Parser: skip UTF-8 BOM

Skip the byte order mark.
BOM is also allowed in .rs or Cargo.toml file.

Fixes #7291

internal/compiler/lexer.rs

@@ -198,6 +198,17 @@ pub fn lex(mut source: &str) -> Vec<crate::parser::Token> {
     let mut result = vec![];
     let mut offset = 0;
     let mut state = LexState::default();
+    if source.starts_with("\u{FEFF}") {
+        // Skip BOM
+        result.push(crate::parser::Token {
+            kind: SyntaxKind::Whitespace,
+            text: source[..3].into(),
+            offset: 0,
+            ..Default::default()
+        });
+        source = &source[3..];
+        offset += 3;
+    }
     while !source.is_empty() {
         if let Some((len, kind)) = crate::parser::lex_next_token(source, &mut state) {
             result.push(crate::parser::Token {

internal/compiler/tests/syntax/basic/bom-simple.slint

@@ -0,0 +1,12 @@
+// Copyright © SixtyFPS GmbH <info@slint.dev>
+// SPDX-License-Identifier: GPL-3.0-only OR LicenseRef-Slint-Royalty-free-2.0 OR LicenseRef-Slint-Software-3.0
+
+// This file contains an UTF-8 BOM at the beginning of the file.
+
+  component Foo {  }
+//^warning{Component is neither used nor exported}
+
+
+export component XX {
+
+}

internal/compiler/tests/syntax_tests.rs

@@ -69,6 +69,12 @@ fn syntax_tests() -> std::io::Result<()> {
 
 fn process_file(path: &std::path::Path) -> std::io::Result<bool> {
     let source = std::fs::read_to_string(path)?;
+    if path.to_str().unwrap_or("").contains("bom-") && !source.starts_with("\u{FEFF}") {
+        // make sure that the file still contains BOM and it wasn't remove by some tools
+        return Err(std::io::Error::other(format!(
+            "{path:?} does not contains BOM while it should"
+        )));
+    }
     std::panic::catch_unwind(|| process_file_source(path, source, false)).unwrap_or_else(|err| {
         println!("Panic while processing {}: {:?}", path.display(), err);
         Ok(false)