xref: /onnv-gate/usr/src/cmd/perl/5.8.4/distrib/lib/utf8.t (revision 0:68f95e015346)
1#!./perl
2
3my $has_perlio;
4
5BEGIN {
6    chdir 't' if -d 't';
7    @INC = '../lib';
8    require './test.pl';
9    unless ($has_perlio = find PerlIO::Layer 'perlio') {
10	print <<EOF;
11# Since you don't have perlio you might get failures with UTF-8 locales.
12EOF
13    }
14}
15
16no utf8; # Ironic, no?
17
18# NOTE!
19#
20# Think carefully before adding tests here.  In general this should be
21# used only for about three categories of tests:
22#
23# (1) tests that absolutely require 'use utf8', and since that in general
24#     shouldn't be needed as the utf8 is being obsoleted, this should
25#     have rather few tests.  If you want to test Unicode and regexes,
26#     you probably want to go to op/regexp or op/pat; if you want to test
27#     split, go to op/split; pack, op/pack; appending or joining,
28#     op/append or op/join, and so forth
29#
30# (2) tests that have to do with Unicode tokenizing (though it's likely
31#     that all the other Unicode tests sprinkled around the t/**/*.t are
32#     going to catch that)
33#
34# (3) complicated tests that simultaneously stress so many Unicode features
35#     that deciding into which other test script the tests should go to
36#     is hard -- maybe consider breaking up the complicated test
37#
38#
39
40plan tests => 145;
41
42{
43    # bug id 20001009.001
44
45    my ($a, $b);
46
47    { use bytes; $a = "\xc3\xa4" }
48    { use utf8;  $b = "\xe4"     }
49
50    my $test = 68;
51
52    ok($a ne $b);
53
54    { use utf8; ok($a ne $b) }
55}
56
57
58{
59    # bug id 20000730.004
60
61    my $smiley = "\x{263a}";
62
63    for my $s ("\x{263a}",
64	       $smiley,
65
66	       "" . $smiley,
67	       "" . "\x{263a}",
68
69	       $smiley    . "",
70	       "\x{263a}" . "",
71	       ) {
72	my $length_chars = length($s);
73	my $length_bytes;
74	{ use bytes; $length_bytes = length($s) }
75	my @regex_chars = $s =~ m/(.)/g;
76	my $regex_chars = @regex_chars;
77	my @split_chars = split //, $s;
78	my $split_chars = @split_chars;
79	ok("$length_chars/$regex_chars/$split_chars/$length_bytes" eq
80	   "1/1/1/3");
81    }
82
83    for my $s ("\x{263a}" . "\x{263a}",
84	       $smiley    . $smiley,
85
86	       "\x{263a}\x{263a}",
87	       "$smiley$smiley",
88
89	       "\x{263a}" x 2,
90	       $smiley    x 2,
91	       ) {
92	my $length_chars = length($s);
93	my $length_bytes;
94	{ use bytes; $length_bytes = length($s) }
95	my @regex_chars = $s =~ m/(.)/g;
96	my $regex_chars = @regex_chars;
97	my @split_chars = split //, $s;
98	my $split_chars = @split_chars;
99	ok("$length_chars/$regex_chars/$split_chars/$length_bytes" eq
100	   "2/2/2/6");
101    }
102}
103
104
105{
106    my $w = 0;
107    local $SIG{__WARN__} = sub { print "#($_[0])\n"; $w++ };
108    my $x = eval q/"\\/ . "\x{100}" . q/"/;;
109
110    ok($w == 0 && $x eq "\x{100}");
111}
112
113{
114    use warnings;
115    use strict;
116
117    my $show = q(
118                 sub show {
119                   my $result;
120                   $result .= '>' . join (',', map {ord} split //, $_) . '<'
121                     foreach @_;
122                   $result;
123                 }
124                 1;
125                );
126    eval $show or die $@; # We don't expect this sub definition to fail.
127    my $progfile = 'utf' . $$;
128    END {unlink_all $progfile}
129
130    # If I'm right 60 is '>' in ASCII, ' ' in EBCDIC
131    # 173 is not punctuation in either ASCII or EBCDIC
132    my (@char);
133    foreach (60, 173, 257, 65532) {
134      my $char = chr $_;
135      utf8::encode($char);
136      # I don't want to use map {ord} and I've no need to hardcode the UTF
137      # version
138      my $charsubst = $char;
139      $charsubst =~ s/(.)/ord ($1) . ','/ge;
140      chop $charsubst;
141      # Not testing this one against map {ord}
142      my $char_as_ord
143          = join " . ", map {sprintf 'chr (%d)', ord $_} split //, $char;
144      push @char, [$_, $char, $charsubst, $char_as_ord];
145    }
146    # Now we've done all the UTF8 munching hopefully we're safe
147    my @tests = (
148             ['check our detection program works',
149              'my @a = ("'.chr(60).'\x2A", ""); $b = show @a', qr/^>60,42<><$/],
150             ['check literal 8 bit input',
151              '$a = "' . chr (173) . '"; $b = show $a', qr/^>173<$/],
152             ['check no utf8; makes no change',
153              'no utf8; $a = "' . chr (173) . '"; $b = show $a', qr/^>173<$/],
154             # Now we do the real byte sequences that are valid UTF8
155             (map {
156               ["the utf8 sequence for chr $_->[0]",
157                qq{\$a = "$_->[1]"; \$b = show \$a}, qr/^>$_->[2]<$/],
158               ["no utf8; for the utf8 sequence for chr $_->[0]",
159                qq(no utf8; \$a = "$_->[1]"; \$b = show \$a), qr/^>$_->[2]<$/],
160               ["use utf8; for the utf8 sequence for chr $_->[0]",
161                qq(use utf8; \$a = "$_->[1]"; \$b = show \$a), qr/^>$_->[0]<$/],
162              } @char),
163             # Interpolation of hex characters needs to take place now, as we're
164             # testing feeding malformed utf8 into perl. Bug now fixed was an
165             # "out of memory" error. We really need the "" [rather than qq()
166             # or q()] to get the best explosion.
167             ["!Feed malformed utf8 into perl.", <<"BANG",
168    use utf8; %a = ("\xE1\xA0"=>"sterling");
169    print 'start'; printf '%x,', ord \$_ foreach keys %a; print "end\n";
170BANG
171	      qr/^Malformed UTF-8 character \(\d bytes?, need \d, .+\).*start\d+,end$/sm
172	     ],
173            );
174    foreach (@tests) {
175        my ($why, $prog, $expect) = @$_;
176        open P, ">$progfile" or die "Can't open '$progfile': $!";
177        binmode(P, ":bytes") if $has_perlio;
178	print P $show, $prog, '; print $b'
179            or die "Print to 'progfile' failed: $!";
180        close P or die "Can't close '$progfile': $!";
181        if ($why =~ s/^!//) {
182            print "# Possible delay...\n";
183        } else {
184            print "# $prog\n";
185        }
186        my $result = runperl ( stderr => 1, progfile => $progfile );
187        like ($result, $expect, $why);
188    }
189    print
190        "# Again! Again! [but this time as eval, and not the explosive one]\n";
191    # and now we've safely done them all as separate files, check that the
192    # evals do the same thing. Hopefully doing it later successfully decouples
193    # the previous tests from anything messy that may go wrong with the evals.
194    foreach (@tests) {
195        my ($why, $prog, $expect) = @$_;
196        next if $why =~ m/^!/; # Goes bang.
197        my $result = eval $prog;
198        if ($@) {
199            print "# prog is $prog\n";
200            print "# \$\@=", _qq($@), "\n";
201        }
202        like ($result, $expect, $why);
203    }
204
205    # See what the tokeniser does with hash keys.
206    print "# What does the tokeniser do with utf8 hash keys?\n";
207    @tests = (map {
208        # This is the control - I don't expect it to fail
209        ["assign utf8 for chr $_->[0] to a hash",
210         qq(my \$a = "$_->[1]"; my %h; \$h{\$a} = 1;
211            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
212         qr/^>$_->[2]<$/],
213        ["no utf8; assign utf8 for chr $_->[0] to a hash",
214         qq(no utf8; my \$a = "$_->[1]"; my %h; \$h{\$a} = 1;
215            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
216         qr/^>$_->[2]<$/],
217        ["use utf8; assign utf8 for chr $_->[0] to a hash",
218         qq(use utf8; my \$a = "$_->[1]"; my %h; \$h{\$a} = 1;
219            my \$b = show keys %h; \$b .= 'F' unless \$h{chr $_->[0]}; \$b),
220         qr/^>$_->[0]<$/],
221        # Now check literal $h{"x"} constructions.
222        ["\$h{\"x\"} construction, where x is utf8 for chr $_->[0]",
223         qq(my \$a = "$_->[1]"; my %h; \$h{"$_->[1]"} = 1;
224            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
225         qr/^>$_->[2]<$/],
226        ["no utf8; \$h{\"x\"} construction, where x is utf8 for chr $_->[0]",
227         qq(no utf8; my \$a = "$_->[1]"; my %h; \$h{"$_->[1]"} = 1;
228            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
229         qr/^>$_->[2]<$/],
230        ["use utf8; \$h{\"x\"} construction, where x is utf8 for chr $_->[0]",
231         qq(use utf8; my \$a = "$_->[1]"; my %h; \$h{"$_->[1]"} = 1;
232            my \$b = show keys %h; \$b .= 'F' unless \$h{chr $_->[0]}; \$b),
233         qr/^>$_->[0]<$/],
234        # Now check "x" => constructions.
235        ["assign \"x\"=>1 to a hash, where x is utf8 for chr $_->[0]",
236         qq(my \$a = "$_->[1]"; my %h; %h = ("$_->[1]" => 1);
237            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
238         qr/^>$_->[2]<$/],
239        ["no utf8; assign \"x\"=>1 to a hash, where x is utf8 for chr $_->[0]",
240         qq(no utf8; my \$a = "$_->[1]"; my %h; %h = ("$_->[1]" => 1);
241            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
242         qr/^>$_->[2]<$/],
243        ["use utf8; assign \"x\"=>1 to a hash, where x is utf8 for chr $_->[0]",
244         qq(use utf8; my \$a = "$_->[1]"; my %h; %h = ("$_->[1]" => 1);
245            my \$b = show keys %h; \$b .= 'F' unless \$h{chr $_->[0]}; \$b),
246         qr/^>$_->[0]<$/],
247        # Check copies of hashes made from literal utf8 keys
248        ["assign utf8 for chr $_->[0] to a hash, then copy it",
249         qq(my \$a = "$_->[1]"; my %i; \$i{\$a} = 1; my %h = %i;
250            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
251         qr/^>$_->[2]<$/],
252        ["no utf8; assign utf8 for chr $_->[0] to a hash, then copy it",
253         qq(no utf8; my \$a = "$_->[1]"; my %i; \$i{\$a} = 1;; my %h = %i;
254            my \$b = show keys %h; \$b .= 'F' unless \$h{$_->[3]}; \$b),
255         qr/^>$_->[2]<$/],
256        ["use utf8; assign utf8 for chr $_->[0] to a hash, then copy it",
257         qq(use utf8; my \$a = "$_->[1]"; my %i; \$i{\$a} = 1; my %h = %i;
258            my \$b = show keys %h; \$b .= 'F' unless \$h{chr $_->[0]}; \$b),
259         qr/^>$_->[0]<$/],
260     } @char);
261    foreach (@tests) {
262        my ($why, $prog, $expect) = @$_;
263        # print "# $prog\n";
264        my $result = eval $prog;
265        like ($result, $expect, $why);
266    }
267}
268
269#
270# bug fixed by change #17928
271# separate perl used because we rely on 'strict' not yet loaded;
272# before the patch, the eval died with an error like:
273#   "my" variable $strict::VERSION can't be in a package
274#
275SKIP: {
276    skip("Embedded UTF-8 does not work in EBCDIC", 1) if ord("A") == 193;
277    ok('' eq runperl(prog => <<'CODE'), "change #17928");
278	my $code = qq{ my \$\xe3\x83\x95\xe3\x83\xbc = 5; };
279    {
280	use utf8;
281	eval $code;
282	print $@ if $@;
283    }
284CODE
285}
286
287{
288    use utf8;
289    $a = <<'END';
2900 ....... 1 ....... 2 ....... 3 ....... 4 ....... 5 ....... 6 ....... 7 .......
291END
292    my (@i, $s);
293
294    @i = ();
295    push @i, $s = index($a, '6');     # 60
296    push @i, $s = index($a, '.', $s); # next . after 60 is 62
297    push @i, $s = index($a, '5');     # 50
298    push @i, $s = index($a, '.', $s); # next . after 52 is 52
299    push @i, $s = index($a, '7');     # 70
300    push @i, $s = index($a, '.', $s); # next . after 70 is 72
301    push @i, $s = index($a, '4');     # 40
302    push @i, $s = index($a, '.', $s); # next . after 40 is 42
303    is("@i", "60 62 50 52 70 72 40 42", "utf8 heredoc index");
304
305    @i = ();
306    push @i, $s = rindex($a, '6');     # 60
307    push @i, $s = rindex($a, '.', $s); # previous . before 60 is 58
308    push @i, $s = rindex($a, '5');     # 50
309    push @i, $s = rindex($a, '.', $s); # previous . before 52 is 48
310    push @i, $s = rindex($a, '7');     # 70
311    push @i, $s = rindex($a, '.', $s); # previous . before 70 is 68
312    push @i, $s = rindex($a, '4');     # 40
313    push @i, $s = rindex($a, '.', $s); # previous . before 40 is 38
314    is("@i", "60 58 50 48 70 68 40 38", "utf8 heredoc rindex");
315
316    @i = ();
317    push @i, $s =  index($a, '6');     # 60
318    push @i,  index($a, '.', $s);      # next     . after  60 is 62
319    push @i, rindex($a, '.', $s);      # previous . before 60 is 58
320    push @i, $s = rindex($a, '5');     # 60
321    push @i,  index($a, '.', $s);      # next     . after  50 is 52
322    push @i, rindex($a, '.', $s);      # previous . before 50 is 48
323    push @i, $s =  index($a, '7', $s); # 70
324    push @i,  index($a, '.', $s);      # next     . after  70 is 72
325    push @i, rindex($a, '.', $s);      # previous . before 70 is 68
326    is("@i", "60 62 58 50 52 48 70 72 68", "utf8 heredoc index and rindex");
327}
328
329SKIP: {
330    skip("Embedded UTF-8 does not work in EBCDIC", 1) if ord("A") == 193;
331    use utf8;
332    eval qq{is(q \xc3\xbc test \xc3\xbc, qq\xc2\xb7 test \xc2\xb7,
333	       "utf8 quote delimiters [perl #16823]");};
334}
335
336# Test the "internals".
337
338{
339    my $a = "A";
340    my $b = chr(0x0FF);
341    my $c = chr(0x100);
342
343    ok( utf8::valid($a), "utf8::valid basic");
344    ok( utf8::valid($b), "utf8::valid beyond");
345    ok( utf8::valid($c), "utf8::valid unicode");
346
347    ok(!utf8::is_utf8($a), "!utf8::is_utf8 basic");
348    ok(!utf8::is_utf8($b), "!utf8::is_utf8 beyond");
349    ok( utf8::is_utf8($c), "utf8::is_utf8 unicode");
350
351    is(utf8::upgrade($a), 1, "utf8::upgrade basic");
352    is(utf8::upgrade($b), 2, "utf8::upgrade beyond");
353    is(utf8::upgrade($c), 2, "utf8::upgrade unicode");
354
355    is($a, "A",       "basic");
356    is($b, "\xFF",    "beyond");
357    is($c, "\x{100}", "unicode");
358
359    ok( utf8::valid($a), "utf8::valid basic");
360    ok( utf8::valid($b), "utf8::valid beyond");
361    ok( utf8::valid($c), "utf8::valid unicode");
362
363    ok( utf8::is_utf8($a), "utf8::is_utf8 basic");
364    ok( utf8::is_utf8($b), "utf8::is_utf8 beyond");
365    ok( utf8::is_utf8($c), "utf8::is_utf8 unicode");
366
367    is(utf8::downgrade($a), 1, "utf8::downgrade basic");
368    is(utf8::downgrade($b), 1, "utf8::downgrade beyond");
369
370    is($a, "A",       "basic");
371    is($b, "\xFF",    "beyond");
372
373    ok( utf8::valid($a), "utf8::valid basic");
374    ok( utf8::valid($b), "utf8::valid beyond");
375
376    ok(!utf8::is_utf8($a), "!utf8::is_utf8 basic");
377    ok(!utf8::is_utf8($b), "!utf8::is_utf8 beyond");
378
379    utf8::encode($a);
380    utf8::encode($b);
381    utf8::encode($c);
382
383    is($a, "A",       "basic");
384    is(length($b), 2, "beyond length");
385    is(length($c), 2, "unicode length");
386
387    ok(utf8::valid($a), "utf8::valid basic");
388    ok(utf8::valid($b), "utf8::valid beyond");
389    ok(utf8::valid($c), "utf8::valid unicode");
390
391    # encode() clears the UTF-8 flag (unlike upgrade()).
392    ok(!utf8::is_utf8($a), "!utf8::is_utf8 basic");
393    ok(!utf8::is_utf8($b), "!utf8::is_utf8 beyond");
394    ok(!utf8::is_utf8($c), "!utf8::is_utf8 unicode");
395
396    utf8::decode($a);
397    utf8::decode($b);
398    utf8::decode($c);
399
400    is($a, "A",       "basic");
401    is($b, "\xFF",    "beyond");
402    is($c, "\x{100}", "unicode");
403
404    ok(utf8::valid($a), "!utf8::valid basic");
405    ok(utf8::valid($b), "!utf8::valid beyond");
406    ok(utf8::valid($c), " utf8::valid unicode");
407
408    ok(!utf8::is_utf8($a), "!utf8::is_utf8 basic");
409    ok( utf8::is_utf8($b), " utf8::is_utf8 beyond"); # $b stays in UTF-8.
410    ok( utf8::is_utf8($c), " utf8::is_utf8 unicode");
411}
412
413{
414    eval {utf8::encode("�")};
415    like($@, qr/^Modification of a read-only value attempted/,
416	 "utf8::encode should refuse to touch read-only values");
417}
418
419{
420    my $a = "456\xb6";
421    utf8::upgrade($a);
422
423    my $b = "123456\xb6";
424    $b =~ s/^...//;
425    utf8::upgrade($b);
426    is($b, $a, "utf8::upgrade OffsetOK");
427}
428