From c5d0e2a0733b5271d1044979cf824dccb2a3c3d5 Mon Sep 17 00:00:00 2001
From: marcell <marcell.pardavi@gmail.com>
Date: Tue, 9 Jun 2015 23:58:00 +0200
Subject: [PATCH 1/2] Add missing #[inline] to methods related to char.

---
 src/libcore/char.rs          | 4 ++++
 src/librustc_unicode/char.rs | 8 ++++++++
 2 files changed, 12 insertions(+)

diff --git a/src/libcore/char.rs b/src/libcore/char.rs
index 48d9869f72ae4..9938c2996158a 100644
--- a/src/libcore/char.rs
+++ b/src/libcore/char.rs
@@ -152,10 +152,12 @@ pub trait CharExt {
 }
 
 impl CharExt for char {
+    #[inline]
     fn is_digit(self, radix: u32) -> bool {
         self.to_digit(radix).is_some()
     }
 
+    #[inline]
     fn to_digit(self, radix: u32) -> Option<u32> {
         if radix > 36 {
             panic!("to_digit: radix is too high (maximum 36)");
@@ -170,10 +172,12 @@ impl CharExt for char {
         else { None }
     }
 
+    #[inline]
     fn escape_unicode(self) -> EscapeUnicode {
         EscapeUnicode { c: self, state: EscapeUnicodeState::Backslash }
     }
 
+    #[inline]
     fn escape_default(self) -> EscapeDefault {
         let init_state = match self {
             '\t' => EscapeDefaultState::Backslash('t'),
diff --git a/src/librustc_unicode/char.rs b/src/librustc_unicode/char.rs
index 9dfd172707d7b..4e0d196e2f83e 100644
--- a/src/librustc_unicode/char.rs
+++ b/src/librustc_unicode/char.rs
@@ -119,6 +119,7 @@ impl char {
     /// assert_eq!('f'.to_digit(16), Some(15));
     /// ```
     #[stable(feature = "rust1", since = "1.0.0")]
+    #[inline]
     pub fn to_digit(self, radix: u32) -> Option<u32> { C::to_digit(self, radix) }
 
     /// Returns an iterator that yields the hexadecimal Unicode escape of a
@@ -157,6 +158,7 @@ impl char {
     /// assert_eq!(heart, r"\u{2764}");
     /// ```
     #[stable(feature = "rust1", since = "1.0.0")]
+    #[inline]
     pub fn escape_unicode(self) -> EscapeUnicode { C::escape_unicode(self) }
 
     /// Returns an iterator that yields the 'default' ASCII and
@@ -195,6 +197,7 @@ impl char {
     /// assert_eq!(quote, "\\\"");
     /// ```
     #[stable(feature = "rust1", since = "1.0.0")]
+    #[inline]
     pub fn escape_default(self) -> EscapeDefault { C::escape_default(self) }
 
     /// Returns the number of bytes this character would need if encoded in
@@ -208,6 +211,7 @@ impl char {
     /// assert_eq!(n, 2);
     /// ```
     #[stable(feature = "rust1", since = "1.0.0")]
+    #[inline]
     pub fn len_utf8(self) -> usize { C::len_utf8(self) }
 
     /// Returns the number of 16-bit code units this character would need if
@@ -221,6 +225,7 @@ impl char {
     /// assert_eq!(n, 1);
     /// ```
     #[stable(feature = "rust1", since = "1.0.0")]
+    #[inline]
     pub fn len_utf16(self) -> usize { C::len_utf16(self) }
 
     /// Encodes this character as UTF-8 into the provided byte buffer, and then
@@ -255,6 +260,7 @@ impl char {
     /// ```
     #[unstable(feature = "unicode",
                reason = "pending decision about Iterator/Writer/Reader")]
+    #[inline]
     pub fn encode_utf8(self, dst: &mut [u8]) -> Option<usize> { C::encode_utf8(self, dst) }
 
     /// Encodes this character as UTF-16 into the provided `u16` buffer, and
@@ -289,6 +295,7 @@ impl char {
     /// ```
     #[unstable(feature = "unicode",
                reason = "pending decision about Iterator/Writer/Reader")]
+    #[inline]
     pub fn encode_utf16(self, dst: &mut [u16]) -> Option<usize> { C::encode_utf16(self, dst) }
 
     /// Returns whether the specified character is considered a Unicode
@@ -451,5 +458,6 @@ impl char {
                  since = "1.0.0")]
     #[unstable(feature = "unicode",
                reason = "needs expert opinion. is_cjk flag stands out as ugly")]
+    #[inline]
     pub fn width(self, is_cjk: bool) -> Option<usize> { charwidth::width(self, is_cjk) }
 }

From e87c62fb12e6b02cfc39fc2a16c315615714757a Mon Sep 17 00:00:00 2001
From: marcell <marcell.pardavi@gmail.com>
Date: Wed, 10 Jun 2015 02:03:56 +0200
Subject: [PATCH 2/2] Modify String::push to reallocate more conservatively in
 case of the character's UTF-8 representation is bigger than 1 byte

---
 src/libcollections/string.rs | 34 +++++++++++++++++-----------------
 1 file changed, 17 insertions(+), 17 deletions(-)

diff --git a/src/libcollections/string.rs b/src/libcollections/string.rs
index 7ede6545b9fb2..6717f2f45fab5 100644
--- a/src/libcollections/string.rs
+++ b/src/libcollections/string.rs
@@ -468,24 +468,24 @@ impl String {
     #[inline]
     #[stable(feature = "rust1", since = "1.0.0")]
     pub fn push(&mut self, ch: char) {
-        if (ch as u32) < 0x80 {
-            self.vec.push(ch as u8);
-            return;
-        }
-
-        let cur_len = self.len();
-        // This may use up to 4 bytes.
-        self.vec.reserve(4);
+        match ch.len_utf8() {
+            1 => self.vec.push(ch as u8),
+            ch_len => {
+                let cur_len = self.len();
+                // This may use up to 4 bytes.
+                self.vec.reserve(ch_len);
 
-        unsafe {
-            // Attempt to not use an intermediate buffer by just pushing bytes
-            // directly onto this string.
-            let slice = slice::from_raw_parts_mut (
-                self.vec.as_mut_ptr().offset(cur_len as isize),
-                4
-            );
-            let used = ch.encode_utf8(slice).unwrap_or(0);
-            self.vec.set_len(cur_len + used);
+                unsafe {
+                    // Attempt to not use an intermediate buffer by just pushing bytes
+                    // directly onto this string.
+                    let slice = slice::from_raw_parts_mut (
+                        self.vec.as_mut_ptr().offset(cur_len as isize),
+                        ch_len
+                    );
+                    let used = ch.encode_utf8(slice).unwrap_or(0);
+                    self.vec.set_len(cur_len + used);
+                }
+            }
         }
     }