xref: /openbsd-src/gnu/usr.bin/perl/ext/XS-APItest/t/utf8.t (revision e068048151d29f2562a32185e21a8ba885482260)
1#!perl -w
2
3use strict;
4use Test::More;
5
6# This file tests various functions and macros in the API related to UTF-8.
7
8BEGIN {
9    use_ok('XS::APItest');
10    require 'charset_tools.pl';
11    require './t/utf8_setup.pl';
12};
13
14$|=1;
15
16use XS::APItest;
17use Config;
18my $word_length = defined $Config{quadkind} ? 8 : 4;
19
20# Below we test some byte-oriented functions that look for UTF-8 variant bytes
21# and we know can work on full words at a time.  Hence this is not black box
22# testing.  We know how long a word is.  Suppose it is 4.  We set things up so
23# that we have a string containing 3 bytes followed by 4, followed by 3, and
24# we tell our APItest functions to position the string so it starts at 1 byte
25# past a word boundary.  That way the first 3 bytes are the final ones of a
26# word, and the final 3 are the initial ones of a non-complete word.  This
27# assumes that the initial and final non-full word bytes are treated
28# individually, so we don't have to test the various combinations of partially
29# filled words.
30
31my $offset = 1;  # Start 1 byte past word boundary.
32
33# We choose an invariant and a variant that are at the boundaries between
34# those two types on ASCII platforms.  And, just in case the EBCDIC ever
35# changes to do per-word, we choose arbitrarily an invariant that has most of
36# its bits set natively, and a variant that has most unset.   First create
37# versions for display in the test names.
38my $display_invariant = isASCII ? "7F" : sprintf "%02X", utf8::unicode_to_native(0x9F);
39my $display_variant =   isASCII ? "80" : sprintf "%02X", utf8::unicode_to_native(0xA0);
40my $invariant = chr hex $display_invariant;
41my $variant = chr hex $display_variant;
42
43# We create a string with the correct number of bytes.  The -1 is to make the
44# final portion not quite fill a full word and $offset to do the same for the
45# initial portion.)
46my $string_length = 3 * $word_length - 1 - $offset;
47my $all_invariants = $invariant x $string_length;
48my $display_all_invariants = $display_invariant x $string_length;
49
50my $ret_ref = test_is_utf8_invariant_string_loc($all_invariants, $offset,
51                                                length $all_invariants);
52pass("The tests below are for is_utf8_invariant_string_loc() with string"
53   . " starting $offset bytes after a word boundary");
54is($ret_ref->[0], 1, "$display_all_invariants contains no variants");
55
56# Just create a string with a single variant, in all the possible positions.
57for my $pos (0.. length($all_invariants) - 1) {
58    my $test_string = $all_invariants;
59    my $test_display = $display_all_invariants;
60
61    substr($test_string, $pos, 1) = $variant;
62    substr($test_display, $pos * 2, 2) = $display_variant;
63    my $ret_ref = test_is_utf8_invariant_string_loc($test_string, $offset,
64                                                    length $test_string);
65    if (is($ret_ref->[0], 0, "$test_display has a variant")) {
66        is($ret_ref->[1], $pos, "   at position $pos");
67    }
68}
69
70# Now work on variant_under_utf8_count().
71pass("The tests below are for variant_under_utf8_count() with string"
72   . " starting $offset bytes after a word boundary");
73is(test_variant_under_utf8_count($all_invariants, $offset,
74                                length $all_invariants),
75                                0,
76                                "$display_all_invariants contains 0 variants");
77
78# First, put a variant in each possible position in the flanking partial words
79for my $pos (0 .. $word_length - $offset,
80             2 * $word_length .. length($all_invariants) - 1)
81{
82    my $test_string = $all_invariants;
83    my $test_display = $display_all_invariants;
84
85    substr($test_string, $pos, 1) = $variant;
86    substr($test_display, $pos * 2, 2) = $display_variant;
87    is(test_variant_under_utf8_count($test_string, $offset, length $test_string),
88                                     1,
89                                     "$test_display contains 1 variant");
90}
91
92# Then try all possible combinations of variant/invariant in the full word in
93# the middle  (We've already tested the case with 0 variants, so start at 1.)
94for my $bit_pattern (1 .. (1 << $word_length) - 1) {
95    my $bits = $bit_pattern;
96    my $display_word = "";
97    my $test_word = "";
98    my $count = 0;
99
100    # Every 1 bit gets the variant for this particular $bit_pattern.
101    for my $bit (0 .. 7) {
102        if ($bits & 1) {
103            $count++;
104            $test_word .= $variant;
105            $display_word .= $display_variant;
106        }
107        else {
108            $test_word .= $invariant;
109            $display_word .= $display_invariant;
110        }
111        $bits >>= 1;
112    }
113
114    my $test_string = $variant x ($word_length - 1)
115                    . $test_word
116                    . $variant x ($word_length - 1);
117    my $display_string = $display_variant x ($word_length - 1)
118                        . $display_word
119                        . $display_variant x ($word_length - 1);
120    my $expected_count = $count + 2 * $word_length - 2;
121    is(test_variant_under_utf8_count($test_string, $offset,
122                        length $test_string), $expected_count,
123                        "$display_string contains $expected_count variants");
124}
125
126
127my $pound_sign = chr utf8::unicode_to_native(163);
128
129# This  test file can't use byte_utf8a_to_utf8n() from t/charset_tools.pl
130# because that uses the same functions we are testing here.  So UTF-EBCDIC
131# strings are hard-coded as I8 strings in this file instead, and we use the
132# translation functions to/from I8 from that file instead.
133
134my $look_for_everything_utf8n_to
135                        = $::UTF8_DISALLOW_SURROGATE
136			| $::UTF8_WARN_SURROGATE
137			| $::UTF8_DISALLOW_NONCHAR
138			| $::UTF8_WARN_NONCHAR
139			| $::UTF8_DISALLOW_SUPER
140			| $::UTF8_WARN_SUPER
141			| $::UTF8_DISALLOW_PERL_EXTENDED
142			| $::UTF8_WARN_PERL_EXTENDED;
143my $look_for_everything_uvchr_to
144                        = $::UNICODE_DISALLOW_SURROGATE
145			| $::UNICODE_WARN_SURROGATE
146			| $::UNICODE_DISALLOW_NONCHAR
147			| $::UNICODE_WARN_NONCHAR
148			| $::UNICODE_DISALLOW_SUPER
149			| $::UNICODE_WARN_SUPER
150			| $::UNICODE_DISALLOW_PERL_EXTENDED
151			| $::UNICODE_WARN_PERL_EXTENDED;
152
153my $highest_non_extended_cp = 2 ** ((isASCII) ? 31 : 30) - 1;
154
155foreach ([0, '', '', 'empty'],
156	 [0, 'N', 'N', '1 char'],
157	 [1, 'NN', 'N', '1 char substring'],
158	 [-2, 'Perl', 'Rules', 'different'],
159	 [0, $pound_sign, $pound_sign, 'pound sign'],
160	 [1, $pound_sign . 10, $pound_sign . 1,
161                                            '10 pounds is more than 1 pound'],
162	 [1, $pound_sign . $pound_sign, $pound_sign,
163                                            '2 pound signs are more than 1'],
164	 [-2, ' $!', " \x{1F42B}!", 'Camels are worth more than 1 dollar'],
165	 [-1, '!', "!\x{1F42A}", 'Initial substrings match'],
166	) {
167    my ($expect, $left, $right, $desc) = @$_;
168    my $copy = $right;
169    utf8::encode($copy);
170    is(bytes_cmp_utf8($left, $copy), $expect, "bytes_cmp_utf8: $desc");
171    next if $right =~ tr/\0-\377//c;
172    utf8::encode($left);
173    is(bytes_cmp_utf8($right, $left), -$expect, "... and $desc reversed");
174}
175
176# The keys to this hash are Unicode code points, their values are the native
177# UTF-8 representations of them.  The code points are chosen because they are
178# "interesting" on either or both ASCII and EBCDIC platforms.  First we add
179# boundaries where the number of bytes required to represent them increase, or
180# are adjacent to problematic code points, so we want to make sure they aren't
181# considered problematic.
182my %code_points = (
183    0x0100     => (isASCII) ? "\xc4\x80"     : I8_to_native("\xc8\xa0"),
184    0x0400 - 1 => (isASCII) ? "\xcf\xbf"     : I8_to_native("\xdf\xbf"),
185    0x0400     => (isASCII) ? "\xd0\x80"     : I8_to_native("\xe1\xa0\xa0"),
186    0x0800 - 1 => (isASCII) ? "\xdf\xbf"     : I8_to_native("\xe1\xbf\xbf"),
187    0x0800     => (isASCII) ? "\xe0\xa0\x80" : I8_to_native("\xe2\xa0\xa0"),
188    0x4000 - 1 => (isASCII) ? "\xe3\xbf\xbf" : I8_to_native("\xef\xbf\xbf"),
189    0x4000     => (isASCII) ? "\xe4\x80\x80" : I8_to_native("\xf0\xb0\xa0\xa0"),
190    0x8000 - 1 => (isASCII) ? "\xe7\xbf\xbf" : I8_to_native("\xf0\xbf\xbf\xbf"),
191
192    # First code point that the implementation of isUTF8_POSSIBLY_PROBLEMATIC,
193    # as of this writing, considers potentially problematic on EBCDIC
194    0x8000     => (isASCII) ? "\xe8\x80\x80" : I8_to_native("\xf1\xa0\xa0\xa0"),
195
196    0xD000 - 1 => (isASCII) ? "\xec\xbf\xbf" : I8_to_native("\xf1\xb3\xbf\xbf"),
197
198    # First code point that the implementation of isUTF8_POSSIBLY_PROBLEMATIC,
199    # as of this writing, considers potentially problematic on ASCII
200    0xD000     => (isASCII) ? "\xed\x80\x80" : I8_to_native("\xf1\xb4\xa0\xa0"),
201
202    # Bracket the surrogates, and include several surrogates
203    0xD7FF     => (isASCII) ? "\xed\x9f\xbf" : I8_to_native("\xf1\xb5\xbf\xbf"),
204    0xD800     => (isASCII) ? "\xed\xa0\x80" : I8_to_native("\xf1\xb6\xa0\xa0"),
205    0xDC00     => (isASCII) ? "\xed\xb0\x80" : I8_to_native("\xf1\xb7\xa0\xa0"),
206    0xDFFF     => (isASCII) ? "\xed\xbf\xbf" : I8_to_native("\xf1\xb7\xbf\xbf"),
207    0xE000     => (isASCII) ? "\xee\x80\x80" : I8_to_native("\xf1\xb8\xa0\xa0"),
208
209    # Include the 32 contiguous non characters, and adjacent code points
210    0xFDCF     => (isASCII) ? "\xef\xb7\x8f" : I8_to_native("\xf1\xbf\xae\xaf"),
211    0xFDD0     => (isASCII) ? "\xef\xb7\x90" : I8_to_native("\xf1\xbf\xae\xb0"),
212    0xFDD1     => (isASCII) ? "\xef\xb7\x91" : I8_to_native("\xf1\xbf\xae\xb1"),
213    0xFDD2     => (isASCII) ? "\xef\xb7\x92" : I8_to_native("\xf1\xbf\xae\xb2"),
214    0xFDD3     => (isASCII) ? "\xef\xb7\x93" : I8_to_native("\xf1\xbf\xae\xb3"),
215    0xFDD4     => (isASCII) ? "\xef\xb7\x94" : I8_to_native("\xf1\xbf\xae\xb4"),
216    0xFDD5     => (isASCII) ? "\xef\xb7\x95" : I8_to_native("\xf1\xbf\xae\xb5"),
217    0xFDD6     => (isASCII) ? "\xef\xb7\x96" : I8_to_native("\xf1\xbf\xae\xb6"),
218    0xFDD7     => (isASCII) ? "\xef\xb7\x97" : I8_to_native("\xf1\xbf\xae\xb7"),
219    0xFDD8     => (isASCII) ? "\xef\xb7\x98" : I8_to_native("\xf1\xbf\xae\xb8"),
220    0xFDD9     => (isASCII) ? "\xef\xb7\x99" : I8_to_native("\xf1\xbf\xae\xb9"),
221    0xFDDA     => (isASCII) ? "\xef\xb7\x9a" : I8_to_native("\xf1\xbf\xae\xba"),
222    0xFDDB     => (isASCII) ? "\xef\xb7\x9b" : I8_to_native("\xf1\xbf\xae\xbb"),
223    0xFDDC     => (isASCII) ? "\xef\xb7\x9c" : I8_to_native("\xf1\xbf\xae\xbc"),
224    0xFDDD     => (isASCII) ? "\xef\xb7\x9d" : I8_to_native("\xf1\xbf\xae\xbd"),
225    0xFDDE     => (isASCII) ? "\xef\xb7\x9e" : I8_to_native("\xf1\xbf\xae\xbe"),
226    0xFDDF     => (isASCII) ? "\xef\xb7\x9f" : I8_to_native("\xf1\xbf\xae\xbf"),
227    0xFDE0     => (isASCII) ? "\xef\xb7\xa0" : I8_to_native("\xf1\xbf\xaf\xa0"),
228    0xFDE1     => (isASCII) ? "\xef\xb7\xa1" : I8_to_native("\xf1\xbf\xaf\xa1"),
229    0xFDE2     => (isASCII) ? "\xef\xb7\xa2" : I8_to_native("\xf1\xbf\xaf\xa2"),
230    0xFDE3     => (isASCII) ? "\xef\xb7\xa3" : I8_to_native("\xf1\xbf\xaf\xa3"),
231    0xFDE4     => (isASCII) ? "\xef\xb7\xa4" : I8_to_native("\xf1\xbf\xaf\xa4"),
232    0xFDE5     => (isASCII) ? "\xef\xb7\xa5" : I8_to_native("\xf1\xbf\xaf\xa5"),
233    0xFDE6     => (isASCII) ? "\xef\xb7\xa6" : I8_to_native("\xf1\xbf\xaf\xa6"),
234    0xFDE7     => (isASCII) ? "\xef\xb7\xa7" : I8_to_native("\xf1\xbf\xaf\xa7"),
235    0xFDE8     => (isASCII) ? "\xef\xb7\xa8" : I8_to_native("\xf1\xbf\xaf\xa8"),
236    0xFDEa     => (isASCII) ? "\xef\xb7\x99" : I8_to_native("\xf1\xbf\xaf\xa9"),
237    0xFDEA     => (isASCII) ? "\xef\xb7\xaa" : I8_to_native("\xf1\xbf\xaf\xaa"),
238    0xFDEB     => (isASCII) ? "\xef\xb7\xab" : I8_to_native("\xf1\xbf\xaf\xab"),
239    0xFDEC     => (isASCII) ? "\xef\xb7\xac" : I8_to_native("\xf1\xbf\xaf\xac"),
240    0xFDED     => (isASCII) ? "\xef\xb7\xad" : I8_to_native("\xf1\xbf\xaf\xad"),
241    0xFDEE     => (isASCII) ? "\xef\xb7\xae" : I8_to_native("\xf1\xbf\xaf\xae"),
242    0xFDEF     => (isASCII) ? "\xef\xb7\xaf" : I8_to_native("\xf1\xbf\xaf\xaf"),
243    0xFDF0     => (isASCII) ? "\xef\xb7\xb0" : I8_to_native("\xf1\xbf\xaf\xb0"),
244
245    # Mostly around non-characters, but some are transitions to longer strings
246    0xFFFD     => (isASCII) ? "\xef\xbf\xbd" : I8_to_native("\xf1\xbf\xbf\xbd"),
247    0x10000 - 1 => (isASCII)
248                   ?              "\xef\xbf\xbf"
249                   : I8_to_native("\xf1\xbf\xbf\xbf"),
250    0x10000     => (isASCII)
251                   ?              "\xf0\x90\x80\x80"
252                   : I8_to_native("\xf2\xa0\xa0\xa0"),
253    0x1FFFD     => (isASCII)
254                   ?              "\xf0\x9f\xbf\xbd"
255                   : I8_to_native("\xf3\xbf\xbf\xbd"),
256    0x1FFFE     => (isASCII)
257                   ?              "\xf0\x9f\xbf\xbe"
258                   : I8_to_native("\xf3\xbf\xbf\xbe"),
259    0x1FFFF     => (isASCII)
260                   ?              "\xf0\x9f\xbf\xbf"
261                   : I8_to_native("\xf3\xbf\xbf\xbf"),
262    0x20000     => (isASCII)
263                   ?              "\xf0\xa0\x80\x80"
264                   : I8_to_native("\xf4\xa0\xa0\xa0"),
265    0x2FFFD     => (isASCII)
266                   ?              "\xf0\xaf\xbf\xbd"
267                   : I8_to_native("\xf5\xbf\xbf\xbd"),
268    0x2FFFE     => (isASCII)
269                   ?              "\xf0\xaf\xbf\xbe"
270                   : I8_to_native("\xf5\xbf\xbf\xbe"),
271    0x2FFFF     => (isASCII)
272                   ?              "\xf0\xaf\xbf\xbf"
273                   : I8_to_native("\xf5\xbf\xbf\xbf"),
274    0x30000     => (isASCII)
275                   ?              "\xf0\xb0\x80\x80"
276                   : I8_to_native("\xf6\xa0\xa0\xa0"),
277    0x3FFFD     => (isASCII)
278                   ?              "\xf0\xbf\xbf\xbd"
279                   : I8_to_native("\xf7\xbf\xbf\xbd"),
280    0x3FFFE     => (isASCII)
281                   ?              "\xf0\xbf\xbf\xbe"
282                   : I8_to_native("\xf7\xbf\xbf\xbe"),
283    0x40000 - 1 => (isASCII)
284                   ?              "\xf0\xbf\xbf\xbf"
285                   : I8_to_native("\xf7\xbf\xbf\xbf"),
286    0x40000     => (isASCII)
287                   ?              "\xf1\x80\x80\x80"
288                   : I8_to_native("\xf8\xa8\xa0\xa0\xa0"),
289    0x4FFFD	=> (isASCII)
290                   ?              "\xf1\x8f\xbf\xbd"
291                   : I8_to_native("\xf8\xa9\xbf\xbf\xbd"),
292    0x4FFFE	=> (isASCII)
293                   ?              "\xf1\x8f\xbf\xbe"
294                   : I8_to_native("\xf8\xa9\xbf\xbf\xbe"),
295    0x4FFFF	=> (isASCII)
296                   ?              "\xf1\x8f\xbf\xbf"
297                   : I8_to_native("\xf8\xa9\xbf\xbf\xbf"),
298    0x50000     => (isASCII)
299                   ?              "\xf1\x90\x80\x80"
300                   : I8_to_native("\xf8\xaa\xa0\xa0\xa0"),
301    0x5FFFD	=> (isASCII)
302                   ?              "\xf1\x9f\xbf\xbd"
303                   : I8_to_native("\xf8\xab\xbf\xbf\xbd"),
304    0x5FFFE	=> (isASCII)
305                   ?              "\xf1\x9f\xbf\xbe"
306                   : I8_to_native("\xf8\xab\xbf\xbf\xbe"),
307    0x5FFFF	=> (isASCII)
308                   ?              "\xf1\x9f\xbf\xbf"
309                   : I8_to_native("\xf8\xab\xbf\xbf\xbf"),
310    0x60000     => (isASCII)
311                   ?              "\xf1\xa0\x80\x80"
312                   : I8_to_native("\xf8\xac\xa0\xa0\xa0"),
313    0x6FFFD	=> (isASCII)
314                   ?              "\xf1\xaf\xbf\xbd"
315                   : I8_to_native("\xf8\xad\xbf\xbf\xbd"),
316    0x6FFFE	=> (isASCII)
317                   ?              "\xf1\xaf\xbf\xbe"
318                   : I8_to_native("\xf8\xad\xbf\xbf\xbe"),
319    0x6FFFF	=> (isASCII)
320                   ?              "\xf1\xaf\xbf\xbf"
321                   : I8_to_native("\xf8\xad\xbf\xbf\xbf"),
322    0x70000     => (isASCII)
323                   ?              "\xf1\xb0\x80\x80"
324                   : I8_to_native("\xf8\xae\xa0\xa0\xa0"),
325    0x7FFFD	=> (isASCII)
326                   ?              "\xf1\xbf\xbf\xbd"
327                   : I8_to_native("\xf8\xaf\xbf\xbf\xbd"),
328    0x7FFFE	=> (isASCII)
329                   ?              "\xf1\xbf\xbf\xbe"
330                   : I8_to_native("\xf8\xaf\xbf\xbf\xbe"),
331    0x7FFFF	=> (isASCII)
332                   ?              "\xf1\xbf\xbf\xbf"
333                   : I8_to_native("\xf8\xaf\xbf\xbf\xbf"),
334    0x80000     => (isASCII)
335                   ?              "\xf2\x80\x80\x80"
336                   : I8_to_native("\xf8\xb0\xa0\xa0\xa0"),
337    0x8FFFD	=> (isASCII)
338                   ?              "\xf2\x8f\xbf\xbd"
339                   : I8_to_native("\xf8\xb1\xbf\xbf\xbd"),
340    0x8FFFE	=> (isASCII)
341                   ?              "\xf2\x8f\xbf\xbe"
342                   : I8_to_native("\xf8\xb1\xbf\xbf\xbe"),
343    0x8FFFF	=> (isASCII)
344                   ?              "\xf2\x8f\xbf\xbf"
345                   : I8_to_native("\xf8\xb1\xbf\xbf\xbf"),
346    0x90000     => (isASCII)
347                   ?              "\xf2\x90\x80\x80"
348                   : I8_to_native("\xf8\xb2\xa0\xa0\xa0"),
349    0x9FFFD	=> (isASCII)
350                   ?              "\xf2\x9f\xbf\xbd"
351                   : I8_to_native("\xf8\xb3\xbf\xbf\xbd"),
352    0x9FFFE	=> (isASCII)
353                   ?              "\xf2\x9f\xbf\xbe"
354                   : I8_to_native("\xf8\xb3\xbf\xbf\xbe"),
355    0x9FFFF	=> (isASCII)
356                   ?              "\xf2\x9f\xbf\xbf"
357                   : I8_to_native("\xf8\xb3\xbf\xbf\xbf"),
358    0xA0000     => (isASCII)
359                   ?              "\xf2\xa0\x80\x80"
360                   : I8_to_native("\xf8\xb4\xa0\xa0\xa0"),
361    0xAFFFD	=> (isASCII)
362                   ?              "\xf2\xaf\xbf\xbd"
363                   : I8_to_native("\xf8\xb5\xbf\xbf\xbd"),
364    0xAFFFE	=> (isASCII)
365                   ?              "\xf2\xaf\xbf\xbe"
366                   : I8_to_native("\xf8\xb5\xbf\xbf\xbe"),
367    0xAFFFF	=> (isASCII)
368                   ?              "\xf2\xaf\xbf\xbf"
369                   : I8_to_native("\xf8\xb5\xbf\xbf\xbf"),
370    0xB0000     => (isASCII)
371                   ?              "\xf2\xb0\x80\x80"
372                   : I8_to_native("\xf8\xb6\xa0\xa0\xa0"),
373    0xBFFFD	=> (isASCII)
374                   ?              "\xf2\xbf\xbf\xbd"
375                   : I8_to_native("\xf8\xb7\xbf\xbf\xbd"),
376    0xBFFFE	=> (isASCII)
377                   ?              "\xf2\xbf\xbf\xbe"
378                   : I8_to_native("\xf8\xb7\xbf\xbf\xbe"),
379    0xBFFFF	=> (isASCII)
380                   ?               "\xf2\xbf\xbf\xbf"
381                   : I8_to_native("\xf8\xb7\xbf\xbf\xbf"),
382    0xC0000     => (isASCII)
383                   ?               "\xf3\x80\x80\x80"
384                   : I8_to_native("\xf8\xb8\xa0\xa0\xa0"),
385    0xCFFFD	=> (isASCII)
386                   ?               "\xf3\x8f\xbf\xbd"
387                   : I8_to_native("\xf8\xb9\xbf\xbf\xbd"),
388    0xCFFFE	=> (isASCII)
389                   ?               "\xf3\x8f\xbf\xbe"
390                   : I8_to_native("\xf8\xb9\xbf\xbf\xbe"),
391    0xCFFFF	=> (isASCII)
392                   ?               "\xf3\x8f\xbf\xbf"
393                   : I8_to_native("\xf8\xb9\xbf\xbf\xbf"),
394    0xD0000     => (isASCII)
395                   ?               "\xf3\x90\x80\x80"
396                   : I8_to_native("\xf8\xba\xa0\xa0\xa0"),
397    0xDFFFD	=> (isASCII)
398                   ?               "\xf3\x9f\xbf\xbd"
399                   : I8_to_native("\xf8\xbb\xbf\xbf\xbd"),
400    0xDFFFE	=> (isASCII)
401                   ?               "\xf3\x9f\xbf\xbe"
402                   : I8_to_native("\xf8\xbb\xbf\xbf\xbe"),
403    0xDFFFF	=> (isASCII)
404                   ?               "\xf3\x9f\xbf\xbf"
405                   : I8_to_native("\xf8\xbb\xbf\xbf\xbf"),
406    0xE0000     => (isASCII)
407                   ?               "\xf3\xa0\x80\x80"
408                   : I8_to_native("\xf8\xbc\xa0\xa0\xa0"),
409    0xEFFFD	=> (isASCII)
410                   ?               "\xf3\xaf\xbf\xbd"
411                   : I8_to_native("\xf8\xbd\xbf\xbf\xbd"),
412    0xEFFFE	=> (isASCII)
413                   ?               "\xf3\xaf\xbf\xbe"
414                   : I8_to_native("\xf8\xbd\xbf\xbf\xbe"),
415    0xEFFFF	=> (isASCII)
416                   ?               "\xf3\xaf\xbf\xbf"
417                   : I8_to_native("\xf8\xbd\xbf\xbf\xbf"),
418    0xF0000     => (isASCII)
419                   ?               "\xf3\xb0\x80\x80"
420                   : I8_to_native("\xf8\xbe\xa0\xa0\xa0"),
421    0xFFFFD	=> (isASCII)
422                   ?               "\xf3\xbf\xbf\xbd"
423                   : I8_to_native("\xf8\xbf\xbf\xbf\xbd"),
424    0xFFFFE	=> (isASCII)
425                   ?               "\xf3\xbf\xbf\xbe"
426                   : I8_to_native("\xf8\xbf\xbf\xbf\xbe"),
427    0xFFFFF	=> (isASCII)
428                   ?               "\xf3\xbf\xbf\xbf"
429                   : I8_to_native("\xf8\xbf\xbf\xbf\xbf"),
430    0x100000    => (isASCII)
431                   ?               "\xf4\x80\x80\x80"
432                   : I8_to_native("\xf9\xa0\xa0\xa0\xa0"),
433    0x10FFFD	=> (isASCII)
434                   ?               "\xf4\x8f\xbf\xbd"
435                   : I8_to_native("\xf9\xa1\xbf\xbf\xbd"),
436    0x10FFFE	=> (isASCII)
437                   ?               "\xf4\x8f\xbf\xbe"
438                   : I8_to_native("\xf9\xa1\xbf\xbf\xbe"),
439    0x10FFFF	=> (isASCII)
440                   ?               "\xf4\x8f\xbf\xbf"
441                   : I8_to_native("\xf9\xa1\xbf\xbf\xbf"),
442    0x110000    => (isASCII)
443                   ?               "\xf4\x90\x80\x80"
444                   : I8_to_native("\xf9\xa2\xa0\xa0\xa0"),
445
446    # Things that would be noncharacters if they were in Unicode, and might be
447    # mistaken, if the C code is bad, to be nonchars
448    0x11FFFE    => (isASCII)
449                   ?               "\xf4\x9f\xbf\xbe"
450                    : I8_to_native("\xf9\xa3\xbf\xbf\xbe"),
451    0x11FFFF    => (isASCII)
452                   ?               "\xf4\x9f\xbf\xbf"
453                    : I8_to_native("\xf9\xa3\xbf\xbf\xbf"),
454    0x20FFFE    => (isASCII)
455                   ?               "\xf8\x88\x8f\xbf\xbe"
456                    : I8_to_native("\xfa\xa1\xbf\xbf\xbe"),
457    0x20FFFF    => (isASCII)
458                   ?               "\xf8\x88\x8f\xbf\xbf"
459                    : I8_to_native("\xfa\xa1\xbf\xbf\xbf"),
460
461    0x200000 - 1 => (isASCII)
462                    ?              "\xf7\xbf\xbf\xbf"
463                    : I8_to_native("\xf9\xbf\xbf\xbf\xbf"),
464    0x200000     => (isASCII)
465                    ?              "\xf8\x88\x80\x80\x80"
466                    : I8_to_native("\xfa\xa0\xa0\xa0\xa0"),
467    0x400000 - 1 => (isASCII)
468                    ?              "\xf8\x8f\xbf\xbf\xbf"
469                    : I8_to_native("\xfb\xbf\xbf\xbf\xbf"),
470    0x400000     => (isASCII)
471                    ?              "\xf8\x90\x80\x80\x80"
472                    : I8_to_native("\xfc\xa4\xa0\xa0\xa0\xa0"),
473    0x4000000 - 1 => (isASCII)
474                     ?              "\xfb\xbf\xbf\xbf\xbf"
475                     : I8_to_native("\xfd\xbf\xbf\xbf\xbf\xbf"),
476    0x4000000     => (isASCII)
477                     ?              "\xfc\x84\x80\x80\x80\x80"
478                     : I8_to_native("\xfe\xa2\xa0\xa0\xa0\xa0\xa0"),
479    0x4000000 - 1 => (isASCII)
480                     ?              "\xfb\xbf\xbf\xbf\xbf"
481                     : I8_to_native("\xfd\xbf\xbf\xbf\xbf\xbf"),
482    0x4000000     => (isASCII)
483                     ?              "\xfc\x84\x80\x80\x80\x80"
484                     : I8_to_native("\xfe\xa2\xa0\xa0\xa0\xa0\xa0"),
485    0x40000000 - 1 => (isASCII)
486                      ?              "\xfc\xbf\xbf\xbf\xbf\xbf"
487                      : I8_to_native("\xfe\xbf\xbf\xbf\xbf\xbf\xbf"),
488    0x40000000     =>
489    (isASCII) ?    "\xfd\x80\x80\x80\x80\x80"
490    : I8_to_native("\xff\xa0\xa0\xa0\xa0\xa0\xa0\xa1\xa0\xa0\xa0\xa0\xa0\xa0"),
491    0x80000000 - 1 =>
492    (isASCII) ?    "\xfd\xbf\xbf\xbf\xbf\xbf"
493    : I8_to_native("\xff\xa0\xa0\xa0\xa0\xa0\xa0\xa1\xbf\xbf\xbf\xbf\xbf\xbf"),
494);
495
496if ($::is64bit) {
497    no warnings qw(overflow portable);
498    $code_points{0x80000000}
499    = (isASCII)
500    ?              "\xfe\x82\x80\x80\x80\x80\x80"
501    : I8_to_native("\xff\xa0\xa0\xa0\xa0\xa0\xa0\xa2\xa0\xa0\xa0\xa0\xa0\xa0");
502    $code_points{0xFFFFFFFF}
503    = (isASCII)
504    ?              "\xfe\x83\xbf\xbf\xbf\xbf\xbf"
505    : I8_to_native("\xff\xa0\xa0\xa0\xa0\xa0\xa0\xa3\xbf\xbf\xbf\xbf\xbf\xbf");
506    $code_points{0x100000000}
507     = (isASCII)
508     ?              "\xfe\x84\x80\x80\x80\x80\x80"
509     : I8_to_native("\xff\xa0\xa0\xa0\xa0\xa0\xa0\xa4\xa0\xa0\xa0\xa0\xa0\xa0");
510    $code_points{0x1000000000 - 1}
511     = (isASCII)
512     ?              "\xfe\xbf\xbf\xbf\xbf\xbf\xbf"
513     : I8_to_native("\xff\xa0\xa0\xa0\xa0\xa0\xa1\xbf\xbf\xbf\xbf\xbf\xbf\xbf");
514    $code_points{0x1000000000}
515     = (isASCII)
516     ?              "\xff\x80\x80\x80\x80\x80\x81\x80\x80\x80\x80\x80\x80"
517     : I8_to_native("\xff\xa0\xa0\xa0\xa0\xa0\xa2\xa0\xa0\xa0\xa0\xa0\xa0\xa0");
518    $code_points{0x7FFFFFFFFFFFFFFF}
519     = (isASCII)
520     ?              "\xff\x80\x87\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf"
521     : I8_to_native("\xff\xa7\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf");
522
523    # This is used when UV_MAX is the upper limit of acceptable code points
524    # $code_points{0xFFFFFFFFFFFFFFFF}
525    # = (isASCII)
526    # ?              "\xff\x80\x8f\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf"
527    # : I8_to_native("\xff\xaf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf\xbf");
528
529    if (isASCII) {  # These could falsely show as overlongs in a naive
530                    # implementation
531        $code_points{0x40000000000}
532                      = "\xff\x80\x80\x80\x80\x81\x80\x80\x80\x80\x80\x80\x80";
533        $code_points{0x1000000000000}
534                      = "\xff\x80\x80\x80\x81\x80\x80\x80\x80\x80\x80\x80\x80";
535        $code_points{0x40000000000000}
536                      = "\xff\x80\x80\x81\x80\x80\x80\x80\x80\x80\x80\x80\x80";
537        $code_points{0x1000000000000000}
538                      = "\xff\x80\x81\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80";
539        # overflows
540        #$code_points{0xfoo}
541        #           = "\xff\x81\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80";
542    }
543}
544elsif (! isASCII) { # 32-bit EBCDIC.  64-bit is clearer to handle, so doesn't
545                    # need this test case
546    no warnings qw(overflow portable);
547    $code_points{0x40000000} = I8_to_native(
548                    "\xff\xa0\xa0\xa0\xa0\xa0\xa0\xa1\xa0\xa0\xa0\xa0\xa0\xa0");
549}
550
551# Now add in entries for each of code points 0-255, which require special
552# handling on EBCDIC.  Remember the keys are Unicode values, and the values
553# are the native UTF-8.  For invariants, the bytes are just the native chr.
554
555my $cp = 0;
556while ($cp < ((isASCII) ? 128 : 160)) {   # This is from the definition of
557                                        # invariant
558    $code_points{$cp} = chr utf8::unicode_to_native($cp);
559    $cp++;
560}
561
562# Done with the invariants.  Now do the variants.  All in this range are 2
563# byte.  Again, we can't use the internal functions to generate UTF-8, as
564# those are what we are trying to test.  In the loop, we know what range the
565# continuation bytes can be in, and what the lowest start byte can be.  So we
566# cycle through them.
567
568my $highest_continuation = 0xBF;
569my $start = (isASCII) ? 0xC2 : 0xC5;
570
571my $continuation = $::lowest_continuation - 1;
572
573while ($cp < 255) {
574    if (++$continuation > $highest_continuation) {
575
576        # Wrap to the next start byte when we reach the final continuation
577        # byte possible
578        $continuation = $::lowest_continuation;
579        $start++;
580    }
581    $code_points{$cp} = I8_to_native(chr($start) . chr($continuation));
582
583    $cp++;
584}
585
586my @warnings;
587
588use warnings 'utf8';
589local $SIG{__WARN__} = sub { push @warnings, @_ };
590
591my %restriction_types;
592
593# This set of tests looks for basic sanity, and lastly tests various routines
594# for the given code point.  If the earlier tests for that code point fail,
595# the later ones probably will too.  Malformations are tested in later
596# segments of code.
597for my $u (sort { utf8::unicode_to_native($a) <=> utf8::unicode_to_native($b) }
598          keys %code_points)
599{
600    my $hex_u = sprintf("0x%02X", $u);
601    my $n = utf8::unicode_to_native($u);
602    my $hex_n = sprintf("0x%02X", $n);
603    my $bytes = $code_points{$u};
604
605    my $offskip_should_be;
606    {
607        no warnings qw(overflow portable);
608        $offskip_should_be = (isASCII)
609            ? ( $u < 0x80           ? 1 :
610                $u < 0x800          ? 2 :
611                $u < 0x10000        ? 3 :
612                $u < 0x200000       ? 4 :
613                $u < 0x4000000      ? 5 :
614                $u < 0x80000000     ? 6 : (($::is64bit)
615                                        ? ($u < 0x1000000000 ? 7 : $::max_bytes)
616                                        : 7)
617              )
618            : ($u < 0xA0        ? 1 :
619               $u < 0x400       ? 2 :
620               $u < 0x4000      ? 3 :
621               $u < 0x40000     ? 4 :
622               $u < 0x400000    ? 5 :
623               $u < 0x4000000   ? 6 :
624               $u < 0x40000000  ? 7 : $::max_bytes );
625    }
626
627    # If this test fails, subsequent ones are meaningless.
628    next unless is(test_OFFUNISKIP($u), $offskip_should_be,
629                   "Verify OFFUNISKIP($hex_u) is $offskip_should_be");
630    my $invariant = $offskip_should_be == 1;
631    my $display_invariant = $invariant || 0;
632    is(test_OFFUNI_IS_INVARIANT($u), $invariant,
633       "Verify OFFUNI_IS_INVARIANT($hex_u) is $display_invariant");
634
635    my $uvchr_skip_should_be = $offskip_should_be;
636    next unless is(test_UVCHR_SKIP($n), $uvchr_skip_should_be,
637                   "Verify UVCHR_SKIP($hex_n) is $uvchr_skip_should_be");
638    is(test_UVCHR_IS_INVARIANT($n), $offskip_should_be == 1,
639       "Verify UVCHR_IS_INVARIANT($hex_n) is $display_invariant");
640
641    my $n_chr = chr $n;
642    utf8::upgrade $n_chr;
643
644    is(test_UTF8_SKIP($n_chr), $uvchr_skip_should_be,
645        "Verify UTF8_SKIP(chr $hex_n) is $uvchr_skip_should_be");
646
647    use bytes;
648    my $byte_length = length $n_chr;
649    for (my $j = 0; $j < $byte_length; $j++) {
650        undef @warnings;
651
652        if ($j == $byte_length - 1) {
653            my $ret
654              = test_is_utf8_valid_partial_char_flags($n_chr, $byte_length, 0);
655            is($ret, 0, "   Verify is_utf8_valid_partial_char_flags("
656                      . display_bytes($n_chr)
657                      . ") returns 0 for full character");
658        }
659        else {
660            my $bytes_so_far = substr($n_chr, 0, $j + 1);
661            my $ret
662             = test_is_utf8_valid_partial_char_flags($bytes_so_far, $j + 1, 0);
663            is($ret, 1, "   Verify is_utf8_valid_partial_char_flags("
664                      . display_bytes($bytes_so_far)
665                      . ") returns 1");
666        }
667
668        is(scalar @warnings, 0, "   Verify is_utf8_valid_partial_char_flags"
669                              . " generated no warnings")
670          or output_warnings(@warnings);
671
672        my $b = substr($n_chr, $j, 1);
673        my $hex_b = sprintf("\"\\x%02x\"", ord $b);
674
675        my $byte_invariant = $j == 0 && $uvchr_skip_should_be == 1;
676        my $display_byte_invariant = $byte_invariant || 0;
677        next unless is(test_UTF8_IS_INVARIANT($b), $byte_invariant,
678                       "   Verify UTF8_IS_INVARIANT($hex_b) for byte $j "
679                     . "is $display_byte_invariant");
680
681        my $is_start = $j == 0 && $uvchr_skip_should_be > 1;
682        my $display_is_start = $is_start || 0;
683        next unless is(test_UTF8_IS_START($b), $is_start,
684                    "      Verify UTF8_IS_START($hex_b) is $display_is_start");
685
686        my $is_continuation = $j != 0 && $uvchr_skip_should_be > 1;
687        my $display_is_continuation = $is_continuation || 0;
688        next unless is(test_UTF8_IS_CONTINUATION($b), $is_continuation,
689                       "      Verify UTF8_IS_CONTINUATION($hex_b) is "
690                     . "$display_is_continuation");
691
692        my $is_continued = $uvchr_skip_should_be > 1;
693        my $display_is_continued = $is_continued || 0;
694        next unless is(test_UTF8_IS_CONTINUED($b), $is_continued,
695                       "      Verify UTF8_IS_CONTINUED($hex_b) is "
696                     . "$display_is_continued");
697
698        my $is_downgradeable_start =    $n < 256
699                                     && $uvchr_skip_should_be > 1
700                                     && $j == 0;
701        my $display_is_downgradeable_start = $is_downgradeable_start || 0;
702        next unless is(test_UTF8_IS_DOWNGRADEABLE_START($b),
703                       $is_downgradeable_start,
704                       "      Verify UTF8_IS_DOWNGRADEABLE_START($hex_b) is "
705                     . "$display_is_downgradeable_start");
706
707        my $is_above_latin1 =  $n > 255 && $j == 0;
708        my $display_is_above_latin1 = $is_above_latin1 || 0;
709        next unless is(test_UTF8_IS_ABOVE_LATIN1($b),
710                       $is_above_latin1,
711                       "      Verify UTF8_IS_ABOVE_LATIN1($hex_b) is "
712                     . "$display_is_above_latin1");
713
714        my $is_possibly_problematic =  $j == 0
715                                    && $n >= ((isASCII)
716                                              ? 0xD000
717                                              : 0x8000);
718        my $display_is_possibly_problematic = $is_possibly_problematic || 0;
719        next unless is(test_isUTF8_POSSIBLY_PROBLEMATIC($b),
720                       $is_possibly_problematic,
721                       "      Verify isUTF8_POSSIBLY_PROBLEMATIC($hex_b) is "
722                     . "$display_is_above_latin1");
723    }
724
725    # We are not trying to look for warnings, etc, so if they should occur, it
726    # is an error.  But some of the code points here do cause warnings, so we
727    # check here and turn off the ones that apply to such code points.  A
728    # later section of the code tests for these kinds of things.
729    my $this_utf8_flags = $look_for_everything_utf8n_to;
730    my $len = length $bytes;
731
732    my $valid_under_strict = 1;
733    my $valid_under_c9strict = 1;
734    my $valid_for_not_extended_utf8 = 1;
735    if ($n > 0x10FFFF) {
736        $this_utf8_flags &= ~($::UTF8_DISALLOW_SUPER|$::UTF8_WARN_SUPER);
737        $valid_under_strict = 0;
738        $valid_under_c9strict = 0;
739        if ($n > $highest_non_extended_cp) {
740            $this_utf8_flags &=
741                ~($::UTF8_DISALLOW_PERL_EXTENDED|$::UTF8_WARN_PERL_EXTENDED);
742            $valid_for_not_extended_utf8 = 0;
743        }
744    }
745    elsif (($n >= 0xFDD0 && $n <= 0xFDEF) || ($n & 0xFFFE) == 0xFFFE) {
746        $this_utf8_flags &= ~($::UTF8_DISALLOW_NONCHAR|$::UTF8_WARN_NONCHAR);
747        $valid_under_strict = 0;
748    }
749    elsif ($n >= 0xD800 && $n <= 0xDFFF) {
750        $this_utf8_flags &= ~($::UTF8_DISALLOW_SURROGATE|$::UTF8_WARN_SURROGATE);
751        $valid_under_c9strict = 0;
752        $valid_under_strict = 0;
753    }
754
755    undef @warnings;
756
757    my $display_flags = sprintf "0x%x", $this_utf8_flags;
758    my $display_bytes = display_bytes($bytes);
759    my $ret_ref = test_utf8n_to_uvchr_error($bytes, $len, $this_utf8_flags);
760
761    # Rest of tests likely meaningless if it gets the wrong code point.
762    next unless is($ret_ref->[0], $n,
763                   "Verify utf8n_to_uvchr_error($display_bytes, $display_flags)"
764                 . "returns $hex_n");
765    is($ret_ref->[1], $len,
766       "Verify utf8n_to_uvchr_error() for $hex_n returns expected length:"
767     . " $len");
768
769    unless (is(scalar @warnings, 0,
770             "Verify utf8n_to_uvchr_error() for $hex_n generated no warnings"))
771    {
772        output_warnings(@warnings);
773    }
774    is($ret_ref->[2], 0,
775       "Verify utf8n_to_uvchr_error() returned no error bits");
776
777    undef @warnings;
778
779    my $ret = test_isUTF8_CHAR($bytes, $len);
780    is($ret, $len,
781            "Verify isUTF8_CHAR($display_bytes) returns expected length: $len");
782
783    unless (is(scalar @warnings, 0,
784               "Verify isUTF8_CHAR() for $hex_n generated no warnings"))
785    {
786        output_warnings(@warnings);
787    }
788
789    undef @warnings;
790
791    $ret = test_isUTF8_CHAR($bytes, $len - 1);
792    is($ret, 0,
793            "Verify isUTF8_CHAR() with too short length parameter returns 0");
794
795    is(scalar @warnings, 0, "Verify isUTF8_CHAR() generated no warnings")
796      or output_warnings(@warnings);
797
798    undef @warnings;
799
800    $ret = test_isUTF8_CHAR_flags($bytes, $len, 0);
801    is($ret, $len, "Verify isUTF8_CHAR_flags($display_bytes, 0)"
802                 . " returns expected length: $len");
803
804    is(scalar @warnings, 0,
805               "Verify isUTF8_CHAR_flags() for $hex_n generated no warnings")
806      or output_warnings(@warnings);
807
808    undef @warnings;
809
810    $ret = test_isUTF8_CHAR_flags($bytes, $len - 1, 0);
811    is($ret, 0,
812        "Verify isUTF8_CHAR_flags() with too short length parameter returns 0");
813
814    is(scalar @warnings, 0, "Verify isUTF8_CHAR_flags() generated no warnings")
815      or output_warnings(@warnings);
816
817    undef @warnings;
818
819    $ret = test_isSTRICT_UTF8_CHAR($bytes, $len);
820    my $expected_len = ($valid_under_strict) ? $len : 0;
821    is($ret, $expected_len, "Verify isSTRICT_UTF8_CHAR($display_bytes)"
822                          . " returns expected length: $expected_len");
823
824    is(scalar @warnings, 0,
825               "Verify isSTRICT_UTF8_CHAR() for $hex_n generated no warnings")
826      or output_warnings(@warnings);
827
828    undef @warnings;
829
830    $ret = test_isSTRICT_UTF8_CHAR($bytes, $len - 1);
831    is($ret, 0,
832       "Verify isSTRICT_UTF8_CHAR() with too short length parameter returns 0");
833
834    is(scalar @warnings, 0, "Verify isSTRICT_UTF8_CHAR() generated no warnings")
835      or output_warnings(@warnings);
836
837    undef @warnings;
838
839    $ret = test_isUTF8_CHAR_flags($bytes, $len,
840                                            $::UTF8_DISALLOW_ILLEGAL_INTERCHANGE);
841    is($ret, $expected_len,
842                    "Verify isUTF8_CHAR_flags('DISALLOW_ILLEGAL_INTERCHANGE')"
843                  . " acts like isSTRICT_UTF8_CHAR");
844
845    is(scalar @warnings, 0,
846               "Verify isUTF8_CHAR() for $hex_n generated no warnings")
847      or output_warnings(@warnings);
848
849    undef @warnings;
850
851    $ret = test_isC9_STRICT_UTF8_CHAR($bytes, $len);
852    $expected_len = ($valid_under_c9strict) ? $len : 0;
853    is($ret, $expected_len, "Verify isC9_STRICT_UTF8_CHAR($display_bytes)"
854                          . " returns expected length: $len");
855
856    is(scalar @warnings, 0,
857            "Verify isC9_STRICT_UTF8_CHAR() for $hex_n generated no warnings")
858      or output_warnings(@warnings);
859
860    undef @warnings;
861
862    $ret = test_isC9_STRICT_UTF8_CHAR($bytes, $len - 1);
863    is($ret, 0,
864    "Verify isC9_STRICT_UTF8_CHAR() with too short length parameter returns 0");
865
866    is(scalar @warnings, 0,
867               "Verify isC9_STRICT_UTF8_CHAR() generated no warnings")
868      or output_warnings(@warnings);
869
870    undef @warnings;
871
872    $ret = test_isUTF8_CHAR_flags($bytes, $len,
873                                        $::UTF8_DISALLOW_ILLEGAL_C9_INTERCHANGE);
874    is($ret, $expected_len,
875                   "Verify isUTF8_CHAR_flags('DISALLOW_ILLEGAL_C9_INTERCHANGE')"
876                  ." acts like isC9_STRICT_UTF8_CHAR");
877
878    is(scalar @warnings, 0,
879               "Verify isUTF8_CHAR() for $hex_n generated no warnings")
880      or output_warnings(@warnings);
881
882    undef @warnings;
883
884    $ret_ref = test_valid_utf8_to_uvchr($bytes);
885    is($ret_ref->[0], $n,
886                   "Verify valid_utf8_to_uvchr($display_bytes) returns $hex_n");
887    is($ret_ref->[1], $len,
888       "Verify valid_utf8_to_uvchr() for $hex_n returns expected length: $len");
889
890    is(scalar @warnings, 0,
891               "Verify valid_utf8_to_uvchr() for $hex_n generated no warnings")
892      or output_warnings(@warnings);
893
894    # Similarly for uvchr_to_utf8
895    my $this_uvchr_flags = $look_for_everything_uvchr_to;
896    if ($n > $highest_non_extended_cp) {
897        $this_uvchr_flags &=
898            ~($::UNICODE_DISALLOW_PERL_EXTENDED|$::UNICODE_WARN_PERL_EXTENDED);
899    }
900    if ($n > 0x10FFFF) {
901        $this_uvchr_flags &= ~($::UNICODE_DISALLOW_SUPER|$::UNICODE_WARN_SUPER);
902    }
903    elsif (($n >= 0xFDD0 && $n <= 0xFDEF) || ($n & 0xFFFE) == 0xFFFE) {
904        $this_uvchr_flags
905                     &= ~($::UNICODE_DISALLOW_NONCHAR|$::UNICODE_WARN_NONCHAR);
906    }
907    elsif ($n >= 0xD800 && $n <= 0xDFFF) {
908        $this_uvchr_flags
909                &= ~($::UNICODE_DISALLOW_SURROGATE|$::UNICODE_WARN_SURROGATE);
910    }
911    $display_flags = sprintf "0x%x", $this_uvchr_flags;
912
913    undef @warnings;
914
915    $ret = test_uvchr_to_utf8_flags($n, $this_uvchr_flags);
916    ok(defined $ret,
917        "Verify uvchr_to_utf8_flags($hex_n, $display_flags) returned success");
918    is($ret, $bytes,
919    "Verify uvchr_to_utf8_flags($hex_n, $display_flags) returns correct bytes");
920
921    is(scalar @warnings, 0,
922                "Verify uvchr_to_utf8_flags($hex_n, $display_flags) for $hex_n"
923              . " generated no warnings")
924      or output_warnings(@warnings);
925
926    # Now append this code point to a string that we will test various
927    # versions of is_foo_utf8_string_bar on, and keep a count of how many code
928    # points are in it.  All the code points in this loop are valid in Perl's
929    # extended UTF-8, but some are not valid under various restrictions.  A
930    # string and count is kept separately that is entirely valid for each
931    # restriction.  And, for each restriction, we note the first occurrence in
932    # the unrestricted string where we find something not in the restricted
933    # string.
934    $restriction_types{""}{'valid_strings'} .= $bytes;
935    $restriction_types{""}{'valid_counts'}++;
936
937    if ($valid_under_c9strict) {
938        $restriction_types{"c9strict"}{'valid_strings'} .= $bytes;
939        $restriction_types{"c9strict"}{'valid_counts'}++;
940    }
941    elsif (! exists $restriction_types{"c9strict"}{'first_invalid_offset'}) {
942        $restriction_types{"c9strict"}{'first_invalid_offset'}
943                    = length $restriction_types{"c9strict"}{'valid_strings'};
944        $restriction_types{"c9strict"}{'first_invalid_count'}
945                            = $restriction_types{"c9strict"}{'valid_counts'};
946    }
947
948    if ($valid_under_strict) {
949        $restriction_types{"strict"}{'valid_strings'} .= $bytes;
950        $restriction_types{"strict"}{'valid_counts'}++;
951    }
952    elsif (! exists $restriction_types{"strict"}{'first_invalid_offset'}) {
953        $restriction_types{"strict"}{'first_invalid_offset'}
954                        = length $restriction_types{"strict"}{'valid_strings'};
955        $restriction_types{"strict"}{'first_invalid_count'}
956                                = $restriction_types{"strict"}{'valid_counts'};
957    }
958
959    if ($valid_for_not_extended_utf8) {
960        $restriction_types{"not_extended_utf8"}{'valid_strings'} .= $bytes;
961        $restriction_types{"not_extended_utf8"}{'valid_counts'}++;
962    }
963    elsif (! exists
964                $restriction_types{"not_extended_utf8"}{'first_invalid_offset'})
965    {
966        $restriction_types{"not_extended_utf8"}{'first_invalid_offset'}
967                = length $restriction_types{"not_extended_utf8"}{'valid_strings'};
968        $restriction_types{"not_extended_utf8"}{'first_invalid_count'}
969                        = $restriction_types{"not_extended_utf8"}{'valid_counts'};
970    }
971}
972
973my $I8c = (isASCII) ? "\x80" : "\xa0";    # A continuation byte
974my $cont_byte = I8_to_native($I8c);
975my $p = (isASCII) ? "\xe1\x80" : I8_to_native("\xE4\xA0");  # partial
976
977# The loop above tested the single or partial character functions/macros,
978# while building up strings to test the string functions, which we do now.
979
980for my $restriction (sort keys %restriction_types) {
981    use bytes;
982
983    for my $use_flags ("", "_flags") {
984
985        # For each restriction, we test it in both the is_foo_flags functions
986        # and the specially named foo function.  But not if there isn't such a
987        # specially named function.  Currently, this is the only tested
988        # restriction that doesn't have a specially named function
989        next if $use_flags eq "" && $restriction eq "not_extended_utf8";
990
991        # Start building up the name of the function we will test.
992        my $base_name = "is_";
993
994        if (! $use_flags  && $restriction ne "") {
995            $base_name .= $restriction . "_";
996        }
997
998        # We test both "is_utf8_string_foo" and "is_fixed_width_buf" functions
999        foreach my $operand ('string', 'fixed_width_buf') {
1000
1001            # Currently, only fixed_width_buf functions have the '_flags'
1002            # suffix.
1003            next if $operand eq 'fixed_width_buf' && $use_flags eq "";
1004
1005            my $name = "${base_name}utf8_$operand";
1006
1007            # We test each version of the function
1008            for my $function ("_loclen", "_loc", "") {
1009
1010                # We test each function against
1011                #   a) valid input
1012                #   b) invalid input created by appending an out-of-place
1013                #      continuation character to the valid string
1014                #   c) input created by appending a partial character.  This
1015                #      is valid in the 'fixed_width' functions, but invalid in
1016                #      the 'string' ones
1017                #   d) invalid input created by calling a function that is
1018                #      expecting a restricted form of the input using the string
1019                #      that's valid when unrestricted
1020                for my $error_type (0, $cont_byte, $p, $restriction) {
1021                    #diag "restriction=$restriction, use_flags=$use_flags, function=$function, error_type=" . display_bytes($error_type);
1022
1023                    # If there is no restriction, the error type will be "",
1024                    # which is redundant with 0.
1025                    next if $error_type eq "";
1026
1027                    my $this_name = "$name$function$use_flags";
1028                    my $bytes
1029                            = $restriction_types{$restriction}{'valid_strings'};
1030                    my $expected_offset = length $bytes;
1031                    my $expected_count
1032                            = $restriction_types{$restriction}{'valid_counts'};
1033                    my $test_name_suffix = "";
1034
1035                    my $this_error_type = $error_type;
1036                    if ($this_error_type) {
1037
1038                        # Appending a bare continuation byte or a partial
1039                        # character doesn't change the character count or
1040                        # offset.  But in the other cases, we have saved where
1041                        # the failures should occur, so use those.  Appending
1042                        # a continuation byte makes it invalid; appending a
1043                        # partial character makes the 'string' form invalid,
1044                        # but not the 'fixed_width_buf' form.
1045                        if (   $this_error_type eq $cont_byte
1046                            || $this_error_type eq $p)
1047                        {
1048                            $bytes .= $this_error_type;
1049                            if ($this_error_type eq $cont_byte) {
1050                                $test_name_suffix
1051                                            = " for an unexpected continuation";
1052                            }
1053                            else {
1054                                $test_name_suffix
1055                                        = " if ends with a partial character";
1056                                $this_error_type
1057                                        = 0 if $operand eq "fixed_width_buf";
1058                            }
1059                        }
1060                        elsif (! exists $restriction_types
1061                                    {$this_error_type}{'first_invalid_count'})
1062                        {
1063                            # If no errors were found, this is entirely valid.
1064                            $this_error_type = 0;
1065                        }
1066                        else {
1067
1068                            if (! exists $restriction_types{$this_error_type}) {
1069                                fail("Internal test error: Unknown error type "
1070                                . "'$this_error_type'");
1071                                next;
1072                            }
1073                            $test_name_suffix
1074                                        = " if contains forbidden code points";
1075
1076                            $bytes = $restriction_types{""}{'valid_strings'};
1077                            $expected_offset
1078                                 = $restriction_types{$this_error_type}
1079                                                     {'first_invalid_offset'};
1080                            $expected_count
1081                                  = $restriction_types{$this_error_type }
1082                                                      {'first_invalid_count'};
1083                        }
1084                    }
1085
1086                    my $length = length $bytes;
1087                    my $ret_ref;
1088
1089                    my $test = "\$ret_ref = test_$this_name(\$bytes, $length";
1090
1091                    # If using the _flags functions, we have to figure out what
1092                    # flags to pass.  This is done to match the restriction.
1093                    if ($use_flags eq "_flags") {
1094                        if (! $restriction) {
1095                            $test .= ", 0";     # The flag
1096
1097                            # Indicate the kind of flag in the test name.
1098                            $this_name .= "(0)";
1099                        }
1100                        else {
1101                            $this_name .= "($restriction)";
1102                            if ($restriction eq "c9strict") {
1103                                $test
1104                                  .= ", $::UTF8_DISALLOW_ILLEGAL_C9_INTERCHANGE";
1105                            }
1106                            elsif ($restriction eq "strict") {
1107                                $test .= ", $::UTF8_DISALLOW_ILLEGAL_INTERCHANGE";
1108                            }
1109                            elsif ($restriction eq "not_extended_utf8") {
1110                                $test .= ", $::UTF8_DISALLOW_PERL_EXTENDED";
1111                            }
1112                            else {
1113                                fail("Internal test error: Unknown restriction "
1114                                . "'$restriction'");
1115                                next;
1116                            }
1117                        }
1118                    }
1119                    $test .= ")";
1120
1121                    # Actually run the test
1122                    eval $test;
1123                    if ($@) {
1124                        fail($test);
1125                        diag $@;
1126                        next;
1127                    }
1128
1129                    my $ret;
1130                    my $error_offset;
1131                    my $cp_count;
1132
1133                    if ($function eq "") {
1134                        $ret = $ret_ref;    # For plain function, there's only a
1135                                            # single return value
1136                    }
1137                    else {  # Otherwise, the multiple values come in an array.
1138                        $ret = shift @$ret_ref ;
1139                        $error_offset = shift @$ret_ref;
1140                        $cp_count = shift@$ret_ref if $function eq "_loclen";
1141                    }
1142
1143                    if ($this_error_type) {
1144                        is($ret, 0,
1145                           "Verify $this_name is FALSE$test_name_suffix");
1146                    }
1147                    else {
1148                        unless(is($ret, 1,
1149                                  "Verify $this_name is TRUE for valid input"
1150                                . "$test_name_suffix"))
1151                        {
1152                            diag("    The bytes starting at offset"
1153                               . " $error_offset are"
1154                               . display_bytes(substr(
1155                                          $restriction_types{$restriction}
1156                                                            {'valid_strings'},
1157                                          $error_offset)));
1158                            next;
1159                        }
1160                    }
1161
1162                    if ($function ne "") {
1163                        unless (is($error_offset, $expected_offset,
1164                                   "\tAnd returns the correct offset"))
1165                        {
1166                            my $min = ($error_offset < $expected_offset)
1167                                    ? $error_offset
1168                                    : $expected_offset;
1169                            diag("    The bytes starting at offset" . $min
1170                              . " are " . display_bytes(substr($bytes, $min)));
1171                        }
1172
1173                        if ($function eq '_loclen') {
1174                            is($cp_count, $expected_count,
1175                               "\tAnd returns the correct character count");
1176                        }
1177                    }
1178                }
1179            }
1180        }
1181    }
1182}
1183
1184SKIP:
1185{
1186    my $simple = join "", "A" .. "J";
1187    my $utf_ch = "\x{3f_ffff}";     # Highest code point that is same number
1188                                    # of bytes on ASCII and EBCDIC: 5
1189    utf8::encode($utf_ch);
1190    my $utf_ch_len = length $utf_ch;
1191    note "utf_ch_len $utf_ch_len";
1192    my $utf = $utf_ch x 10;
1193    my $bad_start = substr($utf, 1);
1194    # $bad_end ends with a start byte and a single continuation
1195    my $bad_end = substr($utf, 0, length($utf)-$utf_ch_len+2);
1196
1197    my @hop_tests =
1198      (  #           start byte      chars
1199       # string      in 'string'     to hop      expected         name
1200       [ $simple,    0,               5,         5,               "simple in range, forward" ],
1201       [ $simple,    10,              -5,        5,               "simple in range, backward" ],
1202       [ $simple,    5,               10,        10,              "simple out of range, forward" ],
1203       [ $simple,    5,               -10,       0,               "simple out of range, backward" ],
1204       [ $utf,       $utf_ch_len * 5, 5,         length($utf),    "utf in range, forward" ],
1205       [ $utf,       $utf_ch_len * 5, -5,        0,               "utf in range, backward" ],
1206       [ $utf,       $utf_ch_len * 5, 4,         $utf_ch_len * 9, "utf in range b, forward" ],
1207       [ $utf,       $utf_ch_len * 5, -4,        $utf_ch_len,     "utf in range b, backward" ],
1208       [ $utf,       $utf_ch_len * 5, 6,         length($utf),    "utf out of range, forward" ],
1209       [ $utf,       $utf_ch_len * 5, -6,        0,               "utf out of range, backward"  ],
1210       [ $bad_start, 0,               1,         $utf_ch_len-1,   "bad start, forward 1 from 0" ],
1211       [ $bad_start, 0,               5,         5 * $utf_ch_len-1, "bad start, forward 5 chars from 0" ],
1212       [ $bad_start, 0,                9,        length($bad_start)-$utf_ch_len, "bad start, forward 9 chars from 0" ],
1213       [ $bad_start, 0,               10,        length $bad_start, "bad start, forward 10 chars from 0" ],
1214       [ $bad_start, $utf_ch_len-1,   -1,        0,                "bad start, back 1 from first start byte" ],
1215       [ $bad_start, $utf_ch_len-2,   -1,        0,                "bad start, back 1 from before first start byte" ],
1216       [ $bad_start, 0,               -1,        0,                "bad start, back 1 from 0" ],
1217       [ $bad_start, length $bad_start, -10,     0,                "bad start, back 10 from end" ],
1218       [ $bad_end,   0,               10,        length $bad_end, "bad end, forward 10 from 0" ],
1219       [ $bad_end,   length($bad_end)-1, 10,     length $bad_end, "bad end, forward 1 from end-1" ],
1220       );
1221
1222    for my $test (@hop_tests) {
1223        my ($str, $s_off, $hop, $want, $name) = @$test;
1224        my $result = test_utf8_hop_safe($str, $s_off, $hop);
1225        is($result, $want, "utf8_hop_safe: $name");
1226    }
1227}
1228
1229{
1230    my $replacement = chr(0xFFFD);
1231    use bytes;
1232    is(test_UTF8_IS_REPLACEMENT($replacement, length $replacement), 1,
1233       "UTF8_IS_REPLACEMENT returns 1 on a REPLACEMENT character");
1234    is(test_UTF8_IS_REPLACEMENT($replacement, length $replacement) - 1, 0,
1235       "UTF8_IS_REPLACEMENT returns 0 on too short an input");
1236}
1237
1238done_testing;
1239