Add feature: non-ASCII identifiers

The feature `non_ascii_idents` gives Rune feature parity with Rust in
supporting Unicode identifiers. Refer to:
rust-lang/rfcs#2457

crates/rune/Cargo.toml

@@ -31,6 +31,7 @@ disable-io = ["alloc"]
 fmt = ["alloc", "syntree"]
 std = ["alloc", "num/std", "serde/std", "rune-core/std", "rune-alloc/std", "musli/std", "musli/std", "once_cell/std", "anyhow/std"]
 alloc = ["anyhow", "rune-alloc/alloc", "rune-core/alloc", "once_cell/alloc", "serde/alloc"]
+non_ascii_idents = ["dep:unicode-ident"]
 
 [dependencies]
 rune-macros = { version = "=0.14.0", path = "../rune-macros" }
@@ -76,6 +77,7 @@ sha2 = { version = "0.10.6", optional = true }
 base64 = { version = "0.21.0", optional = true }
 rand = { version = "0.8.5", optional = true }
 memchr = "2.7.4"
+unicode-ident = { version = "1.0.12", optional = true }
 
 [dev-dependencies]
 tokio = { version = "1.28.1", features = ["full"] }

crates/rune/src/parse/lexer.rs

@@ -7,6 +7,24 @@ use crate::ast::Span;
 use crate::compile::{self, ErrorKind};
 use crate::SourceId;
 
+#[cfg(feature = "non_ascii_idents")]
+use unicode_ident::{is_xid_start, is_xid_continue as is_ident_continue};
+
+#[cfg(feature = "non_ascii_idents")]
+fn is_ident_start(c: char) -> bool {
+    c == '_' || is_xid_start(c)
+}
+
+#[cfg(not(feature = "non_ascii_idents"))]
+fn is_ident_start(c: char) -> bool {
+    matches!(c, '_' | 'a'..='z' | 'A'..='Z')
+}
+
+#[cfg(not(feature = "non_ascii_idents"))]
+fn is_ident_continue(c: char) -> bool {
+    matches!(c, 'a'..='z' | 'A'..='Z' | '_' | '0'..='9')
+}
+
 /// Lexer for the rune language.
 #[derive(Debug)]
 pub struct Lexer<'a> {
@@ -159,7 +177,7 @@ impl<'a> Lexer<'a> {
 
     fn next_ident(&mut self, start: usize) -> compile::Result<Option<ast::Token>> {
         while let Some(c) = self.iter.peek() {
-            if !matches!(c, 'a'..='z' | 'A'..='Z' | '_' | '0'..='9') {
+            if !is_ident_continue(c) {
                 break;
             }
 
@@ -842,7 +860,7 @@ impl<'a> Lexer<'a> {
                     '@' => ast::Kind::At,
                     '$' => ast::Kind::Dollar,
                     '~' => ast::Kind::Tilde,
-                    '_' | 'a'..='z' | 'A'..='Z' => {
+                    c if is_ident_start(c) => {
                         return self.next_ident(start);
                     }
                     '0'..='9' => {

scripts/non_ascii_idents.rn

@@ -0,0 +1,27 @@
+// Test non-ASCII identifiers (feature `non_ascii_idents`).
+
+// Creating a variable.
+let 另一個世界 = "三體世界";
+
+// Reference to a variable.
+let 世界 = 另一個世界;
+
+// In template interpolation.
+let 高論 = `你好，${另一個世界}。`;
+
+// In string formatting.
+println!("我對{另一個世界}說話：「{}」", 高論);
+
+// Compatibility check for alphanumeric characters and underscore.
+let _ = ();
+let aB_1 = ();
+let Ab_2 = ();
+let __甲_乙_丙_丁__ = ();
+
+// Naming functions and function arguments.
+fn 口號(蟲子, 主) {
+    `消除${蟲子}暴政，世界屬於${主}！`
+}
+
+// Function call.
+println!("我們的口號是：「{}」", 口號("人類", "三體"));