4 require 'loc_tools.pl'; # Contains locales_enabled() and
5 # find_utf8_ctype_locale()
14 my $tab = " " x 4; # Indent subsidiary tests this much
16 use Unicode::UCD qw(search_invlist prop_invmap prop_invlist);
17 my ($charname_list, $charname_map, $format, $default) = prop_invmap("Name Alias");
22 # If there is a an abbreviation for the code point name, use it
23 my $name_index = search_invlist(\@{$charname_list}, $cp);
24 if (defined $name_index) {
25 my $synonyms = $charname_map->[$name_index];
27 my $pat = qr/: abbreviation/;
28 my @abbreviations = grep { $_ =~ $pat } @$synonyms;
30 return $abbreviations[0] =~ s/$pat//r;
35 # Otherwise, use the full name
37 return charnames::viacode($cp) // "No name";
40 sub truth($) { # Converts values so is() works
41 return (shift) ? 1 : 0;
46 if(locales_enabled('LC_ALL')) {
48 $base_locale = POSIX::setlocale( &POSIX::LC_ALL, "C");
49 if (defined $base_locale && $base_locale eq 'C') {
50 use locale; # make \w work right in non-ASCII lands
52 # Some locale implementations don't have the 128-255 characters all
53 # mean nothing. Skip the locale tests in that situation
54 for my $i (128 .. 255) {
55 if (chr(utf8::unicode_to_native($i)) =~ /[[:print:]]/) {
61 $utf8_locale = find_utf8_ctype_locale() if $base_locale;
65 sub get_display_locale_or_skip($$) {
67 # Helper function intimately tied to its callers. It knows the loop
68 # iterates with a locale of "", meaning don't use locale; $base_locale
69 # meaning to use a non-UTF-8 locale; and $utf8_locale.
71 # It checks to see if the current test should be skipped or executed,
72 # returning an empty list for the former, and for the latter:
73 # ( 'locale display name',
74 # bool of is this a UTF-8 locale )
76 # The display name is the empty string if not using locale. Functions
77 # with _LC in their name are skipped unless in locale, and functions
78 # without _LC are executed only outside locale. However, if no locales at
79 # all are on the system, the _LC functions are executed outside locale.
81 my ($locale, $suffix) = @_;
83 # The test should be skipped if the input is for a non-existent locale
84 return unless defined $locale;
86 # Here the input is defined, either a locale name or "". If the test is
87 # for not using locales, we want to do the test for non-LC functions,
88 # and skip it for LC ones (except if there are no locales on the system,
89 # we do it for LC ones as if they weren't LC).
91 return ("", 0) if $suffix !~ /LC/ || ! defined $base_locale;
95 # Here the input is for a real locale. We don't test the non-LC functions
97 return if $suffix !~ /LC/;
99 # Here is for a LC function and a real locale. The base locale is not
101 return (" ($locale locale)", 0) if $locale eq $base_locale;
103 # The only other possibility is that we have a UTF-8 locale
104 return (" ($locale)", 1);
107 sub try_malforming($$$)
109 # Determines if the tests for malformed UTF-8 should be done. When done,
110 # the .xs code creates malformations by pretending the length is shorter
111 # than it actually is. Some things can't be malformed, and sometimes this
112 # test knows that the current code doesn't look for a malformation under
113 # various circumstances.
115 my ($i, $function, $using_locale) = @_;
117 # Single bytes can't be malformed
118 return 0 if $i < ((ord "A" == 65) ? 128 : 160);
120 # ASCII doesn't need to ever look beyond the first byte.
121 return 0 if $function eq "ASCII";
123 # No controls above 255, so the code doesn't look at those
124 return 0 if $i > 255 && $function eq "CNTRL";
126 # No non-ASCII digits below 256, except if using locales.
127 return 0 if $i < 256 && ! $using_locale && $function =~ /X?DIGIT/;
133 # name => Lookup-property name
136 alphanumeric => 'Alnum',
137 alpha => 'XPosixAlpha',
143 idfirst => '_Perl_IDStart',
144 idcont => '_Perl_IDCont',
145 lower => 'XPosixLower',
147 psxspc => 'XPosixSpace',
148 punct => 'XPosixPunct',
149 quotemeta => '_Perl_Quotemeta',
150 space => 'XPerlSpace',
151 vertws => 'VertSpace',
152 upper => 'XPosixUpper',
157 local $SIG{__WARN__} = sub { push @warnings, @_ };
159 my %utf8_param_code = (
161 "_safe, malformed" => 1,
165 foreach my $name (sort keys %properties, 'octal') {
167 if ($name eq 'octal') {
168 # Hand-roll an inversion list with 0-7 in it and nothing else.
169 push @invlist, ord "0", ord "8";
172 my $property = $properties{$name};
173 @invlist = prop_invlist($property, '_perl_core_internal_ok');
176 # An empty return could mean an unknown property, or merely that
177 # it is empty. Call in scalar context to differentiate
178 if (! prop_invlist($property, '_perl_core_internal_ok')) {
179 fail("No inversion list found for $property");
185 # Include all the Latin1 code points, plus 0x100.
186 my @code_points = (0 .. 256);
188 # Then include the next few boundaries above those from this property
189 my $above_latins = 0;
190 foreach my $range_start (@invlist) {
191 next if $range_start < 257;
192 push @code_points, $range_start - 1, $range_start;
194 last if $above_latins > 5;
197 # This makes sure we are using the Perl definition of idfirst and idcont,
198 # and not the Unicode. There are a few differences.
199 push @code_points, ord "\N{ESTIMATED SYMBOL}" if $name =~ /^id(first|cont)/;
200 if ($name eq "idcont") { # And some that are continuation but not start
201 push @code_points, ord("\N{GREEK ANO TELEIA}"),
202 ord("\N{COMBINING GRAVE ACCENT}");
205 # And finally one non-Unicode code point.
206 push @code_points, 0x110000; # Above Unicode, no prop should match
207 no warnings 'non_unicode';
209 for my $j (@code_points) {
210 my $i = utf8::native_to_unicode($j);
211 my $function = uc($name);
213 is (@warnings, 0, "Got no unexpected warnings in previous iteration")
214 or diag("@warnings");
217 my $matches = search_invlist(\@invlist, $i);
218 if (! defined $matches) {
222 $matches = truth(! ($matches % 2));
226 my $char_name = get_charname($j);
227 my $display_name = sprintf "\\x{%02X, %s}", $i, $char_name;
228 my $display_call = "is${function}( $display_name )";
230 foreach my $suffix ("", "_A", "_L1", "_LC", "_uni", "_uvchr",
231 "_LC_uvchr", "_utf8", "_LC_utf8")
234 # Not all possible macros have been defined
235 if ($name eq 'vertws') {
237 # vertws is always all of Unicode
238 next if $suffix !~ / ^ _ ( uni | uvchr | utf8 ) $ /x;
240 elsif ($name eq 'alnum') {
242 # ALNUM_A, ALNUM_L1, and ALNUM_uvchr are not defined as these
243 # suffixes were added later, after WORDCHAR was created to be
244 # a clearer synonym for ALNUM
245 next if $suffix eq '_A'
247 || $suffix eq '_uvchr';
249 elsif ($name eq 'octal') {
250 next if $suffix ne "" && $suffix ne '_A' && $suffix ne '_L1';
252 elsif ($name eq 'quotemeta') {
253 # There is only one macro for this, and is defined only for
255 next if $suffix ne ""
258 foreach my $locale ("", $base_locale, $utf8_locale) {
260 my ($display_locale, $locale_is_utf8)
261 = get_display_locale_or_skip($locale, $suffix);
262 next unless defined $display_locale;
264 use if $locale, "locale";
265 POSIX::setlocale( &POSIX::LC_ALL, $locale) if $locale;
267 if ($suffix !~ /utf8/) { # _utf8 has to handled specially
269 = "is${function}$suffix( $display_name )$display_locale";
270 $ret = truth eval "test_is${function}$suffix($i)";
271 if (is ($@, "", "$display_call didn't give error")) {
272 my $truth = $matches;
275 # The single byte functions are false for
279 if $suffix=~ / ^ ( _A | _L [1C] )? $ /x;
281 elsif ( utf8::native_to_unicode($i) >= 128
282 && $name ne 'quotemeta')
285 # The no-suffix and _A functions are false
286 # for non-ASCII. So are _LC functions on a
288 $truth = 0 if $suffix eq "_A"
291 && ! $locale_is_utf8);
295 is ($ret, $truth, "${tab}And correctly returns $truth");
298 else { # _utf8 suffix
300 utf8::upgrade($char);
301 $char = quotemeta $char if $char eq '\\' || $char eq "'";
306 && utf8::native_to_unicode($i) >= 128)
307 { # The C-locale _LC function returns FALSE for Latin1
315 foreach my $utf8_param("_safe",
320 my $utf8_param_code = $utf8_param_code{$utf8_param};
321 my $expect_error = $utf8_param_code > 0;
322 next if $expect_error
323 && ! try_malforming($i, $function, $suffix =~ /LC/);
325 my $display_call = "is${function}$suffix( $display_name"
326 . ", $utf8_param )$display_locale";
327 $ret = truth eval "test_is${function}$suffix('$char',"
328 . " $utf8_param_code)";
331 "expected and got error in $display_call");
332 like($@, qr/Malformed UTF-8 character/,
333 "${tab}And got expected message");
334 if (is (@warnings, 1,
335 "${tab}Got a single warning besides"))
338 qr/Malformed UTF-8 character.*short/,
339 "${tab}Got expected warning");
346 elsif (is ($@, "", "$display_call didn't give error")) {
348 "${tab}And correctly returned $truth");
357 my %to_properties = (
358 FOLD => 'Case_Folding',
359 LOWER => 'Lowercase_Mapping',
360 TITLE => 'Titlecase_Mapping',
361 UPPER => 'Uppercase_Mapping',
365 foreach my $name (sort keys %to_properties) {
366 my $property = $to_properties{$name};
367 my ($list_ref, $map_ref, $format, $missing)
368 = prop_invmap($property, );
369 if (! $list_ref || ! $map_ref) {
370 fail("No inversion map found for $property");
373 if ($format !~ / ^ a l? $ /x) {
374 fail("Unexpected inversion map format ('$format') found for $property");
378 # Include all the Latin1 code points, plus 0x100.
379 my @code_points = (0 .. 256);
381 # Then include the next few multi-char folds above those from this
382 # property, and include the next few single folds as well
383 my $above_latins = 0;
385 for my $i (0 .. @{$list_ref} - 1) {
386 my $range_start = $list_ref->[$i];
387 next if $range_start < 257;
388 if (ref $map_ref->[$i] && $multi_char < 5) {
389 push @code_points, $range_start - 1
390 if $code_points[-1] != $range_start - 1;
391 push @code_points, $range_start;
394 elsif ($above_latins < 5) {
395 push @code_points, $range_start - 1
396 if $code_points[-1] != $range_start - 1;
397 push @code_points, $range_start;
400 last if $above_latins >= 5 && $multi_char >= 5;
403 # And finally one non-Unicode code point.
404 push @code_points, 0x110000; # Above Unicode, no prop should match
405 no warnings 'non_unicode';
407 # $j is native; $i unicode.
408 for my $j (@code_points) {
409 my $i = utf8::native_to_unicode($j);
410 my $function = $name;
412 my $index = search_invlist(\@{$list_ref}, $j);
415 my $char_name = get_charname($j);
416 my $display_name = sprintf "\\N{U+%02X, %s}", $i, $char_name;
418 foreach my $suffix ("", "_L1", "_LC") {
420 # This is the only macro defined for L1
421 next if $suffix eq "_L1" && $function ne "LOWER";
424 foreach my $locale ("", $base_locale, $utf8_locale) {
426 # titlecase is not defined in locales.
427 next if $name eq 'TITLE' && $suffix eq "_LC";
429 my ($display_locale, $locale_is_utf8)
430 = get_display_locale_or_skip($locale, $suffix);
431 next unless defined $display_locale;
433 skip("to${name}_LC does not work for LATIN SMALL LETTER SHARP S"
434 . "$display_locale", 1)
435 if $i == 0xDF && $name =~ / FOLD | UPPER /x
436 && $suffix eq "_LC" && $locale_is_utf8;
438 use if $locale, "locale";
439 POSIX::setlocale( &POSIX::LC_ALL, $locale) if $locale;
441 my $display_call = "to${function}$suffix("
442 . " $display_name )$display_locale";
443 $ret = eval "test_to${function}$suffix($j)";
444 if (is ($@, "", "$display_call didn't give error")) {
451 || ($suffix eq "_LC" && ! $locale_is_utf8)))
455 elsif ($map_ref->[$index] != $missing) {
456 $should_be = $map_ref->[$index] + $j - $list_ref->[$index]
462 is ($ret, $should_be,
463 sprintf("${tab}And correctly returned 0x%02X",
469 # The _uni, uvchr, and _utf8 functions return both the ordinal of the
470 # first code point of the result, and the result in utf8. The .xs
471 # tests return these in an array, in [0] and [1] respectively, with
472 # [2] the length of the utf8 in bytes.
473 my $utf8_should_be = "";
474 my $first_ord_should_be;
475 if (ref $map_ref->[$index]) { # A multi-char result
476 for my $j (0 .. @{$map_ref->[$index]} - 1) {
477 $utf8_should_be .= chr $map_ref->[$index][$j];
480 $first_ord_should_be = $map_ref->[$index][0];
482 else { # A single-char result
483 $first_ord_should_be = ($map_ref->[$index] != $missing)
484 ? $map_ref->[$index] + $j
485 - $list_ref->[$index]
487 $utf8_should_be = chr $first_ord_should_be;
489 utf8::upgrade($utf8_should_be);
492 foreach my $suffix ('_uni', '_uvchr') {
495 my $display_call = "to${function}$suffix( $display_name )";
496 $ret = eval "test_to${function}$suffix($j)";
497 if (is ($@, "", "$display_call didn't give error")) {
498 is ($ret->[0], $first_ord_should_be,
499 sprintf("${tab}And correctly returned 0x%02X",
500 $first_ord_should_be));
501 is ($ret->[1], $utf8_should_be, "${tab}Got correct utf8");
503 is ($ret->[2], length $utf8_should_be,
504 "${tab}Got correct number of bytes for utf8 length");
510 utf8::upgrade($char);
511 $char = quotemeta $char if $char eq '\\' || $char eq "'";
513 my $display_call = "to${function}_utf8($display_name )";
514 $ret = eval "test_to${function}_utf8('$char')";
515 if (is ($@, "", "$display_call didn't give error")) {
516 is ($ret->[0], $first_ord_should_be,
517 sprintf("${tab}And correctly returned 0x%02X",
518 $first_ord_should_be));
519 is ($ret->[1], $utf8_should_be, "${tab}Got correct utf8");
521 is ($ret->[2], length $utf8_should_be,
522 "${tab}Got correct number of bytes for utf8 length");
528 # This is primarily to make sure that no non-Unicode warnings get generated
529 is(scalar @warnings, 0, "No unexpected warnings were generated in the tests")