3 # This file is intentionally encoded in latin-1.
5 # Test uc(), lc(), fc(), ucfirst(), lcfirst(), quotemeta() etc
11 require Config; import Config;
12 skip_all_without_unicode_tables();
13 require './charset_tools.pl';
14 require './loc_tools.pl'; # Contains find_utf8_ctype_locale()
19 plan tests => 139 + 4 * 256;
21 is(lc(undef), "", "lc(undef) is ''");
22 is(lcfirst(undef), "", "lcfirst(undef) is ''");
23 is(uc(undef), "", "uc(undef) is ''");
24 is(ucfirst(undef), "", "ucfirst(undef) is ''");
28 is(CORE::fc(undef), "", "fc(undef) is ''");
29 is(CORE::fc(''), "", "fc('') is ''");
33 like($@, qr/Undefined subroutine &main::fc/, "fc() throws an exception,");
39 ok(!$@, "...but works after requesting the feature");
46 is("\Q$a\E." , "HELLO\\.\\*\\ world.", '\Q\E HELLO.* world');
47 is("\u$a" , "HELLO\.\* world", '\u');
48 is("\l$a" , "hELLO\.\* world", '\l');
49 is("\U$a" , "HELLO\.\* WORLD", '\U');
50 is("\L$a" , "hello\.\* world", '\L');
51 is("\F$a" , "hello\.\* world", '\F');
53 is(quotemeta($a) , "HELLO\\.\\*\\ world", 'quotemeta');
54 is(ucfirst($a) , "HELLO\.\* world", 'ucfirst');
55 is(lcfirst($a) , "hELLO\.\* world", 'lcfirst');
56 is(uc($a) , "HELLO\.\* WORLD", 'uc');
57 is(lc($a) , "hello\.\* world", 'lc');
58 is(fc($a) , "hello\.\* world", 'fc');
60 is("\Q$b\E." , "hello\\.\\*\\ WORLD.", '\Q\E hello.* WORLD');
61 is("\u$b" , "Hello\.\* WORLD", '\u');
62 is("\l$b" , "hello\.\* WORLD", '\l');
63 is("\U$b" , "HELLO\.\* WORLD", '\U');
64 is("\L$b" , "hello\.\* world", '\L');
65 is("\F$b" , "hello\.\* world", '\F');
67 is(quotemeta($b) , "hello\\.\\*\\ WORLD", 'quotemeta');
68 is(ucfirst($b) , "Hello\.\* WORLD", 'ucfirst');
69 is(lcfirst($b) , "hello\.\* WORLD", 'lcfirst');
70 is(uc($b) , "HELLO\.\* WORLD", 'uc');
71 is(lc($b) , "hello\.\* world", 'lc');
72 is(fc($b) , "hello\.\* world", 'fc');
74 # \x{100} is LATIN CAPITAL LETTER A WITH MACRON; its bijective lowercase is
75 # \x{101}, LATIN SMALL LETTER A WITH MACRON.
76 # Which is also its foldcase.
78 $a = "\x{100}\x{101}Aa";
79 $b = "\x{101}\x{100}aA";
81 is("\Q$a\E." , "\x{100}\x{101}Aa.", '\Q\E \x{100}\x{101}Aa');
82 is("\u$a" , "\x{100}\x{101}Aa", '\u');
83 is("\l$a" , "\x{101}\x{101}Aa", '\l');
84 is("\U$a" , "\x{100}\x{100}AA", '\U');
85 is("\L$a" , "\x{101}\x{101}aa", '\L');
86 is("\F$a" , "\x{101}\x{101}aa", '\F');
88 is(quotemeta($a) , "\x{100}\x{101}Aa", 'quotemeta');
89 is(ucfirst($a) , "\x{100}\x{101}Aa", 'ucfirst');
90 is(lcfirst($a) , "\x{101}\x{101}Aa", 'lcfirst');
91 is(uc($a) , "\x{100}\x{100}AA", 'uc');
92 is(lc($a) , "\x{101}\x{101}aa", 'lc');
93 is(fc($a) , "\x{101}\x{101}aa", 'fc');
95 is("\Q$b\E." , "\x{101}\x{100}aA.", '\Q\E \x{101}\x{100}aA');
96 is("\u$b" , "\x{100}\x{100}aA", '\u');
97 is("\l$b" , "\x{101}\x{100}aA", '\l');
98 is("\U$b" , "\x{100}\x{100}AA", '\U');
99 is("\L$b" , "\x{101}\x{101}aa", '\L');
100 is("\F$b" , "\x{101}\x{101}aa", '\F');
102 is(quotemeta($b) , "\x{101}\x{100}aA", 'quotemeta');
103 is(ucfirst($b) , "\x{100}\x{100}aA", 'ucfirst');
104 is(lcfirst($b) , "\x{101}\x{100}aA", 'lcfirst');
105 is(uc($b) , "\x{100}\x{100}AA", 'uc');
106 is(lc($b) , "\x{101}\x{101}aa", 'lc');
107 is(fc($b) , "\x{101}\x{101}aa", 'fc');
109 my $sharp_s = uni_to_native("\x{DF}");
110 # \x{DF} is LATIN SMALL LETTER SHARP S, its uppercase is SS or \x{53}\x{53};
111 # \x{149} is LATIN SMALL LETTER N PRECEDED BY APOSTROPHE, its uppercase is
112 # \x{2BC}\x{E4} or MODIFIER LETTER APOSTROPHE and N.
114 is("\U${sharp_s}aB\x{149}cD", "SSAB\x{2BC}NCD", "multicharacter uppercase");
116 # The \x{DF} is its own lowercase, ditto for \x{149}.
117 # There are no single character -> multiple characters lowercase mappings.
119 is("\L${sharp_s}aB\x{149}cD", "${sharp_s}ab\x{149}cd",
120 "multicharacter lowercase");
122 # \x{DF} is LATIN SMALL LETTER SHARP S, its foldcase is ss or \x{73}\x{73};
123 # \x{149} is LATIN SMALL LETTER N PRECEDED BY APOSTROPHE, its foldcase is
124 # \x{2BC}\x{6E} or MODIFIER LETTER APOSTROPHE and n.
125 # Note that is this further tested in t/uni/fold.t
127 is("\F${sharp_s}aB\x{149}cD", "ssab\x{2BC}ncd", "multicharacter foldcase");
130 # titlecase is used for \u / ucfirst.
132 # \x{587} is ARMENIAN SMALL LIGATURE ECH YIWN and its titlecase is
133 # \x{535}\x{582} ARMENIAN CAPITAL LETTER ECH + ARMENIAN SMALL LETTER YIWN
134 # while its lowercase is
136 # and its uppercase is
137 # \x{535}\x{552} ARMENIAN CAPITAL LETTER ECH + ARMENIAN CAPITAL LETTER YIWN
138 # The foldcase is \x{565}\x{582} ARMENIAN SMALL LETTER ECH + ARMENIAN SMALL LETTER YIWN
142 is("\L\x{587}" , "\x{587}", "ligature lowercase");
143 is("\u\x{587}" , "\x{535}\x{582}", "ligature titlecase");
144 is("\U\x{587}" , "\x{535}\x{552}", "ligature uppercase");
145 is("\F\x{587}" , "\x{565}\x{582}", "ligature foldcase");
147 # mktables had problems where many-to-one case mappings didn't work right.
148 # The lib/uni/fold.t should give the fourth folding, "casefolding", a good
150 # \x{01C4} is LATIN CAPITAL LETTER DZ WITH CARON
151 # \x{01C5} is LATIN CAPITAL LETTER D WITH SMALL LETTER Z WITH CARON
152 # \x{01C6} is LATIN SMALL LETTER DZ WITH CARON
153 # \x{03A3} is GREEK CAPITAL LETTER SIGMA
154 # \x{03C2} is GREEK SMALL LETTER FINAL SIGMA
155 # \x{03C3} is GREEK SMALL LETTER SIGMA
157 is(lc("\x{1C4}") , "\x{1C6}", "U+01C4 lc is U+01C6");
158 is(lc("\x{1C5}") , "\x{1C6}", "U+01C5 lc is U+01C6, too");
160 is(ucfirst("\x{3C2}") , "\x{3A3}", "U+03C2 ucfirst is U+03A3");
161 is(ucfirst("\x{3C3}") , "\x{3A3}", "U+03C3 ucfirst is U+03A3, too");
163 is(uc("\x{1C5}") , "\x{1C4}", "U+01C5 uc is U+01C4");
164 is(uc("\x{1C6}") , "\x{1C4}", "U+01C6 uc is U+01C4, too");
166 # #18107: A host of bugs involving [ul]c{,first}. AMS 20021106
167 $a = "\x{3c3}foo.bar"; # \x{3c3} == GREEK SMALL LETTER SIGMA.
168 $b = "\x{3a3}FOO.BAR"; # \x{3a3} == GREEK CAPITAL LETTER SIGMA.
170 ($c = $b) =~ s/(\w+)/lc($1)/ge;
171 is($c , $a, "Using s///e to change case.");
173 ($c = $a) =~ s/(\p{IsWord}+)/uc($1)/ge;
174 is($c , $b, "Using s///e to change case.");
176 ($c = $a) =~ s/(\p{IsWord}+)/fc($1)/ge;
177 is($c , $a, "Using s///e to foldcase.");
179 ($c = $b) =~ s/(\p{IsWord}+)/lcfirst($1)/ge;
180 is($c , "\x{3c3}FOO.bAR", "Using s///e to change case.");
182 ($c = $a) =~ s/(\p{IsWord}+)/ucfirst($1)/ge;
183 is($c , "\x{3a3}foo.Bar", "Using s///e to change case.");
185 # #18931: perl5.8.0 bug in \U..\E processing
186 # Test case from Nicholas Clark.
192 is(uc($1), "ABCDEFGH", "[perl #18931]");
199 $a =~ s/^(\s*)(\w*)/$1\u$2/;
200 is($a, v10, "[perl #18857]");
205 # [perl #38619] Bug in lc and uc (interaction between UTF-8, substr, and lc/uc)
207 for ("a\x{100}", "xyz\x{100}") {
208 is(substr(uc($_), 0), uc($_), "[perl #38619] uc");
210 for ("A\x{100}", "XYZ\x{100}") {
211 is(substr(lc($_), 0), lc($_), "[perl #38619] lc");
213 for ("a\x{100}", "ßyz\x{100}") { # ß to Ss (different length)
214 is(substr(ucfirst($_), 0), ucfirst($_), "[perl #38619] ucfirst");
217 #fc() didn't exist back then, but coverage is coverage.
218 for ("a\x{100}", "ßyz\x{100}", "xyz\x{100}", "XYZ\x{100}") { # ß to Ss (different length)
219 is(substr(fc($_), 0), fc($_), "[perl #38619] fc");
222 # Related to [perl #38619]
223 # the original report concerns PERL_MAGIC_utf8.
224 # these cases concern PERL_MAGIC_regex_global.
226 for (map { $_ } "a\x{100}", "abc\x{100}", "\x{100}") {
227 chop; # get ("a", "abc", "") in utf8
228 my $return = uc($_) =~ /\G(.?)/g;
229 my $result = $return ? $1 : "not";
230 my $expect = (uc($_) =~ /(.?)/g)[0];
231 is($return, 1, "[perl #38619]");
232 is($result, $expect, "[perl #38619]");
235 for (map { $_ } "A\x{100}", "ABC\x{100}", "\x{100}") {
236 chop; # get ("A", "ABC", "") in utf8
237 my $return = lc($_) =~ /\G(.?)/g;
238 my $result = $return ? $1 : "not";
239 my $expect = (lc($_) =~ /(.?)/g)[0];
240 is($return, 1, "[perl #38619]");
241 is($result, $expect, "[perl #38619]");
244 for (map { $_ } "A\x{100}", "ABC\x{100}", "\x{100}") {
245 chop; # get ("A", "ABC", "") in utf8
246 my $return = fc($_) =~ /\G(.?)/g;
247 my $result = $return ? $1 : "not";
248 my $expect = (fc($_) =~ /(.?)/g)[0];
249 is($return, 1, "[perl #38619]");
250 is($result, $expect, "[perl #38619]");
253 for (1, 4, 9, 16, 25) {
254 is(uc "\x{03B0}" x $_, "\x{3a5}\x{308}\x{301}" x $_,
255 'uc U+03B0 grows threefold');
257 is(lc "\x{0130}" x $_, "i\x{307}" x $_, 'lc U+0130 grows');
259 is(fc "\x{03B0}" x $_, "\x{3C5}\x{308}\x{301}" x $_,
260 'fc U+03B0 grows threefold');
267 is($_, "HellO", '[perl #43207] lc($_) modifying $_');
271 is($_, "HellO", '[perl #43207] fc($_) modifying $_');
275 is($_, "HellO", '[perl #43207] uc($_) modifying $_');
279 is($_, "HellO", '[perl #43207] ucfirst($_) modifying $_');
283 is($_, "HellO", '[perl #43207] lcfirst($_) modifying $_');
286 # new in Unicode 5.1.0
287 is(lc("\x{1E9E}"), uni_to_native("\x{df}"), "lc(LATIN CAPITAL LETTER SHARP S)");
290 use feature 'unicode_strings';
292 is(lc(uni_to_native("\xc0")), uni_to_native("\xc0"), "lc of above-ASCII Latin1 is itself under use bytes");
293 is(lcfirst(uni_to_native("\xc0")), uni_to_native("\xc0"), "lcfirst of above-ASCII Latin1 is itself under use bytes");
294 is(uc(uni_to_native("\xe0")), uni_to_native("\xe0"), "uc of above-ASCII Latin1 is itself under use bytes");
295 is(ucfirst(uni_to_native("\xe0")), uni_to_native("\xe0"), "ucfirst of above-ASCII Latin1 is itself under use bytes");
298 # Brought up in ticket #117855: Constant folding applied to uc() should use
299 # the right set of hints.
300 fresh_perl_like(<<'constantfolding', qr/^(\d+),\1\z/, {},
304 use feature 'unicode_strings';
305 print ord uc($char), ",",
306 ord eval "$function('$char')", "\n";
309 'folded uc() in string eval uses the right hints');
311 # In-place lc/uc should not corrupt string buffers when given a non-utf8-
312 # flagged thingy that stringifies to utf8
313 $h{k} = bless[], "\x{3b0}\x{3b0}\x{3b0}bcde"; # U+03B0 grows with uc()
314 # using delete marks it as TEMP, so uc-in-place is permitted
315 like uc delete $h{k}, qr "^(?:\x{3a5}\x{308}\x{301}){3}BCDE=ARRAY\(.*\)",
316 'uc(TEMP ref) does not produce a corrupt string';
317 $h{k} = bless[], "\x{130}bcde"; # U+0130 grows with lc()
318 # using delete marks it as TEMP, so uc-in-place is permitted
319 like lc delete $h{k}, qr "^i\x{307}bcde=array\(.*\)",
320 'lc(TEMP ref) does not produce a corrupt string';
322 # List::Util::first() etc sets $_ to an SvTEMP without raising its
323 # refcount. This was causing lc() etc to unsafely modify in-place.
324 # see http://nntp.perl.org/group/perl.perl5.porters/228213
327 skip "no List::Util on miniperl", 5, if is_miniperl;
329 my %hl = qw(a 1 b 2 c 3);
330 my %hu = qw(A 1 B 2 C 3);
332 $x = List::Util::first(sub { uc $_ eq 'A' }, keys %hl);
333 is($x, "a", "first { uc }");
334 $x = List::Util::first(sub { ucfirst $_ eq 'A' }, keys %hl);
335 is($x, "a", "first { ucfirst }");
336 $x = List::Util::first(sub { lc $_ eq 'a' }, keys %hu);
337 is($x, "A", "first { lc }");
338 $x = List::Util::first(sub { lcfirst $_ eq 'a' }, keys %hu);
339 is($x, "A", "first { lcfirst }");
340 $x = List::Util::first(sub { fc $_ eq 'a' }, keys %hu);
341 is($x, "A", "first { fc }");
345 my $utf8_locale = find_utf8_ctype_locale();
348 skip 'Can\'t find a UTF-8 locale', 4*256 unless defined $utf8_locale;
350 use feature qw( unicode_strings );
359 # Get all the values outside of 'locale'
360 for my $i (0 .. 255) {
361 push @unicode_lc, lc(chr $i);
362 push @unicode_uc, uc(chr $i);
363 push @unicode_lcfirst, lcfirst(chr $i);
364 push @unicode_ucfirst, ucfirst(chr $i);
368 setlocale(LC_CTYPE, $utf8_locale);
370 for my $i (0 .. 255) {
371 is(lc(chr $i), $unicode_lc[$i], "In a UTF-8 locale, lc(chr $i) is the same as official Unicode");
372 is(uc(chr $i), $unicode_uc[$i], "In a UTF-8 locale, uc(chr $i) is the same as official Unicode");
373 is(lcfirst(chr $i), $unicode_lcfirst[$i], "In a UTF-8 locale, lcfirst(chr $i) is the same as official Unicode");
374 is(ucfirst(chr $i), $unicode_ucfirst[$i], "In a UTF-8 locale, ucfirst(chr $i) is the same as official Unicode");