From 215feb9313bb91666da71e1ae73baaefb8f438ec Mon Sep 17 00:00:00 2001
From: Josh McKinney <joshka@users.noreply.github.com>
Date: Thu, 26 Sep 2024 13:54:28 -0700
Subject: [PATCH] Make AsciiSet consts and fields values

Refs take 8 bytes, whereas the values are only 16 bytes, so there is not
a huge benefit to using references rather than values. PercentEncoding
is changed to store the AsciiSet as a value, and the functions that take
AsciiSet now take Into<AsciiSet> instead of &'static AsciiSet. This
allows existing code to continue to work without modification. The
AsciiSet consts (CONTROLS and NON_ALPHANUMERIC) are also changed to be
values, which is a breaking change, but will only affect code that
attempts to dereference them.

Discussion about the rationale for this is change is at
<https://github.com/servo/rust-url/pull/970#discussion_r1773725026>
---
 percent_encoding/src/ascii_set.rs | 12 +++++++++---
 percent_encoding/src/lib.rs       | 25 ++++++++++++++++++++-----
 2 files changed, 29 insertions(+), 8 deletions(-)

diff --git a/percent_encoding/src/ascii_set.rs b/percent_encoding/src/ascii_set.rs
index 41cd235e..ad70a210 100644
--- a/percent_encoding/src/ascii_set.rs
+++ b/percent_encoding/src/ascii_set.rs
@@ -24,7 +24,7 @@ use core::{mem, ops};
 /// /// https://url.spec.whatwg.org/#fragment-percent-encode-set
 /// const FRAGMENT: &AsciiSet = &CONTROLS.add(b' ').add(b'"').add(b'<').add(b'>').add(b'`');
 /// ```
-#[derive(Debug, PartialEq, Eq)]
+#[derive(Debug, Clone, Copy, PartialEq, Eq)]
 pub struct AsciiSet {
     mask: [Chunk; ASCII_RANGE_LEN / BITS_PER_CHUNK],
 }
@@ -83,6 +83,12 @@ impl AsciiSet {
     }
 }
 
+impl From<&AsciiSet> for AsciiSet {
+    fn from(set: &AsciiSet) -> Self {
+        *set
+    }
+}
+
 impl ops::Add for AsciiSet {
     type Output = Self;
 
@@ -104,7 +110,7 @@ impl ops::Not for AsciiSet {
 /// Note that this includes the newline and tab characters, but not the space 0x20.
 ///
 /// <https://url.spec.whatwg.org/#c0-control-percent-encode-set>
-pub const CONTROLS: &AsciiSet = &AsciiSet {
+pub const CONTROLS: AsciiSet = AsciiSet {
     mask: [
         !0_u32, // C0: 0x00 to 0x1F (32 bits set)
         0,
@@ -134,7 +140,7 @@ static_assert! {
 /// Everything that is not an ASCII letter or digit.
 ///
 /// This is probably more eager than necessary in any context.
-pub const NON_ALPHANUMERIC: &AsciiSet = &CONTROLS
+pub const NON_ALPHANUMERIC: AsciiSet = CONTROLS
     .add(b' ')
     .add(b'!')
     .add(b'"')
diff --git a/percent_encoding/src/lib.rs b/percent_encoding/src/lib.rs
index 6ea05237..b2bf033e 100644
--- a/percent_encoding/src/lib.rs
+++ b/percent_encoding/src/lib.rs
@@ -32,7 +32,7 @@
 //! use percent_encoding::{utf8_percent_encode, AsciiSet, CONTROLS};
 //!
 //! /// https://url.spec.whatwg.org/#fragment-percent-encode-set
-//! const FRAGMENT: &AsciiSet = &CONTROLS.add(b' ').add(b'"').add(b'<').add(b'>').add(b'`');
+//! const FRAGMENT: AsciiSet = CONTROLS.add(b' ').add(b'"').add(b'<').add(b'>').add(b'`');
 //!
 //! assert_eq!(utf8_percent_encode("foo <bar>", FRAGMENT).to_string(), "foo%20%3Cbar%3E");
 //! ```
@@ -114,10 +114,10 @@ pub fn percent_encode_byte(byte: u8) -> &'static str {
 /// assert_eq!(percent_encode(b"foo bar?", NON_ALPHANUMERIC).to_string(), "foo%20bar%3F");
 /// ```
 #[inline]
-pub fn percent_encode<'a>(input: &'a [u8], ascii_set: &'static AsciiSet) -> PercentEncode<'a> {
+pub fn percent_encode<'a, T: Into<AsciiSet>>(input: &'a [u8], ascii_set: T) -> PercentEncode<'a> {
     PercentEncode {
         bytes: input,
-        ascii_set,
+        ascii_set: ascii_set.into(),
     }
 }
 
@@ -133,7 +133,10 @@ pub fn percent_encode<'a>(input: &'a [u8], ascii_set: &'static AsciiSet) -> Perc
 /// assert_eq!(utf8_percent_encode("foo bar?", NON_ALPHANUMERIC).to_string(), "foo%20bar%3F");
 /// ```
 #[inline]
-pub fn utf8_percent_encode<'a>(input: &'a str, ascii_set: &'static AsciiSet) -> PercentEncode<'a> {
+pub fn utf8_percent_encode<'a, T: Into<AsciiSet>>(
+    input: &'a str,
+    ascii_set: T,
+) -> PercentEncode<'a> {
     percent_encode(input.as_bytes(), ascii_set)
 }
 
@@ -141,7 +144,7 @@ pub fn utf8_percent_encode<'a>(input: &'a str, ascii_set: &'static AsciiSet) ->
 #[derive(Debug, Clone, PartialEq, Eq)]
 pub struct PercentEncode<'a> {
     bytes: &'a [u8],
-    ascii_set: &'static AsciiSet,
+    ascii_set: AsciiSet,
 }
 
 impl<'a> Iterator for PercentEncode<'a> {
@@ -383,6 +386,12 @@ mod tests {
         assert_eq!(encoded.collect::<String>(), "foo bar?");
     }
 
+    #[test]
+    fn percent_encode_accepts_value() {
+        let encoded = percent_encode(b"foo bar?", AsciiSet::EMPTY);
+        assert_eq!(encoded.collect::<String>(), "foo bar?");
+    }
+
     #[test]
     fn percent_encode_collect() {
         let encoded = percent_encode(b"foo bar?", NON_ALPHANUMERIC);
@@ -410,6 +419,12 @@ mod tests {
         assert_eq!(encoded.collect::<String>(), "foo bar?");
     }
 
+    #[test]
+    fn utf8_percent_encode_accepts_value() {
+        let encoded = super::utf8_percent_encode("foo bar?", AsciiSet::EMPTY);
+        assert_eq!(encoded.collect::<String>(), "foo bar?");
+    }
+
     #[test]
     fn utf8_percent_encode() {
         assert_eq!(