Disallow mixing of Canadian Syllabary and [a-z]

components/url_formatter/url_formatter_unittest.cc

 // Copyright 2015 The Chromium Authors. All rights reserved.
 // Use of this source code is governed by a BSD-style license that can be
 // found in the LICENSE file.
 
 #include "components/url_formatter/url_formatter.h"
 
 #include <stddef.h>
 #include <string.h>
 
 #include <vector>
@@ skipping 107 matching lines @@
   {"xn--95e5de3ds.cn", L"\x1823\x1837\x1804\x1833\x1824.cn", false},
   // Miao/Pollad
   {"xn--2u0fpf0a.cn", L"\U00016f04\U00016f62\U00016f59.cn", true},
 
   // Script mixing tests
   // The following script combinations are allowed.
   // MODERATELY_RESTRICTIVE with Latin limited to ASCII-Latin.
   // ASCII-Latin + Japn (Kana + Han)
   // ASCII-Latin + Kore (Hangul + Han)
   // ASCII-Latin + Han + Bopomofo
-  // ASCII-Latin + any allowed script other than Cyrillic, Greek and Cherokee
+  // ASCII-Latin + any allowed script other than Cyrillic, Greek, Cherokee
+  // and Unified Canadian Syllabary
   // "payp<alpha>l.com"
   {"www.xn--paypl-g9d.com", L"payp\x03b1l.com", false},
   // google.gr with Greek omicron and epsilon
   {"xn--ggl-6xc1ca.gr", L"g\x03bf\x03bfgl\x03b5.gr", false},
   // google.ru with Cyrillic o
   {"xn--ggl-tdd6ba.ru", L"g\x043e\x043egl\x0435.ru", false},
   // h<e with acute>llo<China in Han>.cn
   {"xn--hllo-bpa7979ih5m.cn", L"h\x00e9llo\x4e2d\x56fd.cn", false},
   // <Greek rho><Cyrillic a><Cyrillic u>.ru
   {"xn--2xa6t2b.ru", L"\x03c1\x0430\x0443.ru", false},
@@ skipping 33 matching lines @@
   {"xn--ab-3ofh8fqbj6h.in", L"ab\x0939\x093f\x0928\x094d\x0926\x0940.in", true},
   // Thai + Latin
   {"xn--ab-jsi9al4bxdb6n.th",
     L"ab\x0e20\x0e32\x0e29\x0e32\x0e44\x0e17\x0e22.th", true},
   // <vitamin in Katakana>b1.com
   {"xn--b1-xi4a7cvc9f.com", L"\x30d3\x30bf\x30df\x30f3" L"b1.com", true},
   // Devanagari + Han
   {"xn--t2bes3ds6749n.com", L"\x0930\x094b\x0932\x0947\x76e7\x0938.com", false},
   // Devanagari + Bengali
   {"xn--11b0x.in", L"\x0915\x0995.in", false},
+  // Canadian Syllabary + Latin
+  {"xn--ab-lym.com", L"ab\x14BF.com", false},
+  {"xn--ab1-p6q.com", L"ab1\x14BF.com", false},
+  {"xn--1ab-m6qd.com", L"\x14BF" L"1ab.com", false},
+  {"xn--ab-jymc.com", L"\x14BF" L"ab.com", false},
 
   // Invisibility check
   // Thai tone mark malek(U+0E48) repeated
   {"xn--03c0b3ca.th", L"\x0e23\x0e35\x0e48\x0e48.th", false},
   // Accute accent repeated
   {"xn--a-xbba.com", L"a\x0301\x0301.com", false},
   // 'a' with acuted accent + another acute accent
   {"xn--1ca20i.com", L"\x00e1\x0301.com", false},
   // Combining mark at the beginning
   {"xn--abc-fdc.jp", L"\x0300" L"abc.jp", false},
@@ skipping 795 matching lines @@
     0, kNpos, kNpos, kNpos, kNpos, kNpos, kNpos, 0, kNpos, kNpos, kNpos, kNpos,
     0, 1, 2, 3, 4, 5, 6, 7
   };
   CheckAdjustedOffsets("http://user@foo.com/", kFormatUrlOmitAll,
                        net::UnescapeRule::NORMAL, omit_all_offsets);
 }
 
 }  // namespace
 
 }  // namespace url_formatter