BOM are handled correctly.

Changes

• Modified get_first_line_syntax in src/assets.rs to strip BOM before pattern matching
• Added comprehensive test coverage for XML, shell scripts, and PHP files with BOM
• Updated CHANGELOG.md

Fixes #3314

[keith-hall]

Nice, thanks