auto merge of #10701 : huonw/rust/rm-from_utf8, r=brson

This function had type &[u8] -> ~str, i.e. it allocates a string
internally, even though the non-allocating version that take &[u8] ->
&str and ~[u8] -> ~str are all that is necessary in most circumstances.

10 files changed, 56 insertions, 159 deletions

diff --git a/src/libstd/c_str.rs b/src/libstd/c_str.rs
index 306ee331929..11845c766ed 100644
--- a/src/libstd/c_str.rs
+++ b/src/libstd/c_str.rs
@@ -167,7 +167,7 @@ impl CString {
         if self.buf.is_null() { return None; }
         let buf = self.as_bytes();
         let buf = buf.slice_to(buf.len()-1); // chop off the trailing NUL
-        str::from_utf8_slice_opt(buf)
+        str::from_utf8_opt(buf)
     }
 
     /// Return a CString iterator.
diff --git a/src/libstd/fmt/mod.rs b/src/libstd/fmt/mod.rs
index c74a9bc9051..463540b3677 100644
--- a/src/libstd/fmt/mod.rs
+++ b/src/libstd/fmt/mod.rs
@@ -802,7 +802,7 @@ impl<'self> Formatter<'self> {
 
     fn runplural(&mut self, value: uint, pieces: &[rt::Piece]) {
         ::uint::to_str_bytes(value, 10, |buf| {
-            let valuestr = str::from_utf8_slice(buf);
+            let valuestr = str::from_utf8(buf);
             for piece in pieces.iter() {
                 self.run(piece, Some(valuestr));
             }
diff --git a/src/libstd/io/flate.rs b/src/libstd/io/flate.rs
index 8a5aa171eb8..4a31449e105 100644
--- a/src/libstd/io/flate.rs
+++ b/src/libstd/io/flate.rs
@@ -107,7 +107,7 @@ mod test {
     fn smoke_test() {
         let mem_writer = MemWriter::new();
         let mut deflate_writer = DeflateWriter::new(mem_writer);
-        let in_msg = "test";
+        let in_msg: &str = "test";
         let in_bytes = in_msg.as_bytes();
         deflate_writer.write(in_bytes);
         deflate_writer.flush();
@@ -118,6 +118,6 @@ mod test {
         let bytes_read = inflate_reader.read(out_bytes).unwrap();
         assert_eq!(bytes_read, in_bytes.len());
         let out_msg = str::from_utf8(out_bytes);
-        assert!(in_msg == out_msg);
+        assert_eq!(in_msg, out_msg);
     }
 }
diff --git a/src/libstd/io/fs.rs b/src/libstd/io/fs.rs
index e239f630f01..f0b51a2c3e0 100644
--- a/src/libstd/io/fs.rs
+++ b/src/libstd/io/fs.rs
@@ -770,9 +770,9 @@ mod test {
             let mut read_buf = [0, .. 1028];
             let read_str = match read_stream.read(read_buf).unwrap() {
                 -1|0 => fail!("shouldn't happen"),
-                n => str::from_utf8(read_buf.slice_to(n))
+                n => str::from_utf8_owned(read_buf.slice_to(n).to_owned())
             };
-            assert!(read_str == message.to_owned());
+            assert_eq!(read_str, message.to_owned());
         }
         unlink(filename);
     })
@@ -801,7 +801,7 @@ mod test {
     })
 
     test!(fn file_test_io_non_positional_read() {
-        let message = "ten-four";
+        let message: &str = "ten-four";
         let mut read_mem = [0, .. 8];
         let tmpdir = tmpdir();
         let filename = &tmpdir.join("file_rt_io_file_test_positional.txt");
@@ -822,7 +822,7 @@ mod test {
         }
         unlink(filename);
         let read_str = str::from_utf8(read_mem);
-        assert!(read_str == message.to_owned());
+        assert_eq!(read_str, message);
     })
 
     test!(fn file_test_io_seek_and_tell_smoke_test() {
@@ -846,9 +846,9 @@ mod test {
         }
         unlink(filename);
         let read_str = str::from_utf8(read_mem);
-        assert!(read_str == message.slice(4, 8).to_owned());
-        assert!(tell_pos_pre_read == set_cursor);
-        assert!(tell_pos_post_read == message.len() as u64);
+        assert_eq!(read_str, message.slice(4, 8));
+        assert_eq!(tell_pos_pre_read, set_cursor);
+        assert_eq!(tell_pos_post_read, message.len() as u64);
     })
 
     test!(fn file_test_io_seek_and_write() {
@@ -877,9 +877,9 @@ mod test {
     test!(fn file_test_io_seek_shakedown() {
         use std::str;          // 01234567890123
         let initial_msg =   "qwer-asdf-zxcv";
-        let chunk_one = "qwer";
-        let chunk_two = "asdf";
-        let chunk_three = "zxcv";
+        let chunk_one: &str = "qwer";
+        let chunk_two: &str = "asdf";
+        let chunk_three: &str = "zxcv";
         let mut read_mem = [0, .. 4];
         let tmpdir = tmpdir();
         let filename = &tmpdir.join("file_rt_io_file_test_seek_shakedown.txt");
@@ -892,18 +892,15 @@ mod test {
 
             read_stream.seek(-4, SeekEnd);
             read_stream.read(read_mem);
-            let read_str = str::from_utf8(read_mem);
-            assert!(read_str == chunk_three.to_owned());
+            assert_eq!(str::from_utf8(read_mem), chunk_three);
 
             read_stream.seek(-9, SeekCur);
             read_stream.read(read_mem);
-            let read_str = str::from_utf8(read_mem);
-            assert!(read_str == chunk_two.to_owned());
+            assert_eq!(str::from_utf8(read_mem), chunk_two);
 
             read_stream.seek(0, SeekSet);
             read_stream.read(read_mem);
-            let read_str = str::from_utf8(read_mem);
-            assert!(read_str == chunk_one.to_owned());
+            assert_eq!(str::from_utf8(read_mem), chunk_one);
         }
         unlink(filename);
     })
@@ -982,7 +979,7 @@ mod test {
                     None|Some("") => fail!("really shouldn't happen.."),
                     Some(n) => prefix+n
                 };
-                assert!(expected == read_str);
+                assert_eq!(expected.as_slice(), read_str);
             }
             unlink(f);
         }
diff --git a/src/libstd/io/mod.rs b/src/libstd/io/mod.rs
index c00233dda55..208c64f5ef4 100644
--- a/src/libstd/io/mod.rs
+++ b/src/libstd/io/mod.rs
@@ -1049,7 +1049,7 @@ pub trait Buffer: Reader {
             Some(n) if n == width => {}
             Some(..) | None => return None // read error
         }
-        match str::from_utf8_slice_opt(buf.slice_to(width)) {
+        match str::from_utf8_opt(buf.slice_to(width)) {
             Some(s) => Some(s.char_at(0)),
             None => None
         }
diff --git a/src/libstd/num/strconv.rs b/src/libstd/num/strconv.rs
index 1028cef9dc6..8e678ab66b2 100644
--- a/src/libstd/num/strconv.rs
+++ b/src/libstd/num/strconv.rs
@@ -426,7 +426,7 @@ pub fn float_to_str_common<T:NumCast+Zero+One+Eq+Ord+NumStrConv+Float+Round+
         sign: SignFormat, digits: SignificantDigits) -> (~str, bool) {
     let (bytes, special) = float_to_str_bytes_common(num, radix,
                                negative_zero, sign, digits);
-    (str::from_utf8(bytes), special)
+    (str::from_utf8_owned(bytes), special)
 }
 
 // Some constants for from_str_bytes_common's input validation,
diff --git a/src/libstd/path/mod.rs b/src/libstd/path/mod.rs
index 8ecfdb3a9e0..79989b838f6 100644
--- a/src/libstd/path/mod.rs
+++ b/src/libstd/path/mod.rs
@@ -176,7 +176,7 @@ pub trait GenericPath: Clone + GenericPathUnsafe {
     /// If the path is not representable in utf-8, this returns None.
     #[inline]
     fn as_str<'a>(&'a self) -> Option<&'a str> {
-        str::from_utf8_slice_opt(self.as_vec())
+        str::from_utf8_opt(self.as_vec())
     }
 
     /// Returns the path as a byte vector
@@ -207,7 +207,7 @@ pub trait GenericPath: Clone + GenericPathUnsafe {
     /// See `dirname` for details.
     #[inline]
     fn dirname_str<'a>(&'a self) -> Option<&'a str> {
-        str::from_utf8_slice_opt(self.dirname())
+        str::from_utf8_opt(self.dirname())
     }
     /// Returns the file component of `self`, as a byte vector.
     /// If `self` represents the root of the file hierarchy, returns None.
@@ -217,7 +217,7 @@ pub trait GenericPath: Clone + GenericPathUnsafe {
     /// See `filename` for details.
     #[inline]
     fn filename_str<'a>(&'a self) -> Option<&'a str> {
-        self.filename().and_then(str::from_utf8_slice_opt)
+        self.filename().and_then(str::from_utf8_opt)
     }
     /// Returns the stem of the filename of `self`, as a byte vector.
     /// The stem is the portion of the filename just before the last '.'.
@@ -239,7 +239,7 @@ pub trait GenericPath: Clone + GenericPathUnsafe {
     /// See `filestem` for details.
     #[inline]
     fn filestem_str<'a>(&'a self) -> Option<&'a str> {
-        self.filestem().and_then(str::from_utf8_slice_opt)
+        self.filestem().and_then(str::from_utf8_opt)
     }
     /// Returns the extension of the filename of `self`, as an optional byte vector.
     /// The extension is the portion of the filename just after the last '.'.
@@ -262,7 +262,7 @@ pub trait GenericPath: Clone + GenericPathUnsafe {
     /// See `extension` for details.
     #[inline]
     fn extension_str<'a>(&'a self) -> Option<&'a str> {
-        self.extension().and_then(str::from_utf8_slice_opt)
+        self.extension().and_then(str::from_utf8_opt)
     }
 
     /// Replaces the filename portion of the path with the given byte vector or string.
@@ -493,12 +493,12 @@ pub trait BytesContainer {
     /// Raises `str::null_byte` if not utf-8
     #[inline]
     fn container_as_str<'a>(&'a self) -> &'a str {
-        str::from_utf8_slice(self.container_as_bytes())
+        str::from_utf8(self.container_as_bytes())
     }
     /// Returns the receiver interpreted as a utf-8 string, if possible
     #[inline]
     fn container_as_str_opt<'a>(&'a self) -> Option<&'a str> {
-        str::from_utf8_slice_opt(self.container_as_bytes())
+        str::from_utf8_opt(self.container_as_bytes())
     }
     /// Returns whether .container_as_str() is guaranteed to not fail
     // FIXME (#8888): Remove unused arg once ::<for T> works
diff --git a/src/libstd/path/posix.rs b/src/libstd/path/posix.rs
index b2bc00dd247..ddf2cce21b0 100644
--- a/src/libstd/path/posix.rs
+++ b/src/libstd/path/posix.rs
@@ -396,13 +396,13 @@ impl Path {
     /// Returns an iterator that yields each component of the path as Option<&str>.
     /// See components() for details.
     pub fn str_components<'a>(&'a self) -> StrComponentIter<'a> {
-        self.components().map(str::from_utf8_slice_opt)
+        self.components().map(str::from_utf8_opt)
     }
 
     /// Returns an iterator that yields each component of the path in reverse as Option<&str>.
     /// See components() for details.
     pub fn rev_str_components<'a>(&'a self) -> RevStrComponentIter<'a> {
-        self.rev_components().map(str::from_utf8_slice_opt)
+        self.rev_components().map(str::from_utf8_opt)
     }
 }
 
@@ -684,7 +684,7 @@ mod tests {
             (s: $path:expr, $op:ident, $exp:expr, opt) => (
                 {
                     let path = Path::init($path);
-                    let left = path.$op().map(|x| str::from_utf8_slice(x));
+                    let left = path.$op().map(|x| str::from_utf8(x));
                     assert_eq!(left, $exp);
                 }
             );
diff --git a/src/libstd/run.rs b/src/libstd/run.rs
index a22f536974b..6cc5e5cc9f2 100644
--- a/src/libstd/run.rs
+++ b/src/libstd/run.rs
@@ -347,7 +347,7 @@ mod tests {
 
         let run::ProcessOutput {status, output, error}
              = run::process_output("echo", [~"hello"]);
-        let output_str = str::from_utf8(output);
+        let output_str = str::from_utf8_owned(output);
 
         assert!(status.success());
         assert_eq!(output_str.trim().to_owned(), ~"hello");
@@ -439,7 +439,7 @@ mod tests {
         let mut prog = run::Process::new("echo", [~"hello"], run::ProcessOptions::new());
         let run::ProcessOutput {status, output, error}
             = prog.finish_with_output();
-        let output_str = str::from_utf8(output);
+        let output_str = str::from_utf8_owned(output);
 
         assert!(status.success());
         assert_eq!(output_str.trim().to_owned(), ~"hello");
@@ -457,7 +457,7 @@ mod tests {
         let run::ProcessOutput {status, output, error}
             = prog.finish_with_output();
 
-        let output_str = str::from_utf8(output);
+        let output_str = str::from_utf8_owned(output);
 
         assert!(status.success());
         assert_eq!(output_str.trim().to_owned(), ~"hello");
@@ -504,7 +504,7 @@ mod tests {
     fn test_keep_current_working_dir() {
         let mut prog = run_pwd(None);
 
-        let output = str::from_utf8(prog.finish_with_output().output);
+        let output = str::from_utf8_owned(prog.finish_with_output().output);
         let parent_dir = os::getcwd();
         let child_dir = Path::init(output.trim());
 
@@ -522,7 +522,7 @@ mod tests {
         let parent_dir = os::getcwd().dir_path();
         let mut prog = run_pwd(Some(&parent_dir));
 
-        let output = str::from_utf8(prog.finish_with_output().output);
+        let output = str::from_utf8_owned(prog.finish_with_output().output);
         let child_dir = Path::init(output.trim());
 
         let parent_stat = parent_dir.stat();
@@ -561,7 +561,7 @@ mod tests {
         if running_on_valgrind() { return; }
 
         let mut prog = run_env(None);
-        let output = str::from_utf8(prog.finish_with_output().output);
+        let output = str::from_utf8_owned(prog.finish_with_output().output);
 
         let r = os::env();
         for &(ref k, ref v) in r.iter() {
@@ -575,7 +575,7 @@ mod tests {
         if running_on_valgrind() { return; }
 
         let mut prog = run_env(None);
-        let output = str::from_utf8(prog.finish_with_output().output);
+        let output = str::from_utf8_owned(prog.finish_with_output().output);
 
         let r = os::env();
         for &(ref k, ref v) in r.iter() {
@@ -594,7 +594,7 @@ mod tests {
         new_env.push((~"RUN_TEST_NEW_ENV", ~"123"));
 
         let mut prog = run_env(Some(new_env));
-        let output = str::from_utf8(prog.finish_with_output().output);
+        let output = str::from_utf8_owned(prog.finish_with_output().output);
 
         assert!(output.contains("RUN_TEST_NEW_ENV=123"));
     }
diff --git a/src/libstd/str.rs b/src/libstd/str.rs
index f65ec6971ab..c1898a9b920 100644
--- a/src/libstd/str.rs
+++ b/src/libstd/str.rs
@@ -124,34 +124,6 @@ condition! {
 Section: Creating a string
 */
 
-/// Convert a vector of bytes to a new UTF-8 string
-///
-/// # Failure
-///
-/// Raises the `not_utf8` condition if invalid UTF-8
-pub fn from_utf8(vv: &[u8]) -> ~str {
-    use str::not_utf8::cond;
-
-    match from_utf8_opt(vv) {
-        None => {
-            let first_bad_byte = *vv.iter().find(|&b| !is_utf8([*b])).unwrap();
-            cond.raise(format!("from_utf8: input is not UTF-8; first bad \
-                                byte is {}", first_bad_byte))
-        }
-        Some(s) => s
-    }
-}
-
-/// Convert a vector of bytes to a new UTF-8 string, if possible.
-/// Returns None if the vector contains invalid UTF-8.
-pub fn from_utf8_opt(vv: &[u8]) -> Option<~str> {
-    if is_utf8(vv) {
-        Some(unsafe { raw::from_utf8(vv) })
-    } else {
-        None
-    }
-}
-
 /// Consumes a vector of bytes to create a new utf-8 string
 ///
 /// # Failure
@@ -187,16 +159,16 @@ pub fn from_utf8_owned_opt(vv: ~[u8]) -> Option<~str> {
 /// # Failure
 ///
 /// Fails if invalid UTF-8
-pub fn from_utf8_slice<'a>(v: &'a [u8]) -> &'a str {
-    from_utf8_slice_opt(v).expect("from_utf8_slice: not utf-8")
+pub fn from_utf8<'a>(v: &'a [u8]) -> &'a str {
+    from_utf8_opt(v).expect("from_utf8: not utf-8")
 }
 
 /// Converts a vector to a string slice without performing any allocations.
 ///
 /// Returns None if the slice is not utf-8.
-pub fn from_utf8_slice_opt<'a>(v: &'a [u8]) -> Option<&'a str> {
+pub fn from_utf8_opt<'a>(v: &'a [u8]) -> Option<&'a str> {
     if is_utf8(v) {
-        Some(unsafe { cast::transmute(v) })
+        Some(unsafe { raw::from_utf8(v) })
     } else { None }
 }
 
@@ -1055,9 +1027,10 @@ pub mod raw {
         from_buf_len(buf as *u8, i as uint)
     }
 
-    /// Converts a vector of bytes to a new owned string.
-    pub unsafe fn from_utf8(v: &[u8]) -> ~str {
-        v.as_imm_buf(|buf, len| from_buf_len(buf, len))
+    /// Converts a slice of bytes to a string slice without checking
+    /// that the string contains valid UTF-8.
+    pub unsafe fn from_utf8<'a>(v: &'a [u8]) -> &'a str {
+        cast::transmute(v)
     }
 
     /// Converts an owned vector of bytes to a new owned string. This assumes
@@ -1068,7 +1041,7 @@ pub mod raw {
     }
 
     /// Converts a byte to a string.
-    pub unsafe fn from_byte(u: u8) -> ~str { from_utf8([u]) }
+    pub unsafe fn from_byte(u: u8) -> ~str { from_utf8_owned(~[u]) }
 
     /// Form a slice from a C string. Unsafe because the caller must ensure the
     /// C string has the static lifetime, or else the return value may be
@@ -3078,33 +3051,6 @@ mod tests {
     }
 
     #[test]
-    fn test_unsafe_from_utf8() {
-        let a = ~[65u8, 65u8, 65u8, 65u8, 65u8, 65u8, 65u8];
-        let b = unsafe { raw::from_utf8(a) };
-        assert_eq!(b, ~"AAAAAAA");
-    }
-
-    #[test]
-    fn test_from_utf8() {
-        let ss = ~"ศไทย中华Việt Nam";
-        let bb = ~[0xe0_u8, 0xb8_u8, 0xa8_u8,
-                  0xe0_u8, 0xb9_u8, 0x84_u8,
-                  0xe0_u8, 0xb8_u8, 0x97_u8,
-                  0xe0_u8, 0xb8_u8, 0xa2_u8,
-                  0xe4_u8, 0xb8_u8, 0xad_u8,
-                  0xe5_u8, 0x8d_u8, 0x8e_u8,
-                  0x56_u8, 0x69_u8, 0xe1_u8,
-                  0xbb_u8, 0x87_u8, 0x74_u8,
-                  0x20_u8, 0x4e_u8, 0x61_u8,
-                  0x6d_u8];
-
-
-        assert_eq!(ss, from_utf8(bb));
-        assert_eq!(~"𐌀𐌖𐌋𐌄𐌑𐌉ปรدولة الكويتทศไทย中华𐍅𐌿𐌻𐍆𐌹𐌻𐌰",
-                   from_utf8(bytes!("𐌀𐌖𐌋𐌄𐌑𐌉ปรدولة الكويتทศไทย中华𐍅𐌿𐌻𐍆𐌹𐌻𐌰")));
-    }
-
-    #[test]
     fn test_is_utf8() {
         // deny overlong encodings
         assert!(!is_utf8([0xc0, 0x80]));
@@ -3129,31 +3075,6 @@ mod tests {
         assert!(is_utf8([0xF4, 0x8F, 0xBF, 0xBF]));
     }
 
-
-    #[test]
-    fn test_from_utf8_fail() {
-        use str::not_utf8::cond;
-
-        let bb = ~[0xff_u8, 0xb8_u8, 0xa8_u8,
-                  0xe0_u8, 0xb9_u8, 0x84_u8,
-                  0xe0_u8, 0xb8_u8, 0x97_u8,
-                  0xe0_u8, 0xb8_u8, 0xa2_u8,
-                  0xe4_u8, 0xb8_u8, 0xad_u8,
-                  0xe5_u8, 0x8d_u8, 0x8e_u8,
-                  0x56_u8, 0x69_u8, 0xe1_u8,
-                  0xbb_u8, 0x87_u8, 0x74_u8,
-                  0x20_u8, 0x4e_u8, 0x61_u8,
-                  0x6d_u8];
-
-        let mut error_happened = false;
-        let _x = cond.trap(|err| {
-            assert_eq!(err, ~"from_utf8: input is not UTF-8; first bad byte is 255");
-            error_happened = true;
-            ~""
-        }).inside(|| from_utf8(bb));
-        assert!(error_happened);
-    }
-
     #[test]
     fn test_raw_from_c_str() {
         unsafe {
@@ -3232,7 +3153,7 @@ mod tests {
         let s1: ~str = ~"All mimsy were the borogoves";
 
         let v: ~[u8] = s1.as_bytes().to_owned();
-        let s2: ~str = from_utf8(v);
+        let s2: ~str = from_utf8(v).to_owned();
         let mut i: uint = 0u;
         let n1: uint = s1.len();
         let n2: uint = v.len();
@@ -3755,49 +3676,28 @@ mod tests {
     }
 
     #[test]
-    fn test_str_from_utf8_slice() {
+    fn test_str_from_utf8() {
         let xs = bytes!("hello");
-        assert_eq!(from_utf8_slice(xs), "hello");
+        assert_eq!(from_utf8(xs), "hello");
 
         let xs = bytes!("ศไทย中华Việt Nam");
-        assert_eq!(from_utf8_slice(xs), "ศไทย中华Việt Nam");
+        assert_eq!(from_utf8(xs), "ศไทย中华Việt Nam");
     }
 
     #[test]
     #[should_fail]
-    fn test_str_from_utf8_slice_invalid() {
+    fn test_str_from_utf8_invalid() {
         let xs = bytes!("hello", 0xff);
-        let _ = from_utf8_slice(xs);
-    }
-
-    #[test]
-    fn test_str_from_utf8_slice_opt() {
-        let xs = bytes!("hello");
-        assert_eq!(from_utf8_slice_opt(xs), Some("hello"));
-
-        let xs = bytes!("ศไทย中华Việt Nam");
-        assert_eq!(from_utf8_slice_opt(xs), Some("ศไทย中华Việt Nam"));
-
-        let xs = bytes!("hello", 0xff);
-        assert_eq!(from_utf8_slice_opt(xs), None);
-    }
-
-    #[test]
-    fn test_str_from_utf8() {
-        let xs = bytes!("hello");
-        assert_eq!(from_utf8(xs), ~"hello");
-
-        let xs = bytes!("ศไทย中华Việt Nam");
-        assert_eq!(from_utf8(xs), ~"ศไทย中华Việt Nam");
+        let _ = from_utf8(xs);
     }
 
     #[test]
     fn test_str_from_utf8_opt() {
-        let xs = bytes!("hello").to_owned();
-        assert_eq!(from_utf8_opt(xs), Some(~"hello"));
+        let xs = bytes!("hello");
+        assert_eq!(from_utf8_opt(xs), Some("hello"));
 
         let xs = bytes!("ศไทย中华Việt Nam");
-        assert_eq!(from_utf8_opt(xs), Some(~"ศไทย中华Việt Nam"));
+        assert_eq!(from_utf8_opt(xs), Some("ศไทย中华Việt Nam"));
 
         let xs = bytes!("hello", 0xff);
         assert_eq!(from_utf8_opt(xs), None);