auto merge of #12797 : pczarn/rust/str_safety, r=huonw

Along the lines of `shift_ref` and `pop_ref`.

src/libglob/lib.rs

@@ -369,11 +369,11 @@ impl Pattern {
                             return EntirePatternDoesntMatch;
                         }
 
-                        let (c, next) = file.slice_shift_char();
-                        if require_literal(c) {
+                        let (some_c, next) = file.slice_shift_char();
+                        if require_literal(some_c.unwrap()) {
                             return SubPatternDoesntMatch;
                         }
-                        prev_char.set(Some(c));
+                        prev_char.set(some_c);
                         file = next;
                     }
                 }
@@ -382,7 +382,8 @@ impl Pattern {
                         return EntirePatternDoesntMatch;
                     }
 
-                    let (c, next) = file.slice_shift_char();
+                    let (some_c, next) = file.slice_shift_char();
+                    let c = some_c.unwrap();
                     let matches = match *token {
                         AnyChar => {
                             !require_literal(c)
@@ -403,7 +404,7 @@ impl Pattern {
                     if !matches {
                         return SubPatternDoesntMatch;
                     }
-                    prev_char.set(Some(c));
+                    prev_char.set(some_c);
                     file = next;
                 }
             }

src/libstd/str.rs

@@ -1355,6 +1355,7 @@ pub mod raw {
     use libc;
     use ptr;
     use ptr::RawPtr;
+    use option::{Option, Some, None};
     use str::{is_utf8, OwnedStr, StrSlice};
     use vec;
     use vec::{MutableVector, ImmutableVector, OwnedVector};
@@ -1464,23 +1465,31 @@ pub mod raw {
     }
 
     /// Removes the last byte from a string and returns it.
+    /// Returns None when an empty string is passed.
     /// The caller must preserve the valid UTF-8 property.
-    pub unsafe fn pop_byte(s: &mut ~str) -> u8 {
+    pub unsafe fn pop_byte(s: &mut ~str) -> Option<u8> {
         let len = s.len();
-        assert!((len > 0u));
-        let b = s[len - 1u];
-        s.set_len(len - 1);
-        return b;
+        if len == 0u {
+            return None;
+        } else {
+            let b = s[len - 1u];
+            s.set_len(len - 1);
+            return Some(b);
+        }
     }
 
     /// Removes the first byte from a string and returns it.
+    /// Returns None when an empty string is passed.
     /// The caller must preserve the valid UTF-8 property.
-    pub unsafe fn shift_byte(s: &mut ~str) -> u8 {
+    pub unsafe fn shift_byte(s: &mut ~str) -> Option<u8> {
         let len = s.len();
-        assert!((len > 0u));
-        let b = s[0];
-        *s = s.slice(1, len).to_owned();
-        return b;
+        if len == 0u {
+            return None;
+        } else {
+            let b = s[0];
+            *s = s.slice(1, len).to_owned();
+            return Some(b);
+        }
     }
 
     /// Access the str in its vector representation.
@@ -2273,25 +2282,22 @@ pub trait StrSlice<'a> {
     /// Retrieves the first character from a string slice and returns
     /// it. This does not allocate a new string; instead, it returns a
     /// slice that point one character beyond the character that was
-    /// shifted.
-    ///
-    /// # Failure
-    ///
-    /// If the string does not contain any characters.
+    /// shifted. If the string does not contain any characters,
+    /// a tuple of None and an empty string is returned instead.
     ///
     /// # Example
     ///
     /// ```rust
     /// let s = "Löwe 老虎 Léopard";
     /// let (c, s1) = s.slice_shift_char();
-    /// assert_eq!(c, 'L');
+    /// assert_eq!(c, Some('L'));
     /// assert_eq!(s1, "öwe 老虎 Léopard");
     ///
     /// let (c, s2) = s1.slice_shift_char();
-    /// assert_eq!(c, 'ö');
+    /// assert_eq!(c, Some('ö'));
     /// assert_eq!(s2, "we 老虎 Léopard");
     /// ```
-    fn slice_shift_char(&self) -> (char, &'a str);
+    fn slice_shift_char(&self) -> (Option<char>, &'a str);
 
     /// Levenshtein Distance between two strings.
     fn lev_distance(&self, t: &str) -> uint;
@@ -2744,10 +2750,14 @@ impl<'a> StrSlice<'a> for &'a str {
     }
 
     #[inline]
-    fn slice_shift_char(&self) -> (char, &'a str) {
-        let CharRange {ch, next} = self.char_range_at(0u);
-        let next_s = unsafe { raw::slice_bytes(*self, next, self.len()) };
-        return (ch, next_s);
+    fn slice_shift_char(&self) -> (Option<char>, &'a str) {
+        if self.is_empty() {
+            return (None, *self);
+        } else {
+            let CharRange {ch, next} = self.char_range_at(0u);
+            let next_s = unsafe { raw::slice_bytes(*self, next, self.len()) };
+            return (Some(ch), next_s);
+        }
     }
 
     fn lev_distance(&self, t: &str) -> uint {
@@ -2810,19 +2820,13 @@ pub trait OwnedStr {
     /// Appends a character to the back of a string
     fn push_char(&mut self, c: char);
 
-    /// Remove the final character from a string and return it
-    ///
-    /// # Failure
-    ///
-    /// If the string does not contain any characters
-    fn pop_char(&mut self) -> char;
+    /// Remove the final character from a string and return it. Return None
+    /// when the string is empty.
+    fn pop_char(&mut self) -> Option<char>;
 
-    /// Remove the first character from a string and return it
-    ///
-    /// # Failure
-    ///
-    /// If the string does not contain any characters
-    fn shift_char(&mut self) -> char;
+    /// Remove the first character from a string and return it. Return None
+    /// when the string is empty.
+    fn shift_char(&mut self) -> Option<char>;
 
     /// Prepend a char to a string
     fn unshift_char(&mut self, ch: char);
@@ -2925,19 +2929,26 @@ impl OwnedStr for ~str {
     }
 
     #[inline]
-    fn pop_char(&mut self) -> char {
+    fn pop_char(&mut self) -> Option<char> {
         let end = self.len();
-        assert!(end > 0u);
-        let CharRange {ch, next} = self.char_range_at_reverse(end);
-        unsafe { self.set_len(next); }
-        return ch;
+        if end == 0u {
+            return None;
+        } else {
+            let CharRange {ch, next} = self.char_range_at_reverse(end);
+            unsafe { self.set_len(next); }
+            return Some(ch);
+        }
     }
 
     #[inline]
-    fn shift_char(&mut self) -> char {
-        let CharRange {ch, next} = self.char_range_at(0u);
-        *self = self.slice(next, self.len()).to_owned();
-        return ch;
+    fn shift_char(&mut self) -> Option<char> {
+        if self.is_empty() {
+            return None;
+        } else {
+            let CharRange {ch, next} = self.char_range_at(0u);
+            *self = self.slice(next, self.len()).to_owned();
+            return Some(ch);
+        }
     }
 
     #[inline]
@@ -3148,7 +3159,7 @@ mod tests {
         let mut data = ~"ประเทศไทย中华";
         let cc = data.pop_char();
         assert_eq!(~"ประเทศไทย中", data);
-        assert_eq!('华', cc);
+        assert_eq!(Some('华'), cc);
     }
 
     #[test]
@@ -3156,14 +3167,15 @@ mod tests {
         let mut data2 = ~"华";
         let cc2 = data2.pop_char();
         assert_eq!(~"", data2);
-        assert_eq!('华', cc2);
+        assert_eq!(Some('华'), cc2);
     }
 
     #[test]
-    #[should_fail]
-    fn test_pop_char_fail() {
+    fn test_pop_char_empty() {
         let mut data = ~"";
-        let _cc3 = data.pop_char();
+        let cc3 = data.pop_char();
+        assert_eq!(~"", data);
+        assert_eq!(None, cc3);
     }
 
     #[test]
@@ -3182,7 +3194,7 @@ mod tests {
         let mut data = ~"ประเทศไทย中";
         let cc = data.shift_char();
         assert_eq!(~"ระเทศไทย中", data);
-        assert_eq!('ป', cc);
+        assert_eq!(Some('ป'), cc);
     }
 
     #[test]
@@ -3599,6 +3611,18 @@ mod tests {
         assert!(!"   _   ".is_whitespace());
     }
 
+    #[test]
+    fn test_slice_shift_char() {
+        let data = "ประเทศไทย中";
+        assert_eq!(data.slice_shift_char(), (Some('ป'), "ระเทศไทย中"));
+    }
+
+    #[test]
+    fn test_slice_shift_char_2() {
+        let empty = "";
+        assert_eq!(empty.slice_shift_char(), (None, ""));
+    }
+
     #[test]
     fn test_push_byte() {
         let mut s = ~"ABC";
@@ -3611,7 +3635,7 @@ mod tests {
         let mut s = ~"ABC";
         let b = unsafe{raw::shift_byte(&mut s)};
         assert_eq!(s, ~"BC");
-        assert_eq!(b, 65u8);
+        assert_eq!(b, Some(65u8));
     }
 
     #[test]
@@ -3619,7 +3643,7 @@ mod tests {
         let mut s = ~"ABC";
         let b = unsafe{raw::pop_byte(&mut s)};
         assert_eq!(s, ~"AB");
-        assert_eq!(b, 67u8);
+        assert_eq!(b, Some(67u8));
     }
 
     #[test]

src/test/run-pass/utf8_chars.rs

@@ -42,12 +42,12 @@ pub fn main() {
     assert!((!str::is_utf8([0xf0_u8, 0xff_u8, 0xff_u8, 0x10_u8])));
 
     let mut stack = ~"a×c€";
-    assert_eq!(stack.pop_char(), '€');
-    assert_eq!(stack.pop_char(), 'c');
+    assert_eq!(stack.pop_char(), Some('€'));
+    assert_eq!(stack.pop_char(), Some('c'));
     stack.push_char('u');
     assert!(stack == ~"a×u");
-    assert_eq!(stack.shift_char(), 'a');
-    assert_eq!(stack.shift_char(), '×');
+    assert_eq!(stack.shift_char(), Some('a'));
+    assert_eq!(stack.shift_char(), Some('×'));
     stack.unshift_char('ß');
     assert!(stack == ~"ßu");
 }