Correctly map upper-case Sigma to lower-case in word-final position. Fix #26035.

diff --git a/src/etc/unicode.py b/src/etc/unicode.py
index e39196b21d977e68efcb870d001b7e5d8480cb88..f580127cddaf8386515919b10902772a45aacbcf 100755 (executable)
--- a/src/etc/unicode.py
+++ b/src/etc/unicode.py
@@ -638,7 +638,8 @@ pub const UNICODE_VERSION: (u64, u64, u64) = (%s, %s, %s);
         (canon_decomp, compat_decomp, gencats, combines,
                 to_upper, to_lower, to_title) = load_unicode_data("UnicodeData.txt")
         load_special_casing("SpecialCasing.txt", to_upper, to_lower, to_title)
-        want_derived = ["XID_Start", "XID_Continue", "Alphabetic", "Lowercase", "Uppercase"]
+        want_derived = ["XID_Start", "XID_Continue", "Alphabetic", "Lowercase", "Uppercase",
+                        "Cased", "Case_Ignorable"]
         derived = load_properties("DerivedCoreProperties.txt", want_derived)
         scripts = load_properties("Scripts.txt", [])
         props = load_properties("PropList.txt",

diff --git a/src/libcollections/str.rs b/src/libcollections/str.rs
index 95bf6d0a9cd609c3f6abb297dcf04050ceaed8f1..913c0bd05325a50d1aa51b3c4b68fb526f94ed51 100644 (file)
--- a/src/libcollections/str.rs
+++ b/src/libcollections/str.rs
@@ -1854,8 +1854,32 @@ pub fn trim_right(&self) -> &str {
     #[unstable(feature = "collections")]
     pub fn to_lowercase(&self) -> String {
         let mut s = String::with_capacity(self.len());
-        s.extend(self[..].chars().flat_map(|c| c.to_lowercase()));
+        for (i, c) in self[..].char_indices() {
+            if c == 'Σ' {
+                map_uppercase_sigma(self, i, &mut s)
+            } else {
+                s.extend(c.to_lowercase());
+            }
+        }
         return s;
+
+        #[cold]
+        #[inline(never)]
+        fn map_uppercase_sigma(from: &str, i: usize, to: &mut String) {
+            debug_assert!('Σ'.len_utf8() == 2);
+            let is_word_final =
+                case_ignoreable_then_cased(from[..i].chars().rev()) &&
+                !case_ignoreable_then_cased(from[i + 2..].chars());
+            to.push_str(if is_word_final { "ς" } else { "σ" });
+        }
+
+        fn case_ignoreable_then_cased<I: Iterator<Item=char>>(iter: I) -> bool {
+            use rustc_unicode::derived_property::{Cased, Case_Ignorable};
+            match iter.skip_while(|&c| Case_Ignorable(c)).next() {
+                Some(c) => Cased(c),
+                None => false,
+            }
+        }
     }
 
     /// Returns the uppercase equivalent of this string.

diff --git a/src/libcollectionstest/str.rs b/src/libcollectionstest/str.rs
index 170f49ab15be553d5177122d4033da1e65bbaf85..0b7a4953edfe008771b371947e61912d9579d63c 100644 (file)
--- a/src/libcollectionstest/str.rs
+++ b/src/libcollectionstest/str.rs
@@ -1687,6 +1687,19 @@ fn trim_ws() {
                "");
 }
 
+#[test]
+fn to_lowercase() {
+    assert_eq!("".to_lowercase(), "");
+    // https://github.com/rust-lang/rust/issues/26035
+    assert_eq!("'Σ AÉΣ'Σ'' Σ DžΣ".to_lowercase(), "'σ aéσ'ς'' σ džς");
+}
+
+#[test]
+fn to_uppercase() {
+    assert_eq!("".to_uppercase(), "");
+    assert_eq!("aéDžßfiᾀ".to_uppercase(), "AÉDŽSSFIἈΙ");
+}
+
 mod pattern {
     use std::str::pattern::Pattern;
     use std::str::pattern::{Searcher, ReverseSearcher};

diff --git a/src/librustc_unicode/lib.rs b/src/librustc_unicode/lib.rs
index 2f72b940e328383af0d6d518d91bf7ba372eb701..a9456cb487c6a71d199414dfd9e251575715ed9c 100644 (file)
--- a/src/librustc_unicode/lib.rs
+++ b/src/librustc_unicode/lib.rs
@@ -49,3 +49,8 @@ pub mod str {
     pub use u_str::{utf8_char_width, is_utf16, Utf16Items, Utf16Item};
     pub use u_str::{utf16_items, Utf16Encoder};
 }
+
+// For use in libcollections, not re-exported in libstd.
+pub mod derived_property {
+    pub use tables::derived_property::{Cased, Case_Ignorable};
+}

diff --git a/src/librustc_unicode/tables.rs b/src/librustc_unicode/tables.rs
index 14e6478788ad03395e5b68a755a80ec89a389ceb..4ebb6a70d291c300cc588ca9aadc7e111be598d2 100644 (file)
--- a/src/librustc_unicode/tables.rs
+++ b/src/librustc_unicode/tables.rs
@@ -251,6 +251,157 @@ pub fn Alphabetic(c: char) -> bool {
         super::bsearch_range_table(c, Alphabetic_table)
     }
 
+    pub const Case_Ignorable_table: &'static [(char, char)] = &[
+        ('\u{27}', '\u{27}'), ('\u{2e}', '\u{2e}'), ('\u{3a}', '\u{3a}'), ('\u{5e}', '\u{5e}'),
+        ('\u{60}', '\u{60}'), ('\u{a8}', '\u{a8}'), ('\u{ad}', '\u{ad}'), ('\u{af}', '\u{af}'),
+        ('\u{b4}', '\u{b4}'), ('\u{b7}', '\u{b8}'), ('\u{2b0}', '\u{36f}'), ('\u{374}', '\u{375}'),
+        ('\u{37a}', '\u{37a}'), ('\u{384}', '\u{385}'), ('\u{387}', '\u{387}'), ('\u{483}',
+        '\u{489}'), ('\u{559}', '\u{559}'), ('\u{591}', '\u{5bd}'), ('\u{5bf}', '\u{5bf}'),
+        ('\u{5c1}', '\u{5c2}'), ('\u{5c4}', '\u{5c5}'), ('\u{5c7}', '\u{5c7}'), ('\u{5f4}',
+        '\u{5f4}'), ('\u{600}', '\u{605}'), ('\u{610}', '\u{61a}'), ('\u{61c}', '\u{61c}'),
+        ('\u{640}', '\u{640}'), ('\u{64b}', '\u{65f}'), ('\u{670}', '\u{670}'), ('\u{6d6}',
+        '\u{6dd}'), ('\u{6df}', '\u{6e8}'), ('\u{6ea}', '\u{6ed}'), ('\u{70f}', '\u{70f}'),
+        ('\u{711}', '\u{711}'), ('\u{730}', '\u{74a}'), ('\u{7a6}', '\u{7b0}'), ('\u{7eb}',
+        '\u{7f5}'), ('\u{7fa}', '\u{7fa}'), ('\u{816}', '\u{82d}'), ('\u{859}', '\u{85b}'),
+        ('\u{8e4}', '\u{902}'), ('\u{93a}', '\u{93a}'), ('\u{93c}', '\u{93c}'), ('\u{941}',
+        '\u{948}'), ('\u{94d}', '\u{94d}'), ('\u{951}', '\u{957}'), ('\u{962}', '\u{963}'),
+        ('\u{971}', '\u{971}'), ('\u{981}', '\u{981}'), ('\u{9bc}', '\u{9bc}'), ('\u{9c1}',
+        '\u{9c4}'), ('\u{9cd}', '\u{9cd}'), ('\u{9e2}', '\u{9e3}'), ('\u{a01}', '\u{a02}'),
+        ('\u{a3c}', '\u{a3c}'), ('\u{a41}', '\u{a42}'), ('\u{a47}', '\u{a48}'), ('\u{a4b}',
+        '\u{a4d}'), ('\u{a51}', '\u{a51}'), ('\u{a70}', '\u{a71}'), ('\u{a75}', '\u{a75}'),
+        ('\u{a81}', '\u{a82}'), ('\u{abc}', '\u{abc}'), ('\u{ac1}', '\u{ac5}'), ('\u{ac7}',
+        '\u{ac8}'), ('\u{acd}', '\u{acd}'), ('\u{ae2}', '\u{ae3}'), ('\u{b01}', '\u{b01}'),
+        ('\u{b3c}', '\u{b3c}'), ('\u{b3f}', '\u{b3f}'), ('\u{b41}', '\u{b44}'), ('\u{b4d}',
+        '\u{b4d}'), ('\u{b56}', '\u{b56}'), ('\u{b62}', '\u{b63}'), ('\u{b82}', '\u{b82}'),
+        ('\u{bc0}', '\u{bc0}'), ('\u{bcd}', '\u{bcd}'), ('\u{c00}', '\u{c00}'), ('\u{c3e}',
+        '\u{c40}'), ('\u{c46}', '\u{c48}'), ('\u{c4a}', '\u{c4d}'), ('\u{c55}', '\u{c56}'),
+        ('\u{c62}', '\u{c63}'), ('\u{c81}', '\u{c81}'), ('\u{cbc}', '\u{cbc}'), ('\u{cbf}',
+        '\u{cbf}'), ('\u{cc6}', '\u{cc6}'), ('\u{ccc}', '\u{ccd}'), ('\u{ce2}', '\u{ce3}'),
+        ('\u{d01}', '\u{d01}'), ('\u{d41}', '\u{d44}'), ('\u{d4d}', '\u{d4d}'), ('\u{d62}',
+        '\u{d63}'), ('\u{dca}', '\u{dca}'), ('\u{dd2}', '\u{dd4}'), ('\u{dd6}', '\u{dd6}'),
+        ('\u{e31}', '\u{e31}'), ('\u{e34}', '\u{e3a}'), ('\u{e46}', '\u{e4e}'), ('\u{eb1}',
+        '\u{eb1}'), ('\u{eb4}', '\u{eb9}'), ('\u{ebb}', '\u{ebc}'), ('\u{ec6}', '\u{ec6}'),
+        ('\u{ec8}', '\u{ecd}'), ('\u{f18}', '\u{f19}'), ('\u{f35}', '\u{f35}'), ('\u{f37}',
+        '\u{f37}'), ('\u{f39}', '\u{f39}'), ('\u{f71}', '\u{f7e}'), ('\u{f80}', '\u{f84}'),
+        ('\u{f86}', '\u{f87}'), ('\u{f8d}', '\u{f97}'), ('\u{f99}', '\u{fbc}'), ('\u{fc6}',
+        '\u{fc6}'), ('\u{102d}', '\u{1030}'), ('\u{1032}', '\u{1037}'), ('\u{1039}', '\u{103a}'),
+        ('\u{103d}', '\u{103e}'), ('\u{1058}', '\u{1059}'), ('\u{105e}', '\u{1060}'), ('\u{1071}',
+        '\u{1074}'), ('\u{1082}', '\u{1082}'), ('\u{1085}', '\u{1086}'), ('\u{108d}', '\u{108d}'),
+        ('\u{109d}', '\u{109d}'), ('\u{10fc}', '\u{10fc}'), ('\u{135d}', '\u{135f}'), ('\u{1712}',
+        '\u{1714}'), ('\u{1732}', '\u{1734}'), ('\u{1752}', '\u{1753}'), ('\u{1772}', '\u{1773}'),
+        ('\u{17b4}', '\u{17b5}'), ('\u{17b7}', '\u{17bd}'), ('\u{17c6}', '\u{17c6}'), ('\u{17c9}',
+        '\u{17d3}'), ('\u{17d7}', '\u{17d7}'), ('\u{17dd}', '\u{17dd}'), ('\u{180b}', '\u{180e}'),
+        ('\u{1843}', '\u{1843}'), ('\u{18a9}', '\u{18a9}'), ('\u{1920}', '\u{1922}'), ('\u{1927}',
+        '\u{1928}'), ('\u{1932}', '\u{1932}'), ('\u{1939}', '\u{193b}'), ('\u{1a17}', '\u{1a18}'),
+        ('\u{1a1b}', '\u{1a1b}'), ('\u{1a56}', '\u{1a56}'), ('\u{1a58}', '\u{1a5e}'), ('\u{1a60}',
+        '\u{1a60}'), ('\u{1a62}', '\u{1a62}'), ('\u{1a65}', '\u{1a6c}'), ('\u{1a73}', '\u{1a7c}'),
+        ('\u{1a7f}', '\u{1a7f}'), ('\u{1aa7}', '\u{1aa7}'), ('\u{1ab0}', '\u{1abe}'), ('\u{1b00}',
+        '\u{1b03}'), ('\u{1b34}', '\u{1b34}'), ('\u{1b36}', '\u{1b3a}'), ('\u{1b3c}', '\u{1b3c}'),
+        ('\u{1b42}', '\u{1b42}'), ('\u{1b6b}', '\u{1b73}'), ('\u{1b80}', '\u{1b81}'), ('\u{1ba2}',
+        '\u{1ba5}'), ('\u{1ba8}', '\u{1ba9}'), ('\u{1bab}', '\u{1bad}'), ('\u{1be6}', '\u{1be6}'),
+        ('\u{1be8}', '\u{1be9}'), ('\u{1bed}', '\u{1bed}'), ('\u{1bef}', '\u{1bf1}'), ('\u{1c2c}',
+        '\u{1c33}'), ('\u{1c36}', '\u{1c37}'), ('\u{1c78}', '\u{1c7d}'), ('\u{1cd0}', '\u{1cd2}'),
+        ('\u{1cd4}', '\u{1ce0}'), ('\u{1ce2}', '\u{1ce8}'), ('\u{1ced}', '\u{1ced}'), ('\u{1cf4}',
+        '\u{1cf4}'), ('\u{1cf8}', '\u{1cf9}'), ('\u{1d2c}', '\u{1d6a}'), ('\u{1d78}', '\u{1d78}'),
+        ('\u{1d9b}', '\u{1df5}'), ('\u{1dfc}', '\u{1dff}'), ('\u{1fbd}', '\u{1fbd}'), ('\u{1fbf}',
+        '\u{1fc1}'), ('\u{1fcd}', '\u{1fcf}'), ('\u{1fdd}', '\u{1fdf}'), ('\u{1fed}', '\u{1fef}'),
+        ('\u{1ffd}', '\u{1ffe}'), ('\u{200b}', '\u{200f}'), ('\u{2018}', '\u{2019}'), ('\u{2024}',
+        '\u{2024}'), ('\u{2027}', '\u{2027}'), ('\u{202a}', '\u{202e}'), ('\u{2060}', '\u{2064}'),
+        ('\u{2066}', '\u{206f}'), ('\u{2071}', '\u{2071}'), ('\u{207f}', '\u{207f}'), ('\u{2090}',
+        '\u{209c}'), ('\u{20d0}', '\u{20f0}'), ('\u{2c7c}', '\u{2c7d}'), ('\u{2cef}', '\u{2cf1}'),
+        ('\u{2d6f}', '\u{2d6f}'), ('\u{2d7f}', '\u{2d7f}'), ('\u{2de0}', '\u{2dff}'), ('\u{2e2f}',
+        '\u{2e2f}'), ('\u{3005}', '\u{3005}'), ('\u{302a}', '\u{302d}'), ('\u{3031}', '\u{3035}'),
+        ('\u{303b}', '\u{303b}'), ('\u{3099}', '\u{309e}'), ('\u{30fc}', '\u{30fe}'), ('\u{a015}',
+        '\u{a015}'), ('\u{a4f8}', '\u{a4fd}'), ('\u{a60c}', '\u{a60c}'), ('\u{a66f}', '\u{a672}'),
+        ('\u{a674}', '\u{a67d}'), ('\u{a67f}', '\u{a67f}'), ('\u{a69c}', '\u{a69d}'), ('\u{a69f}',
+        '\u{a69f}'), ('\u{a6f0}', '\u{a6f1}'), ('\u{a700}', '\u{a721}'), ('\u{a770}', '\u{a770}'),
+        ('\u{a788}', '\u{a78a}'), ('\u{a7f8}', '\u{a7f9}'), ('\u{a802}', '\u{a802}'), ('\u{a806}',
+        '\u{a806}'), ('\u{a80b}', '\u{a80b}'), ('\u{a825}', '\u{a826}'), ('\u{a8c4}', '\u{a8c4}'),
+        ('\u{a8e0}', '\u{a8f1}'), ('\u{a926}', '\u{a92d}'), ('\u{a947}', '\u{a951}'), ('\u{a980}',
+        '\u{a982}'), ('\u{a9b3}', '\u{a9b3}'), ('\u{a9b6}', '\u{a9b9}'), ('\u{a9bc}', '\u{a9bc}'),
+        ('\u{a9cf}', '\u{a9cf}'), ('\u{a9e5}', '\u{a9e6}'), ('\u{aa29}', '\u{aa2e}'), ('\u{aa31}',
+        '\u{aa32}'), ('\u{aa35}', '\u{aa36}'), ('\u{aa43}', '\u{aa43}'), ('\u{aa4c}', '\u{aa4c}'),
+        ('\u{aa70}', '\u{aa70}'), ('\u{aa7c}', '\u{aa7c}'), ('\u{aab0}', '\u{aab0}'), ('\u{aab2}',
+        '\u{aab4}'), ('\u{aab7}', '\u{aab8}'), ('\u{aabe}', '\u{aabf}'), ('\u{aac1}', '\u{aac1}'),
+        ('\u{aadd}', '\u{aadd}'), ('\u{aaec}', '\u{aaed}'), ('\u{aaf3}', '\u{aaf4}'), ('\u{aaf6}',
+        '\u{aaf6}'), ('\u{ab5b}', '\u{ab5f}'), ('\u{abe5}', '\u{abe5}'), ('\u{abe8}', '\u{abe8}'),
+        ('\u{abed}', '\u{abed}'), ('\u{fb1e}', '\u{fb1e}'), ('\u{fbb2}', '\u{fbc1}'), ('\u{fe00}',
+        '\u{fe0f}'), ('\u{fe13}', '\u{fe13}'), ('\u{fe20}', '\u{fe2d}'), ('\u{fe52}', '\u{fe52}'),
+        ('\u{fe55}', '\u{fe55}'), ('\u{feff}', '\u{feff}'), ('\u{ff07}', '\u{ff07}'), ('\u{ff0e}',
+        '\u{ff0e}'), ('\u{ff1a}', '\u{ff1a}'), ('\u{ff3e}', '\u{ff3e}'), ('\u{ff40}', '\u{ff40}'),
+        ('\u{ff70}', '\u{ff70}'), ('\u{ff9e}', '\u{ff9f}'), ('\u{ffe3}', '\u{ffe3}'), ('\u{fff9}',
+        '\u{fffb}'), ('\u{101fd}', '\u{101fd}'), ('\u{102e0}', '\u{102e0}'), ('\u{10376}',
+        '\u{1037a}'), ('\u{10a01}', '\u{10a03}'), ('\u{10a05}', '\u{10a06}'), ('\u{10a0c}',
+        '\u{10a0f}'), ('\u{10a38}', '\u{10a3a}'), ('\u{10a3f}', '\u{10a3f}'), ('\u{10ae5}',
+        '\u{10ae6}'), ('\u{11001}', '\u{11001}'), ('\u{11038}', '\u{11046}'), ('\u{1107f}',
+        '\u{11081}'), ('\u{110b3}', '\u{110b6}'), ('\u{110b9}', '\u{110ba}'), ('\u{110bd}',
+        '\u{110bd}'), ('\u{11100}', '\u{11102}'), ('\u{11127}', '\u{1112b}'), ('\u{1112d}',
+        '\u{11134}'), ('\u{11173}', '\u{11173}'), ('\u{11180}', '\u{11181}'), ('\u{111b6}',
+        '\u{111be}'), ('\u{1122f}', '\u{11231}'), ('\u{11234}', '\u{11234}'), ('\u{11236}',
+        '\u{11237}'), ('\u{112df}', '\u{112df}'), ('\u{112e3}', '\u{112ea}'), ('\u{11301}',
+        '\u{11301}'), ('\u{1133c}', '\u{1133c}'), ('\u{11340}', '\u{11340}'), ('\u{11366}',
+        '\u{1136c}'), ('\u{11370}', '\u{11374}'), ('\u{114b3}', '\u{114b8}'), ('\u{114ba}',
+        '\u{114ba}'), ('\u{114bf}', '\u{114c0}'), ('\u{114c2}', '\u{114c3}'), ('\u{115b2}',
+        '\u{115b5}'), ('\u{115bc}', '\u{115bd}'), ('\u{115bf}', '\u{115c0}'), ('\u{11633}',
+        '\u{1163a}'), ('\u{1163d}', '\u{1163d}'), ('\u{1163f}', '\u{11640}'), ('\u{116ab}',
+        '\u{116ab}'), ('\u{116ad}', '\u{116ad}'), ('\u{116b0}', '\u{116b5}'), ('\u{116b7}',
+        '\u{116b7}'), ('\u{16af0}', '\u{16af4}'), ('\u{16b30}', '\u{16b36}'), ('\u{16b40}',
+        '\u{16b43}'), ('\u{16f8f}', '\u{16f9f}'), ('\u{1bc9d}', '\u{1bc9e}'), ('\u{1bca0}',
+        '\u{1bca3}'), ('\u{1d167}', '\u{1d169}'), ('\u{1d173}', '\u{1d182}'), ('\u{1d185}',
+        '\u{1d18b}'), ('\u{1d1aa}', '\u{1d1ad}'), ('\u{1d242}', '\u{1d244}'), ('\u{1e8d0}',
+        '\u{1e8d6}'), ('\u{e0001}', '\u{e0001}'), ('\u{e0020}', '\u{e007f}'), ('\u{e0100}',
+        '\u{e01ef}')
+    ];
+
+    pub fn Case_Ignorable(c: char) -> bool {
+        super::bsearch_range_table(c, Case_Ignorable_table)
+    }
+
+    pub const Cased_table: &'static [(char, char)] = &[
+        ('\u{41}', '\u{5a}'), ('\u{61}', '\u{7a}'), ('\u{aa}', '\u{aa}'), ('\u{b5}', '\u{b5}'),
+        ('\u{ba}', '\u{ba}'), ('\u{c0}', '\u{d6}'), ('\u{d8}', '\u{f6}'), ('\u{f8}', '\u{1ba}'),
+        ('\u{1bc}', '\u{1bf}'), ('\u{1c4}', '\u{293}'), ('\u{295}', '\u{2b8}'), ('\u{2c0}',
+        '\u{2c1}'), ('\u{2e0}', '\u{2e4}'), ('\u{345}', '\u{345}'), ('\u{370}', '\u{373}'),
+        ('\u{376}', '\u{377}'), ('\u{37a}', '\u{37d}'), ('\u{37f}', '\u{37f}'), ('\u{386}',
+        '\u{386}'), ('\u{388}', '\u{38a}'), ('\u{38c}', '\u{38c}'), ('\u{38e}', '\u{3a1}'),
+        ('\u{3a3}', '\u{3f5}'), ('\u{3f7}', '\u{481}'), ('\u{48a}', '\u{52f}'), ('\u{531}',
+        '\u{556}'), ('\u{561}', '\u{587}'), ('\u{10a0}', '\u{10c5}'), ('\u{10c7}', '\u{10c7}'),
+        ('\u{10cd}', '\u{10cd}'), ('\u{1d00}', '\u{1dbf}'), ('\u{1e00}', '\u{1f15}'), ('\u{1f18}',
+        '\u{1f1d}'), ('\u{1f20}', '\u{1f45}'), ('\u{1f48}', '\u{1f4d}'), ('\u{1f50}', '\u{1f57}'),
+        ('\u{1f59}', '\u{1f59}'), ('\u{1f5b}', '\u{1f5b}'), ('\u{1f5d}', '\u{1f5d}'), ('\u{1f5f}',
+        '\u{1f7d}'), ('\u{1f80}', '\u{1fb4}'), ('\u{1fb6}', '\u{1fbc}'), ('\u{1fbe}', '\u{1fbe}'),
+        ('\u{1fc2}', '\u{1fc4}'), ('\u{1fc6}', '\u{1fcc}'), ('\u{1fd0}', '\u{1fd3}'), ('\u{1fd6}',
+        '\u{1fdb}'), ('\u{1fe0}', '\u{1fec}'), ('\u{1ff2}', '\u{1ff4}'), ('\u{1ff6}', '\u{1ffc}'),
+        ('\u{2071}', '\u{2071}'), ('\u{207f}', '\u{207f}'), ('\u{2090}', '\u{209c}'), ('\u{2102}',
+        '\u{2102}'), ('\u{2107}', '\u{2107}'), ('\u{210a}', '\u{2113}'), ('\u{2115}', '\u{2115}'),
+        ('\u{2119}', '\u{211d}'), ('\u{2124}', '\u{2124}'), ('\u{2126}', '\u{2126}'), ('\u{2128}',
+        '\u{2128}'), ('\u{212a}', '\u{212d}'), ('\u{212f}', '\u{2134}'), ('\u{2139}', '\u{2139}'),
+        ('\u{213c}', '\u{213f}'), ('\u{2145}', '\u{2149}'), ('\u{214e}', '\u{214e}'), ('\u{2160}',
+        '\u{217f}'), ('\u{2183}', '\u{2184}'), ('\u{24b6}', '\u{24e9}'), ('\u{2c00}', '\u{2c2e}'),
+        ('\u{2c30}', '\u{2c5e}'), ('\u{2c60}', '\u{2ce4}'), ('\u{2ceb}', '\u{2cee}'), ('\u{2cf2}',
+        '\u{2cf3}'), ('\u{2d00}', '\u{2d25}'), ('\u{2d27}', '\u{2d27}'), ('\u{2d2d}', '\u{2d2d}'),
+        ('\u{a640}', '\u{a66d}'), ('\u{a680}', '\u{a69d}'), ('\u{a722}', '\u{a787}'), ('\u{a78b}',
+        '\u{a78e}'), ('\u{a790}', '\u{a7ad}'), ('\u{a7b0}', '\u{a7b1}'), ('\u{a7f8}', '\u{a7fa}'),
+        ('\u{ab30}', '\u{ab5a}'), ('\u{ab5c}', '\u{ab5f}'), ('\u{ab64}', '\u{ab65}'), ('\u{fb00}',
+        '\u{fb06}'), ('\u{fb13}', '\u{fb17}'), ('\u{ff21}', '\u{ff3a}'), ('\u{ff41}', '\u{ff5a}'),
+        ('\u{10400}', '\u{1044f}'), ('\u{118a0}', '\u{118df}'), ('\u{1d400}', '\u{1d454}'),
+        ('\u{1d456}', '\u{1d49c}'), ('\u{1d49e}', '\u{1d49f}'), ('\u{1d4a2}', '\u{1d4a2}'),
+        ('\u{1d4a5}', '\u{1d4a6}'), ('\u{1d4a9}', '\u{1d4ac}'), ('\u{1d4ae}', '\u{1d4b9}'),
+        ('\u{1d4bb}', '\u{1d4bb}'), ('\u{1d4bd}', '\u{1d4c3}'), ('\u{1d4c5}', '\u{1d505}'),
+        ('\u{1d507}', '\u{1d50a}'), ('\u{1d50d}', '\u{1d514}'), ('\u{1d516}', '\u{1d51c}'),
+        ('\u{1d51e}', '\u{1d539}'), ('\u{1d53b}', '\u{1d53e}'), ('\u{1d540}', '\u{1d544}'),
+        ('\u{1d546}', '\u{1d546}'), ('\u{1d54a}', '\u{1d550}'), ('\u{1d552}', '\u{1d6a5}'),
+        ('\u{1d6a8}', '\u{1d6c0}'), ('\u{1d6c2}', '\u{1d6da}'), ('\u{1d6dc}', '\u{1d6fa}'),
+        ('\u{1d6fc}', '\u{1d714}'), ('\u{1d716}', '\u{1d734}'), ('\u{1d736}', '\u{1d74e}'),
+        ('\u{1d750}', '\u{1d76e}'), ('\u{1d770}', '\u{1d788}'), ('\u{1d78a}', '\u{1d7a8}'),
+        ('\u{1d7aa}', '\u{1d7c2}'), ('\u{1d7c4}', '\u{1d7cb}'), ('\u{1f130}', '\u{1f149}'),
+        ('\u{1f150}', '\u{1f169}'), ('\u{1f170}', '\u{1f189}')
+    ];
+
+    pub fn Cased(c: char) -> bool {
+        super::bsearch_range_table(c, Cased_table)
+    }
+
     pub const Lowercase_table: &'static [(char, char)] = &[
         ('\u{61}', '\u{7a}'), ('\u{aa}', '\u{aa}'), ('\u{b5}', '\u{b5}'), ('\u{ba}', '\u{ba}'),
         ('\u{df}', '\u{f6}'), ('\u{f8}', '\u{ff}'), ('\u{101}', '\u{101}'), ('\u{103}', '\u{103}'),