Auto merge of #117260 - okaneco:ascii_branchless, r=thomcc

Refactor some `char`, `u8` ASCII functions to be branchless

Extract conditions in singular `matches!` with or-patterns to individual `matches!` statements which enables branchless code output. The following functions were changed:
- `is_ascii_alphanumeric`
- `is_ascii_hexdigit`
- `is_ascii_punctuation`

Added codegen tests

---

Continued from #103024.
Based on the comment from `@scottmcm` #103024 (review).

The unmodified `is_ascii_*` functions didn't seem to benefit from extracting the conditions.

I've never written a codegen test before, but I tried to check that no branches were emitted.

library/core/src/char/methods.rs

@@ -1450,7 +1450,7 @@ impl char {
     #[rustc_const_stable(feature = "const_ascii_ctype_on_intrinsics", since = "1.47.0")]
     #[inline]
     pub const fn is_ascii_alphanumeric(&self) -> bool {
-        matches!(*self, '0'..='9' | 'A'..='Z' | 'a'..='z')
+        matches!(*self, '0'..='9') | matches!(*self, 'A'..='Z') | matches!(*self, 'a'..='z')
     }
 
     /// Checks if the value is an ASCII decimal digit:
@@ -1553,7 +1553,7 @@ impl char {
     #[rustc_const_stable(feature = "const_ascii_ctype_on_intrinsics", since = "1.47.0")]
     #[inline]
     pub const fn is_ascii_hexdigit(&self) -> bool {
-        matches!(*self, '0'..='9' | 'A'..='F' | 'a'..='f')
+        matches!(*self, '0'..='9') | matches!(*self, 'A'..='F') | matches!(*self, 'a'..='f')
     }
 
     /// Checks if the value is an ASCII punctuation character:
@@ -1591,7 +1591,10 @@ impl char {
     #[rustc_const_stable(feature = "const_ascii_ctype_on_intrinsics", since = "1.47.0")]
     #[inline]
     pub const fn is_ascii_punctuation(&self) -> bool {
-        matches!(*self, '!'..='/' | ':'..='@' | '['..='`' | '{'..='~')
+        matches!(*self, '!'..='/')
+            | matches!(*self, ':'..='@')
+            | matches!(*self, '['..='`')
+            | matches!(*self, '{'..='~')
     }
 
     /// Checks if the value is an ASCII graphic character:

library/core/src/num/mod.rs

@@ -791,7 +791,7 @@ impl u8 {
     #[rustc_const_stable(feature = "const_ascii_ctype_on_intrinsics", since = "1.47.0")]
     #[inline]
     pub const fn is_ascii_alphanumeric(&self) -> bool {
-        matches!(*self, b'0'..=b'9' | b'A'..=b'Z' | b'a'..=b'z')
+        matches!(*self, b'0'..=b'9') | matches!(*self, b'A'..=b'Z') | matches!(*self, b'a'..=b'z')
     }
 
     /// Checks if the value is an ASCII decimal digit:
@@ -894,7 +894,7 @@ impl u8 {
     #[rustc_const_stable(feature = "const_ascii_ctype_on_intrinsics", since = "1.47.0")]
     #[inline]
     pub const fn is_ascii_hexdigit(&self) -> bool {
-        matches!(*self, b'0'..=b'9' | b'A'..=b'F' | b'a'..=b'f')
+        matches!(*self, b'0'..=b'9') | matches!(*self, b'A'..=b'F') | matches!(*self, b'a'..=b'f')
     }
 
     /// Checks if the value is an ASCII punctuation character:
@@ -932,7 +932,10 @@ impl u8 {
     #[rustc_const_stable(feature = "const_ascii_ctype_on_intrinsics", since = "1.47.0")]
     #[inline]
     pub const fn is_ascii_punctuation(&self) -> bool {
-        matches!(*self, b'!'..=b'/' | b':'..=b'@' | b'['..=b'`' | b'{'..=b'~')
+        matches!(*self, b'!'..=b'/')
+            | matches!(*self, b':'..=b'@')
+            | matches!(*self, b'['..=b'`')
+            | matches!(*self, b'{'..=b'~')
     }
 
     /// Checks if the value is an ASCII graphic character:

tests/codegen/char-ascii-branchless.rs

@@ -0,0 +1,47 @@
+// Checks that these functions are branchless.
+//
+// compile-flags: -O
+
+#![crate_type = "lib"]
+
+// CHECK-LABEL: @is_ascii_alphanumeric_char
+#[no_mangle]
+pub fn is_ascii_alphanumeric_char(x: char) -> bool {
+    // CHECK-NOT: br
+    x.is_ascii_alphanumeric()
+}
+
+// CHECK-LABEL: @is_ascii_alphanumeric_u8
+#[no_mangle]
+pub fn is_ascii_alphanumeric_u8(x: u8) -> bool {
+    // CHECK-NOT: br
+    x.is_ascii_alphanumeric()
+}
+
+// CHECK-LABEL: @is_ascii_hexdigit_char
+#[no_mangle]
+pub fn is_ascii_hexdigit_char(x: char) -> bool {
+    // CHECK-NOT: br
+    x.is_ascii_hexdigit()
+}
+
+// CHECK-LABEL: @is_ascii_hexdigit_u8
+#[no_mangle]
+pub fn is_ascii_hexdigit_u8(x: u8) -> bool {
+    // CHECK-NOT: br
+    x.is_ascii_hexdigit()
+}
+
+// CHECK-LABEL: @is_ascii_punctuation_char
+#[no_mangle]
+pub fn is_ascii_punctuation_char(x: char) -> bool {
+    // CHECK-NOT: br
+    x.is_ascii_punctuation()
+}
+
+// CHECK-LABEL: @is_ascii_punctuation_u8
+#[no_mangle]
+pub fn is_ascii_punctuation_u8(x: u8) -> bool {
+    // CHECK-NOT: br
+    x.is_ascii_punctuation()
+}