---

yaml
---
r: 152986
b: refs/heads/try2
c: 5cef124
h: refs/heads/master
v: v3

Author: bors

[refs]

@@ -5,7 +5,7 @@ refs/heads/snap-stage3: 78a7676898d9f80ab540c6df5d4c9ce35bb50463
 refs/heads/try: 519addf6277dbafccbb4159db4b710c37eaa2ec5
 refs/tags/release-0.1: 1f5c5126e96c79d22cb7862f75304136e204f105
 refs/heads/ndm: f3868061cd7988080c30d6d5bf352a5a5fe2460b
-refs/heads/try2: 68ead460f9029fd8de508a46fe944afb83e99da3
+refs/heads/try2: 5cef1243a21c162cb9968f9bc14d1a23ba88426e
 refs/heads/dist-snap: ba4081a5a8573875fed17545846f6f6902c8ba8d
 refs/tags/release-0.2: c870d2dffb391e14efb05aa27898f1f6333a9596
 refs/tags/release-0.3: b5f0d0f648d9a6153664837026ba1be43d3e2503

branches/try2/src/doc/guide.md

@@ -160,7 +160,7 @@ Save the file, and then type this into your terminal window:
 
 ```{bash}
 $ rustc hello_world.rs
-$ ./hello_world # just 'hello_world' on Windows
+$ ./hello_world # or hello_world.exe on Windows
 Hello, world
 ```
 
@@ -243,7 +243,7 @@ There are now two files: our source code, with the `.rs` extension, and the
 executable (`hello_world.exe` on Windows, `hello_world` everywhere else)
 
 ```{bash}
-$ ./hello_world  # or ./hello_world.exe on Windows
+$ ./hello_world  # or hello_world.exe on Windows
 ```
 
 This prints out our `Hello, world!` text to our terminal.

branches/try2/src/libcollections/str.rs

@@ -106,6 +106,14 @@ pub fn from_utf8_owned(vv: Vec<u8>) -> Result<String, Vec<u8>> {
 /// # Failure
 ///
 /// Fails if invalid UTF-8
+///
+/// # Example
+///
+/// ```rust
+/// use std::str;
+/// let string = str::from_byte(66u8);
+/// assert_eq!(string.as_slice(), "B");
+/// ```
 pub fn from_byte(b: u8) -> String {
     assert!(b < 128u8);
     String::from_char(1, b as char)
@@ -803,15 +811,9 @@ pub trait StrAllocating: Str {
     }
 
     /// Converts to a vector of `u16` encoded as UTF-16.
+    #[deprecated = "use `utf16_units` instead"]
     fn to_utf16(&self) -> Vec<u16> {
-        let me = self.as_slice();
-        let mut u = Vec::new();
-        for ch in me.chars() {
-            let mut buf = [0u16, ..2];
-            let n = ch.encode_utf16(buf /* as mut slice! */);
-            u.push_all(buf.slice_to(n));
-        }
-        u
+        self.as_slice().utf16_units().collect::<Vec<u16>>()
     }
 
     /// Given a string, make a new string with repeated copies of it.
@@ -1619,14 +1621,17 @@ mod tests {
 
         for p in pairs.iter() {
             let (s, u) = (*p).clone();
+            let s_as_utf16 = s.as_slice().utf16_units().collect::<Vec<u16>>();
+            let u_as_string = from_utf16(u.as_slice()).unwrap();
+
             assert!(is_utf16(u.as_slice()));
-            assert_eq!(s.to_utf16(), u);
+            assert_eq!(s_as_utf16, u);
 
-            assert_eq!(from_utf16(u.as_slice()).unwrap(), s);
+            assert_eq!(u_as_string, s);
             assert_eq!(from_utf16_lossy(u.as_slice()), s);
 
-            assert_eq!(from_utf16(s.to_utf16().as_slice()).unwrap(), s);
-            assert_eq!(from_utf16(u.as_slice()).unwrap().to_utf16(), u);
+            assert_eq!(from_utf16(s_as_utf16.as_slice()).unwrap(), s);
+            assert_eq!(u_as_string.as_slice().utf16_units().collect::<Vec<u16>>(), u);
         }
     }
 

branches/try2/src/libcore/str.rs

@@ -16,6 +16,7 @@
 
 use mem;
 use char;
+use char::Char;
 use clone::Clone;
 use cmp;
 use cmp::{PartialEq, Eq};
@@ -24,7 +25,7 @@ use default::Default;
 use iter::{Filter, Map, Iterator};
 use iter::{DoubleEndedIterator, ExactSize};
 use iter::range;
-use num::Saturating;
+use num::{CheckedMul, Saturating};
 use option::{None, Option, Some};
 use raw::Repr;
 use slice::ImmutableVector;
@@ -557,6 +558,41 @@ impl<'a> Iterator<&'a str> for StrSplits<'a> {
     }
 }
 
+/// External iterator for a string's UTF16 codeunits.
+/// Use with the `std::iter` module.
+#[deriving(Clone)]
+pub struct Utf16CodeUnits<'a> {
+    chars: Chars<'a>,
+    extra: u16
+}
+
+impl<'a> Iterator<u16> for Utf16CodeUnits<'a> {
+    #[inline]
+    fn next(&mut self) -> Option<u16> {
+        if self.extra != 0 {
+            let tmp = self.extra;
+            self.extra = 0;
+            return Some(tmp);
+        }
+
+        let mut buf = [0u16, ..2];
+        self.chars.next().map(|ch| {
+            let n = ch.encode_utf16(buf /* as mut slice! */);
+            if n == 2 { self.extra = buf[1]; }
+            buf[0]
+        })
+    }
+
+    #[inline]
+    fn size_hint(&self) -> (uint, Option<uint>) {
+        let (low, high) = self.chars.size_hint();
+        // every char gets either one u16 or two u16,
+        // so this iterator is between 1 or 2 times as
+        // long as the underlying iterator.
+        (low, high.and_then(|n| n.checked_mul(&2)))
+    }
+}
+
 /*
 Section: Comparing strings
 */
@@ -1609,6 +1645,9 @@ pub trait StrSlice<'a> {
     /// and that it is not reallocated (e.g. by pushing to the
     /// string).
     fn as_ptr(&self) -> *const u8;
+
+    /// Return an iterator of `u16` over the string encoded as UTF-16.
+    fn utf16_units(&self) -> Utf16CodeUnits<'a>;
 }
 
 impl<'a> StrSlice<'a> for &'a str {
@@ -1957,6 +1996,11 @@ impl<'a> StrSlice<'a> for &'a str {
     fn as_ptr(&self) -> *const u8 {
         self.repr().data
     }
+
+    #[inline]
+    fn utf16_units(&self) -> Utf16CodeUnits<'a> {
+        Utf16CodeUnits{ chars: self.chars(), extra: 0}
+    }
 }
 
 impl<'a> Default for &'a str {

branches/try2/src/libnative/io/c_win32.rs

@@ -70,6 +70,7 @@ extern "system" {
 
 pub mod compat {
     use std::intrinsics::{atomic_store_relaxed, transmute};
+    use std::iter::Iterator;
     use libc::types::os::arch::extra::{LPCWSTR, HMODULE, LPCSTR, LPVOID};
 
     extern "system" {
@@ -82,7 +83,8 @@ pub mod compat {
     // layer (after it's loaded) shouldn't be any slower than a regular DLL
     // call.
     unsafe fn store_func(ptr: *mut uint, module: &str, symbol: &str, fallback: uint) {
-        let module = module.to_utf16().append_one(0);
+        let module: Vec<u16> = module.utf16_units().collect();
+        let module = module.append_one(0);
         symbol.with_c_str(|symbol| {
             let handle = GetModuleHandleW(module.as_ptr());
             let func: uint = transmute(GetProcAddress(handle, symbol));

branches/try2/src/libnative/io/file_win32.rs

@@ -255,7 +255,7 @@ impl Drop for Inner {
 
 pub fn to_utf16(s: &CString) -> IoResult<Vec<u16>> {
     match s.as_str() {
-        Some(s) => Ok(s.to_utf16().append_one(0)),
+        Some(s) => Ok(s.utf16_units().collect::<Vec<u16>>().append_one(0)),
         None => Err(IoError {
             code: libc::ERROR_INVALID_NAME as uint,
             extra: 0,

branches/try2/src/libnative/io/process.rs

@@ -294,6 +294,8 @@ fn spawn_process_os(cfg: ProcessConfig,
     use libc::funcs::extra::msvcrt::get_osfhandle;
 
     use std::mem;
+    use std::iter::Iterator;
+    use std::str::StrSlice;
 
     if cfg.gid.is_some() || cfg.uid.is_some() {
         return Err(IoError {
@@ -328,7 +330,8 @@ fn spawn_process_os(cfg: ProcessConfig,
                         lpSecurityDescriptor: ptr::mut_null(),
                         bInheritHandle: 1,
                     };
-                    let filename = "NUL".to_utf16().append_one(0);
+                    let filename: Vec<u16> = "NUL".utf16_units().collect();
+                    let filename = filename.append_one(0);
                     *slot = libc::CreateFileW(filename.as_ptr(),
                                               access,
                                               libc::FILE_SHARE_READ |
@@ -371,7 +374,8 @@ fn spawn_process_os(cfg: ProcessConfig,
 
         with_envp(cfg.env, |envp| {
             with_dirp(cfg.cwd, |dirp| {
-                let mut cmd_str = cmd_str.to_utf16().append_one(0);
+                let mut cmd_str: Vec<u16> = cmd_str.as_slice().utf16_units().collect();
+                cmd_str = cmd_str.append_one(0);
                 let created = CreateProcessW(ptr::null(),
                                              cmd_str.as_mut_ptr(),
                                              ptr::mut_null(),
@@ -770,7 +774,7 @@ fn with_envp<T>(env: Option<&[(CString, CString)]>, cb: |*mut c_void| -> T) -> T
                 let kv = format!("{}={}",
                                  pair.ref0().as_str().unwrap(),
                                  pair.ref1().as_str().unwrap());
-                blk.push_all(kv.to_utf16().as_slice());
+                blk.extend(kv.as_slice().utf16_units());
                 blk.push(0);
             }
 
@@ -788,7 +792,9 @@ fn with_dirp<T>(d: Option<&CString>, cb: |*const u16| -> T) -> T {
       Some(dir) => {
           let dir_str = dir.as_str()
                            .expect("expected workingdirectory to be utf-8 encoded");
-          let dir_str = dir_str.to_utf16().append_one(0);
+          let dir_str: Vec<u16> = dir_str.utf16_units().collect();
+          let dir_str = dir_str.append_one(0);
+
           cb(dir_str.as_ptr())
       },
       None => cb(ptr::null())

branches/try2/src/librustdoc/flock.rs

@@ -162,7 +162,8 @@ mod imp {
 
     impl Lock {
         pub fn new(p: &Path) -> Lock {
-            let p_16 = p.as_str().unwrap().to_utf16().append_one(0);
+            let p_16: Vec<u16> = p.as_str().unwrap().utf16_units().collect();
+            let p_16 = p_16.append_one(0);
             let handle = unsafe {
                 libc::CreateFileW(p_16.as_ptr(),
                                   libc::FILE_GENERIC_READ |

branches/try2/src/librustdoc/lib.rs

@@ -408,18 +408,17 @@ fn json_output(krate: clean::Crate, res: Vec<plugins::PluginJson> ,
     //   "crate": { parsed crate ... },
     //   "plugins": { output of plugins ... }
     // }
-    let mut json = box std::collections::TreeMap::new();
-    json.insert("schema".to_string(),
-                json::String(SCHEMA_VERSION.to_string()));
-    let plugins_json = box res.move_iter()
-                              .filter_map(|opt| {
-                                  match opt {
-                                      None => None,
-                                      Some((string, json)) => {
-                                          Some((string.to_string(), json))
-                                      }
+    let mut json = std::collections::TreeMap::new();
+    json.insert("schema".to_string(), json::String(SCHEMA_VERSION.to_string()));
+    let plugins_json = res.move_iter()
+                          .filter_map(|opt| {
+                              match opt {
+                                  None => None,
+                                  Some((string, json)) => {
+                                      Some((string.to_string(), json))
                                   }
-                              }).collect();
+                              }
+                          }).collect();
 
     // FIXME #8335: yuck, Rust -> str -> JSON round trip! No way to .encode
     // straight to the Rust JSON representation.
@@ -429,7 +428,7 @@ fn json_output(krate: clean::Crate, res: Vec<plugins::PluginJson> ,
             let mut encoder = json::Encoder::new(&mut w as &mut io::Writer);
             krate.encode(&mut encoder).unwrap();
         }
-        str::from_utf8(w.unwrap().as_slice()).unwrap().to_string()
+        str::from_utf8_owned(w.unwrap()).unwrap()
     };
     let crate_json = match json::from_str(crate_json_str.as_slice()) {
         Ok(j) => j,
@@ -440,6 +439,5 @@ fn json_output(krate: clean::Crate, res: Vec<plugins::PluginJson> ,
     json.insert("plugins".to_string(), json::Object(plugins_json));
 
     let mut file = try!(File::create(&dst));
-    try!(json::Object(json).to_writer(&mut file));
-    Ok(())
+    json::Object(json).to_writer(&mut file)
 }