add xtask

.cargo/config.toml

@@ -0,0 +1,2 @@
+[alias]
+xtask = "run --package xtask --"

Cargo.toml

@@ -1,3 +1,9 @@
 [workspace]
 resolver = "2"
-members = ["starpls", "starpls_lexer", "starpls_parser", "starpls_syntax"]
+members = [
+    "starpls",
+    "starpls_lexer",
+    "starpls_parser",
+    "starpls_syntax",
+    "xtask",
+]

starpls_lexer/src/unescape.rs

@@ -151,7 +151,7 @@ fn scan_byte_string_escape<F>(
             'x' => scan_byte_string_hexadecimal_escape(chars),
 
             // unicode escapes start with either `u` or `U`
-            'u' => {
+            'u' | 'U' => {
                 let res = scan_unicode_escape(chars);
                 let end = input_len - chars.as_str().len();
                 match res {
@@ -271,19 +271,15 @@ fn scan_unicode_escape(chars: &mut Chars<'_>) -> Result<char, EscapeError> {
                 let digit = c.to_digit(16).expect("invalid hexadecimal digit");
                 value = value * 16 + digit;
                 num_digits += 1;
-                if num_digits == 4 || num_digits == 8 {
-                    break char::from_u32(value).ok_or_else(|| {
-                        if value > 0x10FFFF {
-                            EscapeError::OutOfRangeUnicodeEscape
-                        } else {
-                            EscapeError::LoneSurrogateUnicodeEscape
-                        }
-                    });
+                if num_digits == 8 {
+                    break;
                 }
             }
             _ => {
-                break Err(if num_digits == 0 {
+                return Err(if num_digits == 0 {
                     EscapeError::EmptyUnicodeEscape
+                } else if num_digits == 4 {
+                    break;
                 } else if num_digits < 4 {
                     EscapeError::TooShort16BitUnicodeEscape
                 } else if num_digits < 8 {
@@ -294,4 +290,12 @@ fn scan_unicode_escape(chars: &mut Chars<'_>) -> Result<char, EscapeError> {
             }
         }
     }
+
+    char::from_u32(value).ok_or_else(|| {
+        if value > 0x10FFFF {
+            EscapeError::OutOfRangeUnicodeEscape
+        } else {
+            EscapeError::LoneSurrogateUnicodeEscape
+        }
+    })
 }

starpls_lexer/src/unescape/tests.rs

@@ -42,6 +42,15 @@ fn test_unescape_byte_string() {
             6..10 "😿"
         "#]],
     );
+    check(
+        r#"\x41\u0400\u4e16\U0001F63F"#,
+        expect![[r#"
+        0..4 "A"
+        4..10 "Ѐ"
+        10..16 "世"
+        16..26 "😿"
+    "#]],
+    );
     check(
         r#"\377\378\x80\xff\xff"#,
         expect![[r#"

xtask/Cargo.toml

@@ -0,0 +1,10 @@
+[package]
+name = "xtask"
+version = "0.1.0"
+edition = "2021"
+
+# See more keys and their definitions at https://doc.rust-lang.org/cargo/reference/manifest.html
+
+[dependencies]
+anyhow = "1.0.76"
+clap = { version = "4.4.11", features = ["derive"] }