fix(css): preserve CSS unicode escapes in noUselessEscapeInString

.changeset/fix-css-unicode-escape.md

@@ -0,0 +1,5 @@
+---
+"@biomejs/biome": patch
+---
+
+Fixed [#9385](https://github.com/biomejs/biome/issues/9385): the [`noUselessEscapeInString`](https://biomejs.dev/linter/rules/no-useless-escape-in-string/) rule no longer removes valid CSS unicode escape sequences (e.g. `\e7bb`, `\e644`) from string literals. Previously, backslashes followed by hex digits `8`–`9` and `a`–`f`/`A`–`F` were incorrectly treated as useless escapes, breaking iconfont `content` properties.

crates/biome_css_analyze/src/lint/suspicious/no_useless_escape_in_string.rs

@@ -19,7 +19,7 @@ declare_lint_rule! {
     ///
     /// ```css,expect_diagnostic
     /// a::after {
-    ///   content: "\a"
+    ///   content: "\g"
     /// }
     /// ```
     ///
@@ -43,6 +43,12 @@ declare_lint_rule! {
     /// }
     /// ```
     ///
+    /// ```css
+    /// a::after {
+    ///   content: "\e7bb"
+    /// }
+    /// ```
+    ///
     pub NoUselessEscapeInString {
         version: "2.0.0",
         name: "noUselessEscapeInString",
@@ -109,16 +115,17 @@ fn next_useless_escape(str: &str, quote: u8) -> Option<usize> {
                 b'^'
                 | b'\r'
                 | b'\n'
-                | b'0'..=b'7'
                 | b'\\'
-                | b'b'
-                | b'f'
                 | b'n'
                 | b'r'
                 | b't'
                 | b'u'
                 | b'v'
-                | b'x' => {}
+                | b'x'
+                // CSS unicode escape: \ followed by 1-6 hex digits
+                | b'0'..=b'9'
+                | b'a'..=b'f'
+                | b'A'..=b'F' => {}
                 // Preserve escaping of Unicode characters U+2028 and U+2029
                 0xE2 => {
                     if !(matches!(it.next(), Some((_, 0x80)))

crates/biome_css_analyze/tests/specs/suspicious/noUselessEscapeInString/invalid.css

@@ -1,5 +1,5 @@
 .a::after {
-    content: /*before*/ "useless \a" /*after*/
+    content: /*before*/ "useless \g" /*after*/
 }
 .b::after {
     content: "useless \'"

crates/biome_css_analyze/tests/specs/suspicious/noUselessEscapeInString/invalid.css.snap

@@ -5,7 +5,7 @@ expression: invalid.css
 # Input
 ```css
 .a::after {
-    content: /*before*/ "useless \a" /*after*/
+    content: /*before*/ "useless \g" /*after*/
 }
 .b::after {
     content: "useless \'"
@@ -19,7 +19,7 @@ invalid.css:2:35 lint/suspicious/noUselessEscapeInString  FIXABLE  ━━━━
   ! The character doesn't need to be escaped.
 
     1 │ .a::after {
-  > 2 │     content: /*before*/ "useless \a" /*after*/
+  > 2 │     content: /*before*/ "useless \g" /*after*/
       │                                   ^
     3 │ }
     4 │ .b::after {
@@ -28,7 +28,7 @@ invalid.css:2:35 lint/suspicious/noUselessEscapeInString  FIXABLE  ━━━━
 
   i Safe fix: Unescape the character.
 
-    2 │ ····content:·/*before*/·"useless·\a"·/*after*/
+    2 │ ····content:·/*before*/·"useless·\g"·/*after*/
       │                                  -            
 
 ```

crates/biome_css_analyze/tests/specs/suspicious/noUselessEscapeInString/valid.css

@@ -4,4 +4,20 @@
 }
 .b::after {
     content: "\t"
+}
+/* CSS unicode escapes: \ followed by 1-6 hex digits */
+.c::after {
+    content: "\e7bb"
+}
+.d::after {
+    content: "\e644"
+}
+.e::after {
+    content: "\a"
+}
+.f::after {
+    content: "\0a"
+}
+.g::after {
+    content: "\AB12CD"
 }

crates/biome_css_analyze/tests/specs/suspicious/noUselessEscapeInString/valid.css.snap

@@ -11,4 +11,22 @@ expression: valid.css
 .b::after {
     content: "\t"
 }
+/* CSS unicode escapes: \ followed by 1-6 hex digits */
+.c::after {
+    content: "\e7bb"
+}
+.d::after {
+    content: "\e644"
+}
+.e::after {
+    content: "\a"
+}
+.f::after {
+    content: "\0a"
+}
+.g::after {
+    content: "\AB12CD"
+}
 ```
+
+_Note: The parser emitted 2 diagnostics which are not shown here._