1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
|
# Test the /a, /d, etc regex modifiers
BEGIN {
chdir 't' if -d 't';
@INC = '../lib';
require './test.pl';
require './loc_tools.pl';
}
use strict;
use warnings;
use Config;
plan('no_plan');
# Each case is a valid element of its hash key. Choose, where available, an
# ASCII-range, Latin-1 non-ASCII range, and above Latin1 range code point.
my %testcases = (
'\w' => [ ord("A"), 0xE2, 0x16B ], # Below expects these to all be alpha
'\d' => [ ord("0"), 0x0662 ],
'\s' => [ ord("\t"), 0xA0, 0x1680 ], # Below expects these to be [:blank:]
'[:cntrl:]' => [ 0x00, 0x88 ],
'[:graph:]' => [ ord("&"), 0xF7, 0x02C7 ], # Below expects these to be
# [:print:]
'[:lower:]' => [ ord("g"), 0xE3, 0x0127 ],
'[:punct:]' => [ ord("!"), 0xBF, 0x055C ],
'[:upper:]' => [ ord("G"), 0xC3, 0x0126 ],
'[:xdigit:]' => [ ord("4"), 0xFF15 ],
);
$testcases{'[:digit:]'} = $testcases{'\d'};
$testcases{'[:alnum:]'} = $testcases{'\w'};
$testcases{'[:alpha:]'} = $testcases{'\w'};
$testcases{'[:blank:]'} = $testcases{'\s'};
$testcases{'[:print:]'} = $testcases{'[:graph:]'};
$testcases{'[:space:]'} = $testcases{'\s'};
$testcases{'[:word:]'} = $testcases{'\w'};
my $utf8_locale;
my @charsets = qw(a d u aa);
if (! is_miniperl() && $Config{d_setlocale}) {
require POSIX;
my $current_locale = POSIX::setlocale( &POSIX::LC_ALL, "C") // "";
if ($current_locale eq 'C') {
# test for d_setlocale is repeated here because this one is compile
# time, and the one above is run time
use if $Config{d_setlocale}, 'locale';
# Some implementations don't have the 128-255 range characters all
# mean nothing under the C locale (an example being VMS). This is
# legal, but since we don't know what the right answers should be,
# skip the locale tests in that situation.
for my $i (128 .. 255) {
goto skip_adding_C_locale if chr($i) =~ /[[:print:]]/;
}
push @charsets, 'l';
skip_adding_C_locale:
# Use a pseudo-modifier 'L' to indicate to use /l with a UTF-8 locale
$utf8_locale = find_utf8_ctype_locale();
push @charsets, 'L' if defined $utf8_locale;
}
}
# For each possible character set...
foreach my $charset (@charsets) {
my $locale;
my $charset_mod = lc $charset;
my $charset_display;
if ($charset_mod eq 'l') {
$locale = POSIX::setlocale(&POSIX::LC_ALL, ($charset eq 'l')
? "C"
: $utf8_locale
);
die "Couldn't change locale" unless $locale;
$charset_display = $charset_mod . " ($locale)";
}
else {
$charset_display = $charset_mod;
}
# And in utf8 or not
foreach my $upgrade ("", 'utf8::upgrade($a); ') {
# reverse gets the, \w, \s, \d first.
for my $class (reverse sort keys %testcases) {
# The complement of \w is \W; of [:posix:] is [:^posix:]
my $complement = $class;
if ($complement !~ s/ ( \[: ) /$1^/x) {
$complement = uc($class);
}
# For each test case
foreach my $ord (@{$testcases{$class}}) {
my $char = display(chr($ord));
# > 255 already implies upgraded. Skip the ones that don't
# have an explicit upgrade. This shows more clearly in the
# output which tests are in utf8, or not.
next if $ord > 255 && ! $upgrade;
my $reason = ""; # Explanation output with each test
my $neg_reason = "";
my $match = 1; # Calculated whether test regex should
# match or not
# Everything always matches in ASCII, or under /u
if ($ord < 128 || $charset eq 'u' || $charset eq 'L') {
$reason = "\"$char\" is a $class under /$charset_display";
$neg_reason = "\"$char\" is not a $complement under /$charset_display";
}
elsif ($charset eq "a" || $charset eq "aa") {
$match = 0;
$reason = "\"$char\" is non-ASCII, which can't be a $class under /a";
$neg_reason = "\"$char\" is non-ASCII, which is a $complement under /a";
}
elsif ($ord > 255) {
$reason = "\"$char\" is a $class under /$charset_display";
$neg_reason = "\"$char\" is not a $complement under /$charset_display";
}
elsif ($charset eq 'l') {
# We are using the C locale, which is essentially ASCII,
# but under utf8, the above-latin1 chars are treated as
# Unicode)
$reason = "\"$char\" is not a $class in the C locale under /l";
$neg_reason = "\"$char\" is a $complement in the C locale under /l";
$match = 0;
}
elsif ($upgrade) {
$reason = "\"$char\" is a $class in utf8 under /d";
$neg_reason = "\"$char\" is not a $complement in utf8 under /d";
}
else {
$reason = "\"$char\" is above-ASCII latin1, which requires utf8 to be a $class under /d";
$neg_reason = "\"$char\" is above-ASCII latin1, which is a $complement under /d (unless in utf8)";
$match = 0;
}
$reason = "; $reason" if $reason;
$neg_reason = "; $neg_reason" if $neg_reason;
my $op;
my $neg_op;
if ($match) {
$op = '=~';
$neg_op = '!~';
}
else {
$op = '!~';
$neg_op = '=~';
}
# In [...] or not
foreach my $bracketed (0, 1) {
my $lb = "";
my $rb = "";
if ($bracketed) {
# Adds an extra char to the character class to make sure
# that the class doesn't get optimized away.
$lb = ($bracketed) ? '[_' : "";
$rb = ($bracketed) ? ']' : "";
}
else { # [:posix:] must be inside outer [ ]
next if $class =~ /\[/;
}
my $length = 10; # For regexec.c regrepeat() cases by
# matching more than one item
# Test both class and its complement, and with one or more
# than one item to match.
foreach my $eval (
qq[my \$a = "$char"; $upgrade\$a $op qr/ (?$charset_mod: $lb$class$rb ) /x],
qq[my \$a = "$char" x $length; $upgrade\$a $op qr/ (?$charset_mod: $lb$class$rb\{$length} ) /x],
) {
ok (eval $eval, $eval . $reason);
}
foreach my $eval (
qq[my \$a = "$char"; $upgrade\$a $neg_op qr/ (?$charset_mod: $lb$complement$rb ) /x],
qq[my \$a = "$char" x $length; $upgrade\$a $neg_op qr/ (?$charset_mod: $lb$complement$rb\{$length} ) /x],
) {
ok (eval $eval, $eval . $neg_reason);
}
}
next if $class ne '\w';
# Test \b, \B at beginning and end of string
foreach my $eval (
qq[my \$a = "$char"; $upgrade\$a $op qr/ (?$charset_mod: ^ \\b . ) /x],
qq[my \$a = "$char"; $upgrade\$a $op qr/ (?$charset_mod: . \\b \$) /x],
) {
ok (eval $eval, $eval . $reason);
}
foreach my $eval (
qq[my \$a = "$char"; $upgrade\$a $neg_op qr/(?$charset_mod: ^ \\B . ) /x],
qq[my \$a = "$char"; $upgrade\$a $neg_op qr/(?$charset_mod: . \\B \$ ) /x],
) {
ok (eval $eval, $eval . $neg_reason);
}
# Test \b, \B adjacent to a non-word char, both before it and
# after. We test with ASCII, Latin1 and Unicode non-word chars
foreach my $space_ord (@{$testcases{'\s'}}) {
# Useless to try to test non-utf8 when the ord itself
# forces utf8
next if $space_ord > 255 && ! $upgrade;
my $space = display(chr $space_ord);
foreach my $eval (
qq[my \$a = "$space$char"; $upgrade\$a $op qr/ (?$charset_mod: . \\b . ) /x],
qq[my \$a = "$char$space"; $upgrade\$a $op qr/ (?$charset_mod: . \\b . ) /x],
) {
ok (eval $eval, $eval . $reason . "; \"$space\" is not a \\w");
}
foreach my $eval (
qq[my \$a = "$space$char"; $upgrade\$a $neg_op qr/ (?$charset_mod: . \\B . ) /x],
qq[my \$a = "$char$space"; $upgrade\$a $neg_op qr/ (?$charset_mod: . \\B . ) /x],
) {
ok (eval $eval, $eval . $neg_reason . "; \"$space\" is not a \\w");
}
}
# Test \b, \B in the middle of two nominally word chars, but
# one or both may be considered non-word depending on range
# and charset.
foreach my $other_ord (@{$testcases{'\w'}}) {
next if $other_ord > 255 && ! $upgrade;
my $other = display(chr $other_ord);
# Determine if the other char is a word char in current
# circumstances
my $other_is_word = 1;
my $other_reason = "\"$other\" is a $class under /$charset_display";
my $other_neg_reason = "\"$other\" is not a $complement under /$charset_display";
if ($other_ord > 127
&& $charset ne 'u' && $charset ne 'L'
&& (($charset eq "a" || $charset eq "aa")
|| ($other_ord < 256 && ($charset eq 'l' || ! $upgrade))))
{
$other_is_word = 0;
$other_reason = "\"$other\" is not a $class under /$charset_display";
$other_neg_reason = "\"$other\" is a $complement under /$charset_display";
}
my $both_reason = $reason;
$both_reason .= "; $other_reason" if $other_ord != $ord;
my $both_neg_reason = $neg_reason;
$both_neg_reason .= "; $other_neg_reason" if $other_ord != $ord;
# If both are the same wordness, then \b will fail; \B
# succeed
if ($match == $other_is_word) {
$op = '!~';
$neg_op = '=~';
}
else {
$op = '=~';
$neg_op = '!~';
}
foreach my $eval (
qq[my \$a = "$other$char"; $upgrade\$a $op qr/ (?$charset_mod: $other \\b $char ) /x],
qq[my \$a = "$char$other"; $upgrade\$a $op qr/ (?$charset_mod: $char \\b $other ) /x],
) {
ok (eval $eval, $eval . $both_reason);
}
foreach my $eval (
qq[my \$a = "$other$char"; $upgrade\$a $neg_op qr/ (?$charset_mod: $other \\B $char ) /x],
qq[my \$a = "$char$other"; $upgrade\$a $neg_op qr/ (?$charset_mod: $char \\B $other ) /x],
) {
ok (eval $eval, $eval . $both_neg_reason);
}
next if $other_ord == $ord;
# These start with the \b or \B. They are included, based
# on source code analysis, to force the testing of the FBC
# (find_by_class) portions of regexec.c.
foreach my $eval (
qq[my \$a = "$other$char"; $upgrade\$a $op qr/ (?$charset_mod: \\b $char ) /x],
qq[my \$a = "$char$other"; $upgrade\$a $op qr/ (?$charset_mod: \\b $other ) /x],
) {
ok (eval $eval, $eval . $both_reason);
}
foreach my $eval (
qq[my \$a = "$other$char"; $upgrade\$a $neg_op qr/ (?$charset_mod: \\B $char ) /x],
qq[my \$a = "$char$other"; $upgrade\$a $neg_op qr/ (?$charset_mod: \\B $other ) /x],
) {
ok (eval $eval, $eval . $both_neg_reason);
}
}
} # End of each test case in a class
} # End of \w, \s, ...
} # End of utf8 upgraded or not
}
plan(curr_test() - 1);
|