xref: /openbsd/gnu/usr.bin/perl/t/op/utfhash.t (revision 9f11ffb7) 
1#!./perl -w
2
3BEGIN {
4    chdir 't' if -d 't';
5    require './test.pl';
6    set_up_inc('../lib');
7}
8
9plan(tests => 99);
10
11use strict;
12
13# Two hashes one with all 8-bit possible keys (initially), other
14# with a utf8 requiring key from the outset.
15
16my %hash8 = ( "\xff" => 0xff,
17              "\x7f" => 0x7f,
18            );
19my %hashu = ( "\xff" => 0xff,
20              "\x7f" => 0x7f,
21              "\x{1ff}" => 0x1ff,
22            );
23
24# Check that we can find the 8-bit things by various literals
25is($hash8{"\x{00ff}"},0xFF);
26is($hash8{"\x{007f}"},0x7F);
27is($hash8{"\xff"},0xFF);
28is($hash8{"\x7f"},0x7F);
29is($hashu{"\x{00ff}"},0xFF);
30is($hashu{"\x{007f}"},0x7F);
31is($hashu{"\xff"},0xFF);
32is($hashu{"\x7f"},0x7F);
33
34# Now try same thing with variables forced into various forms.
35foreach ("\x7f","\xff")
36 {
37  my $a = $_; # Force a copy
38  utf8::upgrade($a);
39  is($hash8{$a},ord($a));
40  is($hashu{$a},ord($a));
41  utf8::downgrade($a);
42  is($hash8{$a},ord($a));
43  is($hashu{$a},ord($a));
44  my $b = $a.chr(100);
45  chop($b);
46  is($hash8{$b},ord($b));
47  is($hashu{$b},ord($b));
48 }
49
50# Check we have not got an spurious extra keys
51is(join('',sort { ord $a <=> ord $b } keys %hash8),"\x7f\xff");
52is(join('',sort { ord $a <=> ord $b } keys %hashu),"\x7f\xff\x{1ff}");
53
54# Now add a utf8 key to the 8-bit hash
55$hash8{chr(0x1ff)} = 0x1ff;
56
57# Check we have not got an spurious extra keys
58is(join('',sort { ord $a <=> ord $b } keys %hash8),"\x7f\xff\x{1ff}");
59
60foreach ("\x7f","\xff","\x{1ff}")
61 {
62  my $a = $_;
63  utf8::upgrade($a);
64  is($hash8{$a},ord($a));
65  my $b = $a.chr(100);
66  chop($b);
67  is($hash8{$b},ord($b));
68 }
69
70# and remove utf8 from the other hash
71is(delete $hashu{chr(0x1ff)},0x1ff);
72is(join('',sort keys %hashu),"\x7f\xff");
73
74foreach ("\x7f","\xff")
75 {
76  my $a = $_;
77  utf8::upgrade($a);
78  is($hashu{$a},ord($a));
79  utf8::downgrade($a);
80  is($hashu{$a},ord($a));
81  my $b = $a.chr(100);
82  chop($b);
83  is($hashu{$b},ord($b));
84 }
85
86
87
88{
89  print "# Unicode hash keys and \\w\n";
90  # This is not really a regex test but regexes bring
91  # out the issue nicely.
92  use strict;
93  my $u3 = "f\x{df}\x{100}";
94  my $u2 = substr($u3,0,2);
95  my $u1 = substr($u2,0,1);
96  my $u0 = chr (0xdf)x4; # Make this 4 chars so that all lengths are distinct.
97
98  my @u = ($u0, $u1, $u2, $u3);
99
100  while (@u) {
101    my %u = (map {( $_, $_)} @u);
102    my $keys = scalar @u;
103    $keys .= ($keys == 1) ? " key" : " keys";
104
105    for (keys %u) {
106        my $l = 0 + /^\w+$/;
107        my $r = 0 + $u{$_} =~ /^\w+$/;
108	is ($l, $r, "\\w on keys with $keys, key of length " . length $_);
109    }
110
111    my $more;
112    do {
113      $more = 0;
114      # Want to do this direct, rather than copying to a temporary variable
115      # The first time each will return key and value at the start of the hash.
116      # each will return () after we've done the last pair. $more won't get
117      # set then, and the do will exit.
118      for (each %u) {
119        $more = 1;
120        my $l = 0 + /^\w+$/;
121        my $r = 0 + $u{$_} =~ /^\w+$/;
122        is ($l, $r, "\\w on each, with $keys, key of length " . length $_);
123      }
124    } while ($more);
125
126    for (%u) {
127      my $l = 0 + /^\w+$/;
128      my $r = 0 + $u{$_} =~ /^\w+$/;
129      is ($l, $r, "\\w on hash with $keys, key of length " . length $_);
130    }
131    pop @u;
132    undef %u;
133  }
134}
135
136{
137  my $utf8_sz = my $bytes_sz = "\x{df}";
138  $utf8_sz .= chr 256;
139  chop ($utf8_sz);
140
141  my (%bytes_first, %utf8_first);
142
143  $bytes_first{$bytes_sz} = $bytes_sz;
144
145  for (keys %bytes_first) {
146    my $l = 0 + /^\w+$/;
147    my $r = 0 + $bytes_first{$_} =~ /^\w+$/;
148    is ($l, $r, "\\w on each, bytes");
149  }
150
151  $bytes_first{$utf8_sz} = $utf8_sz;
152
153  for (keys %bytes_first) {
154    my $l = 0 + /^\w+$/;
155    my $r = 0 + $bytes_first{$_} =~ /^\w+$/;
156    is ($l, $r, "\\w on each, bytes now utf8");
157  }
158
159  $utf8_first{$utf8_sz} = $utf8_sz;
160
161  for (keys %utf8_first) {
162    my $l = 0 + /^\w+$/;
163    my $r = 0 + $utf8_first{$_} =~ /^\w+$/;
164    is ($l, $r, "\\w on each, utf8");
165  }
166
167  $utf8_first{$bytes_sz} = $bytes_sz;
168
169  for (keys %utf8_first) {
170    my $l = 0 + /^\w+$/;
171    my $r = 0 + $utf8_first{$_} =~ /^\w+$/;
172    is ($l, $r, "\\w on each, utf8 now bytes");
173  }
174
175}
176
177{
178    local $/; # Slurp.
179    my $data = <DATA>;
180    my ($utf8, $utf1047ebcdic) = split /__SPLIT__/, $data;
181    $utf8 = $utf1047ebcdic if $::IS_EBCDIC;
182    eval $utf8;
183}
184__END__
185{
186  # See if utf8 barewords work [perl #22969]
187  use utf8;
188  my %hash = (тест => 123);
189  is($hash{тест}, $hash{'тест'});
190  is($hash{тест}, 123);
191  is($hash{'тест'}, 123);
192  %hash = (тест => 123);
193  is($hash{тест}, $hash{'тест'});
194  is($hash{тест}, 123);
195  is($hash{'тест'}, 123);
196
197  # See if plain ASCII strings quoted with '=>' erroneously get utf8 flag [perl #68812]
198  my %foo = (a => 'b', 'c' => 'd');
199  for my $key (keys %foo) {
200    ok !utf8::is_utf8($key), "'$key' shouldn't have utf8 flag";
201  }
202}
203__SPLIT__
204{   # This is 1047 UTF-EBCDIC; won't work on other code pages.
205  # See if utf8 barewords work [perl #22969]
206  use utf8; # UTF-EBCDIC, really.
207  my %hash = (���������� => 123);
208  is($hash{����������}, $hash{'����������'});
209  is($hash{����������}, 123);
210  is($hash{'����������'}, 123);
211  %hash = (���������� => 123);
212  is($hash{����������}, $hash{'����������'});
213  is($hash{����������}, 123);
214  is($hash{'����������'}, 123);
215
216  # See if plain ASCII strings quoted with '=>' erroneously get utf8 flag [perl #68812]
217  my %foo = (a => 'b', 'c' => 'd');
218  for my $key (keys %foo) {
219    ok !utf8::is_utf8($key), "'$key' shouldn't have utf8 flag";
220  }
221}
222