---

yaml
---
r: 226935
b: refs/heads/master
c: fe5336b
h: refs/heads/master
i:
  226933: 3a35d32
  226931: 0674c47
  226927: e5141a3
v: v3

Author: Dave Huseby

[refs]

@@ -1,5 +1,5 @@
 ---
-refs/heads/master: 741a19f8efcb060febfaf29bee13860a55ba8bf2
+refs/heads/master: fe5336b04d134559d910ce8f27e6deb3ca997119
 refs/heads/snap-stage3: 1af31d4974e33027a68126fa5a5a3c2c6491824f
 refs/heads/try: b53c0f93eedcdedd4fd89bccc5a3a09d1c5cd23e
 refs/tags/release-0.1: 1f5c5126e96c79d22cb7862f75304136e204f105

trunk/src/doc/trpl/the-stack-and-the-heap.md

@@ -549,7 +549,7 @@ reuse.
 If you’d like to dive into this topic in greater detail, [this paper][wilson]
 is a great introduction.
 
-[wilson]: http://citeseerx.ist.psu.edu/viewdoc/summary?doi=10.1.1.143.4688
+[wilson]: http://www.cs.northwestern.edu/~pdinda/icsclass/doc/dsa.pdf
 
 ## Semantic impact
 

trunk/src/etc/unicode.py

@@ -395,6 +395,46 @@ def emit_conversions_module(f, to_upper, to_lower, to_title):
         is_pub=False, t_type = t_type, pfun=pfun)
     f.write("}\n\n")
 
+def emit_grapheme_module(f, grapheme_table, grapheme_cats):
+    f.write("""pub mod grapheme {
+    use core::slice::SliceExt;
+    pub use self::GraphemeCat::*;
+    use core::result::Result::{Ok, Err};
+
+    #[allow(non_camel_case_types)]
+    #[derive(Clone, Copy)]
+    pub enum GraphemeCat {
+""")
+    for cat in grapheme_cats + ["Any"]:
+        f.write("        GC_" + cat + ",\n")
+    f.write("""    }
+
+    fn bsearch_range_value_table(c: char, r: &'static [(char, char, GraphemeCat)]) -> GraphemeCat {
+        use core::cmp::Ordering::{Equal, Less, Greater};
+        match r.binary_search_by(|&(lo, hi, _)| {
+            if lo <= c && c <= hi { Equal }
+            else if hi < c { Less }
+            else { Greater }
+        }) {
+            Ok(idx) => {
+                let (_, _, cat) = r[idx];
+                cat
+            }
+            Err(_) => GC_Any
+        }
+    }
+
+    pub fn grapheme_category(c: char) -> GraphemeCat {
+        bsearch_range_value_table(c, grapheme_cat_table)
+    }
+
+""")
+
+    emit_table(f, "grapheme_cat_table", grapheme_table, "&'static [(char, char, GraphemeCat)]",
+        pfun=lambda x: "(%s,%s,GC_%s)" % (escape_char(x[0]), escape_char(x[1]), x[2]),
+        is_pub=False)
+    f.write("}\n")
+
 def emit_charwidth_module(f, width_table):
     f.write("pub mod charwidth {\n")
     f.write("    use core::option::Option;\n")
@@ -457,6 +497,79 @@ def emit_norm_module(f, canon, compat, combine, norm_props):
     canon_comp_keys = canon_comp.keys()
     canon_comp_keys.sort()
 
+    f.write("pub mod normalization {\n")
+
+    def mkdata_fun(table):
+        def f(char):
+            data = "(%s,&[" % escape_char(char)
+            first = True
+            for d in table[char]:
+                if not first:
+                    data += ","
+                first = False
+                data += escape_char(d)
+            data += "])"
+            return data
+        return f
+
+    f.write("    // Canonical decompositions\n")
+    emit_table(f, "canonical_table", canon_keys, "&'static [(char, &'static [char])]",
+        pfun=mkdata_fun(canon))
+
+    f.write("    // Compatibility decompositions\n")
+    emit_table(f, "compatibility_table", compat_keys, "&'static [(char, &'static [char])]",
+        pfun=mkdata_fun(compat))
+
+    def comp_pfun(char):
+        data = "(%s,&[" % escape_char(char)
+        canon_comp[char].sort(lambda x, y: x[0] - y[0])
+        first = True
+        for pair in canon_comp[char]:
+            if not first:
+                data += ","
+            first = False
+            data += "(%s,%s)" % (escape_char(pair[0]), escape_char(pair[1]))
+        data += "])"
+        return data
+
+    f.write("    // Canonical compositions\n")
+    emit_table(f, "composition_table", canon_comp_keys,
+        "&'static [(char, &'static [(char, char)])]", pfun=comp_pfun)
+
+    f.write("""
+    fn bsearch_range_value_table(c: char, r: &'static [(char, char, u8)]) -> u8 {
+        use core::cmp::Ordering::{Equal, Less, Greater};
+        use core::slice::SliceExt;
+        use core::result::Result::{Ok, Err};
+        match r.binary_search_by(|&(lo, hi, _)| {
+            if lo <= c && c <= hi { Equal }
+            else if hi < c { Less }
+            else { Greater }
+        }) {
+            Ok(idx) => {
+                let (_, _, result) = r[idx];
+                result
+            }
+            Err(_) => 0
+        }
+    }\n
+""")
+
+    emit_table(f, "combining_class_table", combine, "&'static [(char, char, u8)]", is_pub=False,
+            pfun=lambda x: "(%s,%s,%s)" % (escape_char(x[0]), escape_char(x[1]), x[2]))
+
+    f.write("""    #[deprecated(reason = "use the crates.io `unicode-normalization` lib instead",
+                 since = "1.0.0")]
+    #[unstable(feature = "unicode",
+               reason = "this functionality will be moved to crates.io")]
+    pub fn canonical_combining_class(c: char) -> u8 {
+        bsearch_range_value_table(c, combining_class_table)
+    }
+
+}
+
+""")
+
 def remove_from_wtable(wtable, val):
     wtable_out = []
     while wtable:
@@ -536,3 +649,53 @@ def optimize_width_table(wtable):
         # normalizations and conversions module
         emit_norm_module(rf, canon_decomp, compat_decomp, combines, norm_props)
         emit_conversions_module(rf, to_upper, to_lower, to_title)
+
+        ### character width module
+        width_table = []
+        for zwcat in ["Me", "Mn", "Cf"]:
+            width_table.extend(map(lambda (lo, hi): (lo, hi, 0, 0), gencats[zwcat]))
+        width_table.append((4448, 4607, 0, 0))
+
+        # get widths, except those that are explicitly marked zero-width above
+        ea_widths = load_east_asian_width(["W", "F", "A"], ["Me", "Mn", "Cf"])
+        # these are doublewidth
+        for dwcat in ["W", "F"]:
+            width_table.extend(map(lambda (lo, hi): (lo, hi, 2, 2), ea_widths[dwcat]))
+        width_table.extend(map(lambda (lo, hi): (lo, hi, 1, 2), ea_widths["A"]))
+
+        width_table.sort(key=lambda w: w[0])
+
+        # soft hyphen is not zero width in preformatted text; it's used to indicate
+        # a hyphen inserted to facilitate a linebreak.
+        width_table = remove_from_wtable(width_table, 173)
+
+        # optimize the width table by collapsing adjacent entities when possible
+        width_table = optimize_width_table(width_table)
+        emit_charwidth_module(rf, width_table)
+
+        ### grapheme cluster module
+        # from http://www.unicode.org/reports/tr29/#Grapheme_Cluster_Break_Property_Values
+        grapheme_cats = load_properties("auxiliary/GraphemeBreakProperty.txt", [])
+
+        # Control
+        #  Note 1:
+        # This category also includes Cs (surrogate codepoints), but Rust's `char`s are
+        # Unicode Scalar Values only, and surrogates are thus invalid `char`s.
+        # Thus, we have to remove Cs from the Control category
+        #  Note 2:
+        # 0x0a and 0x0d (CR and LF) are not in the Control category for Graphemes.
+        # However, the Graphemes iterator treats these as a special case, so they
+        # should be included in grapheme_cats["Control"] for our implementation.
+        grapheme_cats["Control"] = group_cat(list(
+            (set(ungroup_cat(grapheme_cats["Control"]))
+             | set(ungroup_cat(grapheme_cats["CR"]))
+             | set(ungroup_cat(grapheme_cats["LF"])))
+            - set(ungroup_cat([surrogate_codepoints]))))
+        del(grapheme_cats["CR"])
+        del(grapheme_cats["LF"])
+
+        grapheme_table = []
+        for cat in grapheme_cats:
+            grapheme_table.extend([(x, y, cat) for (x, y) in grapheme_cats[cat]])
+        grapheme_table.sort(key=lambda w: w[0])
+        emit_grapheme_module(rf, grapheme_table, grapheme_cats.keys())

trunk/src/liballoc/arc.rs

@@ -272,6 +272,18 @@ impl<T: ?Sized> Arc<T> {
     }
 }
 
+/// Get the number of weak references to this value.
+#[inline]
+#[unstable(feature = "arc_counts")]
+#[deprecated(since = "1.2.0", reason = "renamed to Arc::weak_count")]
+pub fn weak_count<T: ?Sized>(this: &Arc<T>) -> usize { Arc::weak_count(this) }
+
+/// Get the number of strong references to this value.
+#[inline]
+#[unstable(feature = "arc_counts")]
+#[deprecated(since = "1.2.0", reason = "renamed to Arc::strong_count")]
+pub fn strong_count<T: ?Sized>(this: &Arc<T>) -> usize { Arc::strong_count(this) }
+
 #[stable(feature = "rust1", since = "1.0.0")]
 impl<T: ?Sized> Clone for Arc<T> {
     /// Makes a clone of the `Arc<T>`.
@@ -472,6 +484,13 @@ impl<T: ?Sized> Arc<T> {
     }
 }
 
+#[inline]
+#[unstable(feature = "arc_unique")]
+#[deprecated(since = "1.2", reason = "use Arc::get_mut instead")]
+pub fn get_mut<T: ?Sized>(this: &mut Arc<T>) -> Option<&mut T> {
+    Arc::get_mut(this)
+}
+
 #[stable(feature = "rust1", since = "1.0.0")]
 impl<T: ?Sized> Drop for Arc<T> {
     /// Drops the `Arc<T>`.
@@ -841,7 +860,7 @@ mod tests {
     use std::sync::atomic::Ordering::{Acquire, SeqCst};
     use std::thread;
     use std::vec::Vec;
-    use super::{Arc, Weak};
+    use super::{Arc, Weak, get_mut, weak_count, strong_count};
     use std::sync::Mutex;
 
     struct Canary(*mut atomic::AtomicUsize);
@@ -879,39 +898,43 @@ mod tests {
 
     #[test]
     fn test_arc_get_mut() {
-        let mut x = Arc::new(3);
-        *Arc::get_mut(&mut x).unwrap() = 4;
-        assert_eq!(*x, 4);
-        let y = x.clone();
-        assert!(Arc::get_mut(&mut x).is_none());
-        drop(y);
-        assert!(Arc::get_mut(&mut x).is_some());
-        let _w = x.downgrade();
-        assert!(Arc::get_mut(&mut x).is_none());
+        unsafe {
+            let mut x = Arc::new(3);
+            *get_mut(&mut x).unwrap() = 4;
+            assert_eq!(*x, 4);
+            let y = x.clone();
+            assert!(get_mut(&mut x).is_none());
+            drop(y);
+            assert!(get_mut(&mut x).is_some());
+            let _w = x.downgrade();
+            assert!(get_mut(&mut x).is_none());
+        }
     }
 
     #[test]
     fn test_cowarc_clone_make_unique() {
-        let mut cow0 = Arc::new(75);
-        let mut cow1 = cow0.clone();
-        let mut cow2 = cow1.clone();
-
-        assert!(75 == *Arc::make_unique(&mut cow0));
-        assert!(75 == *Arc::make_unique(&mut cow1));
-        assert!(75 == *Arc::make_unique(&mut cow2));
-
-        *Arc::make_unique(&mut cow0) += 1;
-        *Arc::make_unique(&mut cow1) += 2;
-        *Arc::make_unique(&mut cow2) += 3;
-
-        assert!(76 == *cow0);
-        assert!(77 == *cow1);
-        assert!(78 == *cow2);
-
-        // none should point to the same backing memory
-        assert!(*cow0 != *cow1);
-        assert!(*cow0 != *cow2);
-        assert!(*cow1 != *cow2);
+        unsafe {
+            let mut cow0 = Arc::new(75);
+            let mut cow1 = cow0.clone();
+            let mut cow2 = cow1.clone();
+
+            assert!(75 == *Arc::make_unique(&mut cow0));
+            assert!(75 == *Arc::make_unique(&mut cow1));
+            assert!(75 == *Arc::make_unique(&mut cow2));
+
+            *Arc::make_unique(&mut cow0) += 1;
+            *Arc::make_unique(&mut cow1) += 2;
+            *Arc::make_unique(&mut cow2) += 3;
+
+            assert!(76 == *cow0);
+            assert!(77 == *cow1);
+            assert!(78 == *cow2);
+
+            // none should point to the same backing memory
+            assert!(*cow0 != *cow1);
+            assert!(*cow0 != *cow2);
+            assert!(*cow1 != *cow2);
+        }
     }
 
     #[test]
@@ -924,7 +947,9 @@ mod tests {
         assert!(75 == *cow1);
         assert!(75 == *cow2);
 
-        *Arc::make_unique(&mut cow0) += 1;
+        unsafe {
+            *Arc::make_unique(&mut cow0) += 1;
+        }
 
         assert!(76 == *cow0);
         assert!(75 == *cow1);
@@ -945,7 +970,9 @@ mod tests {
         assert!(75 == *cow0);
         assert!(75 == *cow1_weak.upgrade().unwrap());
 
-        *Arc::make_unique(&mut cow0) += 1;
+        unsafe {
+            *Arc::make_unique(&mut cow0) += 1;
+        }
 
         assert!(76 == *cow0);
         assert!(cow1_weak.upgrade().is_none());
@@ -1001,40 +1028,40 @@ mod tests {
     #[test]
     fn test_strong_count() {
         let a = Arc::new(0u32);
-        assert!(Arc::strong_count(&a) == 1);
+        assert!(strong_count(&a) == 1);
         let w = a.downgrade();
-        assert!(Arc::strong_count(&a) == 1);
+        assert!(strong_count(&a) == 1);
         let b = w.upgrade().expect("");
-        assert!(Arc::strong_count(&b) == 2);
-        assert!(Arc::strong_count(&a) == 2);
+        assert!(strong_count(&b) == 2);
+        assert!(strong_count(&a) == 2);
         drop(w);
         drop(a);
-        assert!(Arc::strong_count(&b) == 1);
+        assert!(strong_count(&b) == 1);
         let c = b.clone();
-        assert!(Arc::strong_count(&b) == 2);
-        assert!(Arc::strong_count(&c) == 2);
+        assert!(strong_count(&b) == 2);
+        assert!(strong_count(&c) == 2);
     }
 
     #[test]
     fn test_weak_count() {
         let a = Arc::new(0u32);
-        assert!(Arc::strong_count(&a) == 1);
-        assert!(Arc::weak_count(&a) == 0);
+        assert!(strong_count(&a) == 1);
+        assert!(weak_count(&a) == 0);
         let w = a.downgrade();
-        assert!(Arc::strong_count(&a) == 1);
-        assert!(Arc::weak_count(&a) == 1);
+        assert!(strong_count(&a) == 1);
+        assert!(weak_count(&a) == 1);
         let x = w.clone();
-        assert!(Arc::weak_count(&a) == 2);
+        assert!(weak_count(&a) == 2);
         drop(w);
         drop(x);
-        assert!(Arc::strong_count(&a) == 1);
-        assert!(Arc::weak_count(&a) == 0);
+        assert!(strong_count(&a) == 1);
+        assert!(weak_count(&a) == 0);
         let c = a.clone();
-        assert!(Arc::strong_count(&a) == 2);
-        assert!(Arc::weak_count(&a) == 0);
+        assert!(strong_count(&a) == 2);
+        assert!(weak_count(&a) == 0);
         let d = c.downgrade();
-        assert!(Arc::weak_count(&c) == 1);
-        assert!(Arc::strong_count(&c) == 2);
+        assert!(weak_count(&c) == 1);
+        assert!(strong_count(&c) == 2);
 
         drop(a);
         drop(c);