4 # re/fold_grind.t has more complex tests, but doesn't test every fold
5 # This file also tests the fc() keyword.
11 skip_all_without_unicode_tables();
12 skip_all_if_miniperl("miniperl, no Unicode::Normalize");
13 require Config; import Config;
14 require './charset_tools.pl';
15 require './loc_tools.pl'; # Contains find_utf8_ctype_locale()
18 use feature 'unicode_strings';
19 use Unicode::UCD qw(all_casefolds);
21 binmode *STDOUT, ":utf8";
27 # Read in the official case folding definitions.
28 my $casefolds = all_casefolds();
36 foreach my $decimal_code_point (sort { $a <=> $b } keys %$casefolds) {
37 # We only use simple folds in fc(), since the regex engine uses full case
40 my $name = charnames::viacode($decimal_code_point);
41 my $type = $casefolds->{$decimal_code_point}{'status'};
42 my $code = $casefolds->{$decimal_code_point}{'code'};
43 my $simple = $casefolds->{$decimal_code_point}{'simple'};
44 my $full = $casefolds->{$decimal_code_point}{'full'};
46 if ($simple && $simple ne $full) { # If there is a distinction
47 push @simple_folds, [ $code, $simple, $type, $name ];
50 push @CF, [ $code, $full, $type, $name ];
52 # Get the inverse fold for single-char mappings.
53 $reverse_fold{pack "W*", hex $simple} = pack "W*", $decimal_code_point if $simple;
56 foreach my $test_ref ( @simple_folds ) {
58 my ($code, $mapping, $type, $name) = @$test_ref;
59 my $c = pack("W*", hex $code);
61 my $f = pack("W*", map { hex } split " ", $mapping);
63 my $against = join "", "qq{", map("\\x{$_}", split " ", $mapping), "}";
65 isnt(fc($c), $f, "$code - $name - $mapping - $type - Full casefolding, fc(\\x{$code}) ne $against");
66 isnt("\F$c", $f, "$code - $name - $mapping - $type - Full casefolding, qq{\\F\\x{$code}} ne $against");
70 foreach my $test_ref (@CF) {
71 my ($code, $mapping, $type, $name) = @$test_ref;
72 my $c = pack("W*", hex $code);
74 my $f = pack("W*", map { hex } split " ", $mapping);
75 my $f_length = length $f;
78 qq[":$c:" =~ /:$c:/i],
79 qq[":$c:" =~ /:[_$c]:/], # Place two chars in [] so doesn't get
80 # optimized to a non-charclass
81 qq[":$c:" =~ /:[_$c]:/i],
82 qq[":$c:" =~ /:$f:/i],
83 qq[":$f:" =~ /:$c:/i],
85 ok eval $test, "$code - $name - $mapping - $type - $test";
90 my $against = join "", "qq{", map("\\x{$_}", split " ", $mapping), "}";
92 "$code - $name - $mapping - $type - fc(\\x{$code}) eq $against");
93 is("\F$c", $f, "$code - $name - $mapping - $type - qq{\\F\\x{$code}} eq $against");
95 # And here we test bytes. For [A-Za-z0-9], the fold is the same as lc under
96 # bytes. For everything else, it's the bytes that formed the original string.
97 if ( $c =~ /[A-Za-z0-9]/ ) {
99 is(CORE::fc($c), lc($c), "$code - $name - fc and use bytes, ascii");
103 is($copy, do { use bytes; CORE::fc($c) }, "$code - $name - fc and use bytes");
106 # Certain tests weren't convenient to put in the list above since they are
107 # TODO's in multi-character folds.
108 if ($f_length == 1) {
110 # The qq loses the utf8ness of ":$f:". These tests are not about
111 # finding bugs in utf8ness, so make sure it's utf8.
112 my $test = qq[my \$s = ":$f:"; utf8::upgrade(\$s); \$s =~ /:[_$c]:/i];
113 ok eval $test, "$code - $name - $mapping - $type - $test";
114 $test = qq[":$c:" =~ /:[_$f]:/i];
115 ok eval $test, "$code - $name - $mapping - $type - $test";
119 # There are two classes of multi-char folds that need more work. For
121 # ":ß:" =~ /:[_s]{2}:/i
122 # ":ss:" =~ /:[_ß]:/i
124 # Some of the old tests for the second case happened to pass somewhat
125 # coincidentally. But none would pass if changed to this.
126 # ":SS:" =~ /:[_ß]:/i
128 # As the capital SS doesn't get folded. When those pass, it means
129 # that the code has been changed to take into account folding in the
130 # string, and all should pass, capitalized or not (this wouldn't be
131 # true for [^complemented character classes], for which the fold case
132 # is better, but these aren't used in this .t currently. So, what is
133 # done is to essentially upper-case the string for this class (but use
134 # the reverse fold not uc(), as that is more correct)
136 for my $i (0 .. $f_length - 1) {
137 my $cur_char = substr($f, $i, 1);
138 $u .= $reverse_fold{$cur_char} || $cur_char;
142 # A multi-char fold should not match just one char;
143 # e.g., ":ß:" !~ /:[_s]:/i
144 $test = qq[":$c:" !~ /:[_$f]:/i];
145 ok eval $test, "$code - $name - $mapping - $type - $test";
147 TODO: { # e.g., ":ß:" =~ /:[_s]{2}:/i
148 local $TODO = 'Multi-char fold in [character class]';
150 $test = qq[":$c:" =~ /:[_$f]{$f_length}:/i];
151 ok eval $test, "$code - $name - $mapping - $type - $test";
154 # e.g., ":SS:" =~ /:[_ß]:/i now pass, so TODO has been removed, but
155 # since they use '$u', they are left out of the main loop
156 $test = qq[ my \$s = ":$u:"; utf8::upgrade(\$s); \$s =~ /:[_$c]:/i];
157 ok eval $test, "$code - $name - $mapping - $type - $test";
164 # These three come from the ICU project's test suite, more especifically
165 # http://icu.sourcearchive.com/documentation/4.4~rc1-1/strcase_8cpp-source.html
167 my $s = "A\N{U+00df}\N{U+00b5}\N{U+fb03}\N{U+1040C}\N{U+0130}\N{U+0131}";
168 #\N{LATIN CAPITAL LETTER A}\N{LATIN SMALL LETTER SHARP S}\N{MICRO SIGN}\N{LATIN SMALL LIGATURE FFI}\N{DESERET CAPITAL LETTER AY}\N{LATIN CAPITAL LETTER I WITH DOT ABOVE}\N{LATIN SMALL LETTER DOTLESS I}
170 my $f = "ass\N{U+03bc}ffi\N{U+10434}i\N{U+0307}\N{U+0131}";
171 #\N{LATIN SMALL LETTER A}\N{LATIN SMALL LETTER S}\N{LATIN SMALL LETTER S}\N{GREEK SMALL LETTER MU}\N{LATIN SMALL LETTER F}\N{LATIN SMALL LETTER F}\N{LATIN SMALL LETTER I}\N{DESERET SMALL LETTER AY}\N{LATIN SMALL LETTER I}\N{COMBINING DOT ABOVE}\N{LATIN SMALL LETTER DOTLESS I}
173 is(fc($s), $f, "ICU's casefold test passes");
174 is("\F$s", $f, "ICU's casefold test passes");
176 is( fc("aBİIıϐßffi"), "abi̇iıβssffi" );
177 is( "\FaBİIıϐßffi", "abi̇iıβssffi" );
179 # local $::TODO = "turkic special cases";
180 # is( fc "aBİIıϐßffi", "abiııβssffi" );
183 # The next batch come from http://www.devdaily.com/java/jwarehouse/lucene/contrib/icu/src/test/org/apache/lucene/analysis/icu/TestICUFoldingFilter.java.shtml
184 # Except the article got most casings wrong. Or maybe Lucene does.
186 is( fc("This is a test"), "this is a test" );
187 is( fc("Ruß"), "russ" );
188 is( fc("ΜΆΪΟΣ"), "μάϊοσ" );
189 is( fc("Μάϊος"), "μάϊοσ" );
191 is( fc("r" . uni_to_native("\xe9") . "sum" . uni_to_native("\xe9")),
192 "r" . uni_to_native("\xe9") . "sum" . uni_to_native("\xe9") );
193 is( fc("re\x{0301}sume\x{0301}"), "re\x{301}sume\x{301}" );
194 is( fc("ELİF"), "eli\x{307}f" );
195 is( fc("eli\x{307}f"), "eli\x{307}f");
197 # This batch comes from
198 # http://www.java2s.com/Open-Source/Java-Document/Internationalization-Localization/icu4j/com/ibm/icu/dev/test/lang/UCharacterCaseTest.java.htm
199 # Which uses ICU as the backend.
201 my @folding_mixed = (
202 uni_to_native("\x{61}\x{42}\x{130}\x{49}\x{131}\x{3d0}\x{df}\x{fb03}"),
203 "A" . uni_to_native("\x{df}\x{b5}\x{fb03}\x{1040C}\x{130}\x{131}"),
206 my @folding_default = (
207 uni_to_native("\x{61}\x{62}\x{69}\x{307}\x{69}\x{131}\x{3b2}\x{73}\x{73}\x{66}\x{66}\x{69}"),
208 "ass\x{3bc}ffi\x{10434}i\x{307}\x{131}"
211 my @folding_exclude_turkic = (
212 uni_to_native("\x{61}\x{62}\x{69}\x{131}\x{131}\x{3b2}\x{73}\x{73}\x{66}\x{66}\x{69}"),
213 "ass\x{3bc}ffi\x{10434}i\x{131}",
216 is( fc($folding_mixed[1]), $folding_default[1] );
218 is( fc($folding_mixed[0]), $folding_default[0] );
224 # Table stolen from tchrist's mail in
225 # http://bugs.python.org/file23051/casing-tests.py
226 # and http://98.245.80.27/tcpc/OSCON2011/case-test.python3
227 # For reference, it's a longer version of what he posted here:
228 # http://stackoverflow.com/questions/6991038/case-insensitive-storage-and-unicode-compatibility
230 #Couple of repeats because I'm lazy, not tchrist's fault.
232 #This should probably go in t/op/lc.t
235 # ORIG LC_SIMPLE TC_SIMPLE UC_SIMPLE LC_FULL TC_FULL UC_FULL FC_SIMPLE FC_TURKIC FC_FULL
236 [ 'þǽr rihtes', 'þǽr rihtes', 'Þǽr Rihtes', 'ÞǼR RIHTES', 'þǽr rihtes', 'Þǽr Rihtes', 'ÞǼR RIHTES', 'þǽr rihtes', 'þǽr rihtes', 'þǽr rihtes', ],
237 [ 'duȝeðlice', 'duȝeðlice', 'Duȝeðlice', 'DUȜEÐLICE', 'duȝeðlice', 'Duȝeðlice', 'DUȜEÐLICE', 'duȝeðlice', 'duȝeðlice', 'duȝeðlice', ],
238 [ 'Ævar Arnfjörð Bjarmason', 'ævar arnfjörð bjarmason', 'Ævar Arnfjörð Bjarmason', 'ÆVAR ARNFJÖRÐ BJARMASON', 'ævar arnfjörð bjarmason', 'Ævar Arnfjörð Bjarmason', 'ÆVAR ARNFJÖRÐ BJARMASON', 'ævar arnfjörð bjarmason', 'ævar arnfjörð bjarmason', 'ævar arnfjörð bjarmason', ],
239 [ 'Кириллица', 'кириллица', 'Кириллица', 'КИРИЛЛИЦА', 'кириллица', 'Кириллица', 'КИРИЛЛИЦА', 'кириллица', 'кириллица', 'кириллица', ],
240 [ 'ij', 'ij', 'IJ', 'IJ', 'ij', 'IJ', 'IJ', 'ij', 'ij', 'ij', ],
241 [ 'Van Dijke', 'van dijke', 'Van Dijke', 'VAN DIJKE', 'van dijke', 'Van Dijke', 'VAN DIJKE', 'van dijke', 'van dijke', 'van dijke', ],
242 [ 'VAN DIJKE', 'van dijke', 'Van Dijke', 'VAN DIJKE', 'van dijke', 'Van Dijke', 'VAN DIJKE', 'van dijke', 'van dijke', 'van dijke', ],
243 [ 'efficient', 'efficient', 'Efficient', 'EffiCIENT', 'efficient', 'Efficient', 'EFFICIENT', 'efficient', 'efficient', 'efficient', ],
244 [ 'flour', 'flour', 'flour', 'flOUR', 'flour', 'Flour', 'FLOUR', 'flour', 'flour', 'flour', ],
245 [ 'flour and water', 'flour and water', 'flour And Water', 'flOUR AND WATER', 'flour and water', 'Flour And Water', 'FLOUR AND WATER', 'flour and water', 'flour and water', 'flour and water', ],
246 [ 'dzur', 'dzur', 'Dzur', 'DZUR', 'dzur', 'Dzur', 'DZUR', 'dzur', 'dzur', 'dzur', ],
247 [ 'Dzur', 'dzur', 'Dzur', 'DZUR', 'dzur', 'Dzur', 'DZUR', 'dzur', 'dzur', 'dzur', ],
248 [ 'DZUR', 'dzur', 'Dzur', 'DZUR', 'dzur', 'Dzur', 'DZUR', 'dzur', 'dzur', 'dzur', ],
249 [ 'dzur mountain', 'dzur mountain', 'Dzur Mountain', 'DZUR MOUNTAIN', 'dzur mountain', 'Dzur Mountain', 'DZUR MOUNTAIN', 'dzur mountain', 'dzur mountain', 'dzur mountain', ],
250 [ 'Dzur Mountain', 'dzur mountain', 'Dzur Mountain', 'DZUR MOUNTAIN', 'dzur mountain', 'Dzur Mountain', 'DZUR MOUNTAIN', 'dzur mountain', 'dzur mountain', 'dzur mountain', ],
251 [ 'DZUR MOUNTAIN', 'dzur mountain', 'Dzur Mountain', 'DZUR MOUNTAIN', 'dzur mountain', 'Dzur Mountain', 'DZUR MOUNTAIN', 'dzur mountain', 'dzur mountaın', 'dzur mountain', ],
252 [ 'poſt', 'poſt', 'Poſt', 'POST', 'poſt', 'Poſt', 'POST', 'post', 'post', 'post', ],
253 [ 'poſt', 'poſt', 'Poſt', 'POſt', 'poſt', 'Poſt', 'POST', 'poſt', 'post', 'post', ],
254 [ 'ſtop', 'ſtop', 'ſtop', 'ſtOP', 'ſtop', 'Stop', 'STOP', 'ſtop', 'stop', 'stop', ],
255 [ 'tschüß', 'tschüß', 'Tschüß', 'TSCHÜß', 'tschüß', 'Tschüß', 'TSCHÜSS', 'tschüß', 'tschüss', 'tschüss', ],
256 [ 'TSCHÜẞ', 'tschüß', 'Tschüß', 'TSCHÜẞ', 'tschüß', 'Tschüß', 'TSCHÜẞ', 'tschüß', 'tschüss', 'tschüss', ],
257 [ 'weiß', 'weiß', 'Weiß', 'WEIß', 'weiß', 'Weiß', 'WEISS', 'weiß', 'weiss', 'weiss', ],
258 [ 'WEIẞ', 'weiß', 'Weiß', 'WEIẞ', 'weiß', 'Weiß', 'WEIẞ', 'weiß', 'weıss', 'weiss', ],
259 [ 'ẞIEW', 'ßiew', 'ẞiew', 'ẞIEW', 'ßiew', 'ẞiew', 'ẞIEW', 'ßiew', 'ssıew', 'ssiew', ],
260 [ 'ᾲ', 'ᾲ', 'Ὰͅ', 'ᾺΙ', 'ᾲ', 'Ὰͅ', 'ᾺΙ', 'ὰι', 'ὰι', 'ὰι', ],
261 [ 'Ὰι', 'ὰι', 'Ὰι', 'ᾺΙ', 'ὰι', 'Ὰι', 'ᾺΙ', 'ὰι', 'ὰι', 'ὰι', ],
262 [ 'ᾺΙ', 'ὰι', 'Ὰι', 'ᾺΙ', 'ὰι', 'Ὰι', 'ᾺΙ', 'ὰι', 'ὰι', 'ὰι', ],
263 [ 'ᾲ', 'ᾲ', 'ᾲ', 'ᾲ', 'ᾲ', 'Ὰͅ', 'ᾺΙ', 'ᾲ', 'ὰι', 'ὰι', ],
264 [ 'Ὰͅ', 'ᾲ', 'Ὰͅ', 'ᾺΙ', 'ᾲ', 'Ὰͅ', 'ᾺΙ', 'ὰι', 'ὰι', 'ὰι', ],
265 [ 'ᾺΙ', 'ὰι', 'Ὰι', 'ᾺΙ', 'ὰι', 'Ὰι', 'ᾺΙ', 'ὰι', 'ὰι', 'ὰι', ],
266 [ 'ᾲ στο διάολο', 'ᾲ στο διάολο', 'ᾲ Στο Διάολο', 'ᾲ ΣΤΟ ΔΙΆΟΛΟ', 'ᾲ στο διάολο', 'Ὰͅ Στο Διάολο', 'ᾺΙ ΣΤΟ ΔΙΆΟΛΟ', 'ᾲ στο διάολο', 'ὰι στο διάολο', 'ὰι στο διάολο', ],
267 [ 'ᾲ στο διάολο', 'ᾲ στο διάολο', 'Ὰͅ Στο Διάολο', 'ᾺΙ ΣΤΟ ΔΙΆΟΛΟ', 'ᾲ στο διάολο', 'Ὰͅ Στο Διάολο', 'ᾺΙ ΣΤΟ ΔΙΆΟΛΟ', 'ὰι στο διάολο', 'ὰι στο διάολο', 'ὰι στο διάολο', ],
268 [ '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐇𐐝𐐀𐐡𐐇𐐓', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐇𐐝𐐀𐐡𐐇𐐓', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', ],
269 [ '𐐔𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐇𐐝𐐀𐐡𐐇𐐓', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐇𐐝𐐀𐐡𐐇𐐓', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', ],
270 [ '𐐔𐐇𐐝𐐀𐐡𐐇𐐓', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐇𐐝𐐀𐐡𐐇𐐓', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐯𐑅𐐨𐑉𐐯𐐻', '𐐔𐐇𐐝𐐀𐐡𐐇𐐓', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', '𐐼𐐯𐑅𐐨𐑉𐐯𐐻', ],
271 [ 'henry ⅷ', 'henry ⅷ', 'Henry Ⅷ', 'HENRY Ⅷ', 'henry ⅷ', 'Henry Ⅷ', 'HENRY Ⅷ', 'henry ⅷ', 'henry ⅷ', 'henry ⅷ', ],
272 [ 'Henry Ⅷ', 'henry ⅷ', 'Henry Ⅷ', 'HENRY Ⅷ', 'henry ⅷ', 'Henry Ⅷ', 'HENRY Ⅷ', 'henry ⅷ', 'henry ⅷ', 'henry ⅷ', ],
273 [ 'HENRY Ⅷ', 'henry ⅷ', 'Henry Ⅷ', 'HENRY Ⅷ', 'henry ⅷ', 'Henry Ⅷ', 'HENRY Ⅷ', 'henry ⅷ', 'henry ⅷ', 'henry ⅷ', ],
274 [ 'i work at ⓚ', 'i work at ⓚ', 'I Work At Ⓚ', 'I WORK AT Ⓚ', 'i work at ⓚ', 'I Work At Ⓚ', 'I WORK AT Ⓚ', 'i work at ⓚ', 'i work at ⓚ', 'i work at ⓚ', ],
275 [ 'I Work At Ⓚ', 'i work at ⓚ', 'I Work At Ⓚ', 'I WORK AT Ⓚ', 'i work at ⓚ', 'I Work At Ⓚ', 'I WORK AT Ⓚ', 'i work at ⓚ', 'ı work at ⓚ', 'i work at ⓚ', ],
276 [ 'I WORK AT Ⓚ', 'i work at ⓚ', 'I Work At Ⓚ', 'I WORK AT Ⓚ', 'i work at ⓚ', 'I Work At Ⓚ', 'I WORK AT Ⓚ', 'i work at ⓚ', 'ı work at ⓚ', 'i work at ⓚ', ],
277 [ 'istambul', 'istambul', 'Istambul', 'ISTAMBUL', 'istambul', 'Istambul', 'ISTAMBUL', 'istambul', 'istambul', 'istambul', ],
278 [ 'i̇stanbul', 'i̇stanbul', 'İstanbul', 'İSTANBUL', 'i̇stanbul', 'İstanbul', 'İSTANBUL', 'i̇stanbul', 'i̇stanbul', 'i̇stanbul', ],
279 [ 'İstanbul', 'i̇stanbul', 'İstanbul', 'İSTANBUL', 'i̇stanbul', 'İstanbul', 'İSTANBUL', 'i̇stanbul', 'ı̇stanbul', 'i̇stanbul', ],
280 [ 'İSTANBUL', 'istanbul', 'İstanbul', 'İSTANBUL', 'i̇stanbul', 'İstanbul', 'İSTANBUL', 'İstanbul', 'istanbul', 'i̇stanbul', ],
281 [ 'στιγμας', 'στιγμας', 'Στιγμας', 'ΣΤΙΓΜΑΣ', 'στιγμας', 'Στιγμας', 'ΣΤΙΓΜΑΣ', 'στιγμασ', 'στιγμασ', 'στιγμασ', ],
282 [ 'στιγμασ', 'στιγμασ', 'Στιγμασ', 'ΣΤΙΓΜΑΣ', 'στιγμασ', 'Στιγμασ', 'ΣΤΙΓΜΑΣ', 'στιγμασ', 'στιγμασ', 'στιγμασ', ],
283 [ 'ΣΤΙΓΜΑΣ', 'στιγμασ', 'Στιγμασ', 'ΣΤΙΓΜΑΣ', 'στιγμασ', 'Στιγμασ', 'ΣΤΙΓΜΑΣ', 'στιγμασ', 'στιγμασ', 'στιγμασ', ],
284 [ 'ʀᴀʀᴇ', 'ʀᴀʀᴇ', 'Ʀᴀʀᴇ', 'ƦᴀƦᴇ', 'ʀᴀʀᴇ', 'Ʀᴀʀᴇ', 'ƦᴀƦᴇ', 'ʀᴀʀᴇ', 'ʀᴀʀᴇ', 'ʀᴀʀᴇ', ],
285 [ 'Ʀᴀʀᴇ', 'ʀᴀʀᴇ', 'Ʀᴀʀᴇ', 'ƦᴀƦᴇ', 'ʀᴀʀᴇ', 'Ʀᴀʀᴇ', 'ƦᴀƦᴇ', 'ʀᴀʀᴇ', 'ʀᴀʀᴇ', 'ʀᴀʀᴇ', ],
286 [ 'ƦᴀƦᴇ', 'ʀᴀʀᴇ', 'Ʀᴀʀᴇ', 'ƦᴀƦᴇ', 'ʀᴀʀᴇ', 'Ʀᴀʀᴇ', 'ƦᴀƦᴇ', 'ʀᴀʀᴇ', 'ʀᴀʀᴇ', 'ʀᴀʀᴇ', ],
287 [ 'Ԧԧ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'ԧԧ', 'ԧԧ', ],
288 [ 'ԧԧ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'ԧԧ', 'ԧԧ', ],
289 [ 'Ԧԧ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'ԧԧ', 'ԧԧ', ],
290 [ 'ԦԦ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'Ԧԧ', 'ԦԦ', 'ԧԧ', 'ԧԧ', 'ԧԧ', ],
291 [ "þǽr rihtes", "þǽr rihtes", "Þǽr Rihtes", "ÞǼR RIHTES", "þǽr rihtes", "Þǽr Rihtes", "ÞǼR RIHTES", "þǽr rihtes", "þǽr rihtes", "þǽr rihtes", ],
292 [ "duȝeðlice", "duȝeðlice", "Duȝeðlice", "DUȜEÐLICE", "duȝeðlice", "Duȝeðlice", "DUȜEÐLICE", "duȝeðlice", "duȝeðlice", "duȝeðlice", ],
293 [ "Van Dijke", "van dijke", "Van Dijke", "VAN DIJKE", "van dijke", "Van Dijke", "VAN DIJKE", "van dijke", "van dijke", "van dijke", ],
294 [ "fi", "fi", "fi", "fi", "fi", "Fi", "FI", "fi", "fi", "fi", ],
295 [ "filesystem", "filesystem", "filesystem", "fiLESYSTEM", "filesystem", "Filesystem", "FILESYSTEM", "filesystem", "filesystem", "filesystem", ],
296 [ "efficient", "efficient", "Efficient", "EffiCIENT", "efficient", "Efficient", "EFFICIENT", "efficient", "efficient", "efficient", ],
297 [ "flour and water", "flour and water", "flour And Water", "flOUR AND WATER", "flour and water", "Flour And Water", "FLOUR AND WATER", "flour and water", "flour and water", "flour and water", ],
298 [ "dz", "dz", "Dz", "DZ", "dz", "Dz", "DZ", "dz", "dz", "dz", ],
299 [ "dzur mountain", "dzur mountain", "Dzur Mountain", "DZUR MOUNTAIN", "dzur mountain", "Dzur Mountain", "DZUR MOUNTAIN", "dzur mountain", "dzur mountain", "dzur mountain", ],
300 [ "poſt", "poſt", "Poſt", "POST", "poſt", "Poſt", "POST", "post", "post", "post", ],
301 [ "poſt", "poſt", "Poſt", "POſt", "poſt", "Poſt", "POST", "poſt", "post", "post", ],
302 [ "ſtop", "ſtop", "ſtop", "ſtOP", "ſtop", "Stop", "STOP", "ſtop", "stop", "stop", ],
303 [ "tschüß", "tschüß", "Tschüß", "TSCHÜß", "tschüß", "Tschüß", "TSCHÜSS", "tschüß", "tschüss", "tschüss", ],
304 [ "TSCHÜẞ", "tschüß", "Tschüß", "TSCHÜẞ", "tschüß", "Tschüß", "TSCHÜẞ", "tschüß", "tschüss", "tschüss", ],
305 [ "rußland", "rußland", "Rußland", "RUßLAND", "rußland", "Rußland", "RUSSLAND", "rußland", "russland", "russland", ],
306 [ "RUẞLAND", "rußland", "Rußland", "RUẞLAND", "rußland", "Rußland", "RUẞLAND", "rußland", "russland", "russland", ],
307 [ "weiß", "weiß", "Weiß", "WEIß", "weiß", "Weiß", "WEISS", "weiß", "weiss", "weiss", ],
308 [ "WEIẞ", "weiß", "Weiß", "WEIẞ", "weiß", "Weiß", "WEIẞ", "weiß", "weıss", "weiss", ],
309 [ "ẞIEW", "ßiew", "ẞiew", "ẞIEW", "ßiew", "ẞiew", "ẞIEW", "ßiew", "ssıew", "ssiew", ],
310 [ "ͅ", "ͅ", "Ι", "Ι", "ͅ", "Ι", "Ι", "ι", "ι", "ι", ],
311 [ "ᾲ", "ᾲ", "Ὰͅ", "ᾺΙ", "ᾲ", "Ὰͅ", "ᾺΙ", "ὰι", "ὰι", "ὰι", ],
312 [ "Ὰι", "ὰι", "Ὰι", "ᾺΙ", "ὰι", "Ὰι", "ᾺΙ", "ὰι", "ὰι", "ὰι", ],
313 [ "ᾺΙ", "ὰι", "Ὰι", "ᾺΙ", "ὰι", "Ὰι", "ᾺΙ", "ὰι", "ὰι", "ὰι", ],
314 [ "ᾲ", "ᾲ", "ᾲ", "ᾲ", "ᾲ", "Ὰͅ", "ᾺΙ", "ᾲ", "ὰι", "ὰι", ],
315 [ "Ὰͅ", "ᾲ", "Ὰͅ", "ᾺΙ", "ᾲ", "Ὰͅ", "ᾺΙ", "ὰι", "ὰι", "ὰι", ],
316 [ "ᾺΙ", "ὰι", "Ὰι", "ᾺΙ", "ὰι", "Ὰι", "ᾺΙ", "ὰι", "ὰι", "ὰι", ],
317 [ "ᾲ στο διάολο", "ᾲ στο διάολο", "ᾲ Στο Διάολο", "ᾲ ΣΤΟ ΔΙΆΟΛΟ", "ᾲ στο διάολο", "Ὰͅ Στο Διάολο", "ᾺΙ ΣΤΟ ΔΙΆΟΛΟ", "ᾲ στο διάολο", "ὰι στο διάολο", "ὰι στο διάολο", ],
318 [ "ᾲ στο διάολο", "ᾲ στο διάολο", "Ὰͅ Στο Διάολο", "ᾺΙ ΣΤΟ ΔΙΆΟΛΟ", "ᾲ στο διάολο", "Ὰͅ Στο Διάολο", "ᾺΙ ΣΤΟ ΔΙΆΟΛΟ", "ὰι στο διάολο", "ὰι στο διάολο", "ὰι στο διάολο", ],
319 [ "ⅷ", "ⅷ", "Ⅷ", "Ⅷ", "ⅷ", "Ⅷ", "Ⅷ", "ⅷ", "ⅷ", "ⅷ", ],
320 [ "henry ⅷ", "henry ⅷ", "Henry Ⅷ", "HENRY Ⅷ", "henry ⅷ", "Henry Ⅷ", "HENRY Ⅷ", "henry ⅷ", "henry ⅷ", "henry ⅷ", ],
321 [ "ⓚ", "ⓚ", "Ⓚ", "Ⓚ", "ⓚ", "Ⓚ", "Ⓚ", "ⓚ", "ⓚ", "ⓚ", ],
322 [ "i work at ⓚ", "i work at ⓚ", "I Work At Ⓚ", "I WORK AT Ⓚ", "i work at ⓚ", "I Work At Ⓚ", "I WORK AT Ⓚ", "i work at ⓚ", "i work at ⓚ", "i work at ⓚ", ],
323 [ "istambul", "istambul", "Istambul", "ISTAMBUL", "istambul", "Istambul", "ISTAMBUL", "istambul", "istambul", "istambul", ],
324 [ "i̇stanbul", "i̇stanbul", "İstanbul", "İSTANBUL", "i̇stanbul", "İstanbul", "İSTANBUL", "i̇stanbul", "i̇stanbul", "i̇stanbul", ],
325 [ "İstanbul", "i̇stanbul", "İstanbul", "İSTANBUL", "i̇stanbul", "İstanbul", "İSTANBUL", "i̇stanbul", "ı̇stanbul", "i̇stanbul", ],
326 [ "İSTANBUL", "istanbul", "İstanbul", "İSTANBUL", "i̇stanbul", "İstanbul", "İSTANBUL", "İstanbul", "istanbul", "i̇stanbul", ],
327 [ "στιγμας", "στιγμας", "Στιγμας", "ΣΤΙΓΜΑΣ", "στιγμας", "Στιγμας", "ΣΤΙΓΜΑΣ", "στιγμασ", "στιγμασ", "στιγμασ", ],
328 [ "στιγμασ", "στιγμασ", "Στιγμασ", "ΣΤΙΓΜΑΣ", "στιγμασ", "Στιγμασ", "ΣΤΙΓΜΑΣ", "στιγμασ", "στιγμασ", "στιγμασ", ],
329 [ "ΣΤΙΓΜΑΣ", "στιγμασ", "Στιγμασ", "ΣΤΙΓΜΑΣ", "στιγμασ", "Στιγμασ", "ΣΤΙΓΜΑΣ", "στιγμασ", "στιγμασ", "στιγμασ", ],
330 [ "ʀᴀʀᴇ", "ʀᴀʀᴇ", "Ʀᴀʀᴇ", "ƦᴀƦᴇ", "ʀᴀʀᴇ", "Ʀᴀʀᴇ", "ƦᴀƦᴇ", "ʀᴀʀᴇ", "ʀᴀʀᴇ", "ʀᴀʀᴇ", ],
331 [ "𐐼𐐯𐑅𐐨𐑉𐐯𐐻", "𐐼𐐯𐑅𐐨𐑉𐐯𐐻", "𐐔𐐯𐑅𐐨𐑉𐐯𐐻", "𐐔𐐇𐐝𐐀𐐡𐐇𐐓", "𐐼𐐯𐑅𐐨𐑉𐐯𐐻", "𐐔𐐯𐑅𐐨𐑉𐐯𐐻", "𐐔𐐇𐐝𐐀𐐡𐐇𐐓", "𐐼𐐯𐑅𐐨𐑉𐐯𐐻", "𐐼𐐯𐑅𐐨𐑉𐐯𐐻", "𐐼𐐯𐑅𐐨𐑉𐐯𐐻", ],
332 [ "Ԧԧ", "ԧԧ", "Ԧԧ", "ԦԦ", "ԧԧ", "Ԧԧ", "ԦԦ", "ԧԧ", "ԧԧ", "ԧԧ", ],
333 [ "ﬓﬔﬕﬖﬗ", "ﬓﬔﬕﬖﬗ", "ﬓﬔﬕﬖﬗ", "ﬓﬔﬕﬖﬗ", "ﬓﬔﬕﬖﬗ", "Մնﬔﬕﬖﬗ", "ՄՆՄԵՄԻՎՆՄԽ", "ﬓﬔﬕﬖﬗ", "մնմեմիվնմխ", "մնմեմիվնմխ", ],
334 [ "ʼn groot", "ʼn groot", "ʼn Groot", "ʼn GROOT", "ʼn groot", "ʼN Groot", "ʼN GROOT", "ʼn groot", "ʼn groot", "ʼn groot", ],
335 [ "ẚ", "ẚ", "ẚ", "ẚ", "ẚ", "Aʾ", "Aʾ", "ẚ", "aʾ", "aʾ", ],
336 [ "ff", "ff", "ff", "ff", "ff", "Ff", "FF", "ff", "ff", "ff", ],
337 [ "ǰ", "ǰ", "ǰ", "ǰ", "ǰ", "J̌", "J̌", "ǰ", "ǰ", "ǰ", ],
338 [ "550 nm or Å", "550 nm or å", "550 Nm Or Å", "550 NM OR Å", "550 nm or å", "550 Nm Or Å", "550 NM OR Å", "550 nm or å", "550 nm or å", "550 nm or å", ],
344 my ($simple_lc, $simple_tc, $simple_uc, $simple_fc) = @{$_}[1, 2, 3, 7];
345 my ($orig, $lower, $titlecase, $upper, $fc_turkic, $fc_full) = @{$_}[0,4,5,6,8,9];
347 if ($orig =~ /(\P{Assigned})/) { # So can fail gracefully in earlier
349 fail(sprintf "because U+%04X is unassigned", ord($1));
352 is( fc($orig), $fc_full, "fc('$orig') returns '$fc_full'" );
353 is( "\F$orig", $fc_full, '\F works' );
354 is( lc($orig), $lower, "lc('$orig') returns '$lower'" );
355 is( "\L$orig", $lower, '\L works' );
356 is( uc($orig), $upper, "uc('$orig') returns '$upper'" );
357 is( "\U$orig", $upper, '\U works' );
363 package Eeyup { use overload q{""} => sub { main::uni_to_native("\x{df}") }, fallback => 1 }
364 package Uunope { use overload q{""} => sub { "\x{30cb}" }, fallback => 1 }
365 package Undef { use overload q{""} => sub { undef }, fallback => 1 }
367 my $obj = bless {}, "Eeyup";
368 is(fc($obj), "ss", "fc() works on overloaded objects returning latin-1");
369 $obj = bless {}, "Eeyup";
370 is("\F$obj", "ss", '\F works on overloaded objects returning latin-1');
372 $obj = bless {}, "Uunope";
373 is(fc($obj), "\x{30cb}", "fc() works on overloaded objects returning UTF-8");
374 $obj = bless {}, "Uunope";
375 is("\F$obj", "\x{30cb}", '\F works on overloaded objects returning UTF-8');
377 $obj = bless {}, "Undef";
381 use warnings "uninitialized";
382 local $SIG{__WARN__} = sub { $warnings++; like(shift, qr/Use of uninitialized value (?:\$obj )?in fc/) };
386 is( $warnings, 2, "correct number of warnings" );
389 package Derpy { sub TIESCALAR { bless {}, shift } sub FETCH { $fetched++; main::uni_to_native("\x{df}") } }
393 is( fc($x), "ss", "fc() works on tied values" );
394 is( $fetched, 1, "and only calls the magic once" );
399 use feature qw( fc );
400 my $troublesome1 = uni_to_native("\xdf") x 11; #SvLEN should be 12, SvCUR should be 11
401 #So this should force fc() to grow the string.
403 is( fc($troublesome1), "ss" x 11, "fc() grows the string" );
405 my $troublesome2 = "abcdef:" . uni_to_native("\x{df}")
406 . ":fjksjs"; #SvLEN should be 16, SvCUR should be 15
407 is( fc($troublesome2), "abcdef:ss:fjksjs", "fc() expands \\x{DF} in the middle of a string that needs to grow" );
409 my $troublesome3 = ":" . uni_to_native("\x{df}") . ":";
410 is( fc($troublesome3), ":ss:", "fc() expands \\x{DF} in the middle of a string" );
413 my $troublesome4 = uni_to_native("\x{B5}"); #\N{MICRON SIGN} is latin-1, but its foldcase is in UTF-8
415 is( fc($troublesome4), "\x{3BC}", "fc() for a latin-1 \x{B5} returns UTF-8" );
416 ok( !utf8::is_utf8($troublesome4), "fc() doesn't upgrade the original string" );
419 my $troublesome5 = uni_to_native("\x{C9}") . "abda"
420 . uni_to_native("\x{B5}") . "aaf"
421 . uni_to_native("\x{C8}"); # Up until foldcasing \x{B5}, the string
422 # was in Latin-1. This tests that the
423 # results don't have illegal UTF-8
424 # (i.e. leftover latin-1) in them
426 is( fc($troublesome5), uni_to_native("\x{E9}") . "abda\x{3BC}aaf"
427 . uni_to_native("\x{E8}") );
432 use feature qw( fc unicode_strings );
434 skip "locales not available", 256 unless locales_enabled('LC_ALL');
436 setlocale(&POSIX::LC_ALL, "C");
438 # This tests both code paths in pp_fc
443 utf8::downgrade($latin1); #No-op, but doesn't hurt
444 utf8::upgrade($utf8);
445 is(fc($latin1), fc($utf8), "fc() gives the same results for \\x{$_} in Latin-1 and UTF-8 under unicode_strings");
447 skip 'Locales not available', 2 unless locales_enabled('LC_CTYPE');
449 is(fc($latin1), lc($latin1), "use locale; fc(qq{\\x{$_}}), lc(qq{\\x{$_}}) when qq{\\x{$_}} is in latin-1");
450 is(fc($utf8), lc($utf8), "use locale; fc(qq{\\x{$_}}), lc(qq{\\x{$_}}) when qq{\\x{$_}} is in latin-1");
453 no feature 'unicode_strings';
454 is(fc($latin1), lc($latin1), "under nothing, fc() for <256 is the same as lc");
459 my $utf8_locale = find_utf8_ctype_locale();
462 use feature qw( fc );
464 no warnings 'locale'; # Would otherwise warn
465 is(fc("\x{1E9E}"), fc("\x{17F}\x{17F}"), 'fc("\x{1E9E}") eq fc("\x{17F}\x{17F}")');
466 use warnings 'locale';
468 skip 'Can\'t find a UTF-8 locale', 1 unless defined $utf8_locale;
469 setlocale(&LC_CTYPE, $utf8_locale);
470 is(fc("\x{1E9E}"), "ss", 'fc("\x{1E9E}") eq "ss" in a UTF-8 locale)');
475 skip 'Can\'t find a UTF-8 locale', 256 unless defined $utf8_locale;
477 use feature qw( fc unicode_strings );
479 # Get the official fc values outside locale.
483 push @unicode_fc, fc(chr);
486 # These should match the UTF-8 locale values
487 setlocale(&LC_CTYPE, $utf8_locale);
490 is(fc(chr), $unicode_fc[$_], "In a UTF-8 locale, fc(chr $_) is the same as official Unicode");
495 my $num_tests = curr_test() - 1;