This is a live mirror of the Perl 5 development currently hosted at https://github.com/perl/perl5
Fix skipping char in (*sr:...) defn
[perl5.git] / t / re / script_run.t
CommitLineData
034602eb
KW
1BEGIN {
2 chdir 't' if -d 't';
3 require './test.pl';
4 set_up_inc('../lib');
5}
6
7use strict;
8use warnings;
9
10$|=1;
11
12# The Script_Extension property has only recently become reasonably stable, so
13# later Unicode releases may change things. Some of these tests were
14# designed to provide more code covereage in regexec.c, so changes in it or
15# later Standards could cause them to not test what they originally were aimed
16# to do.
17
18no warnings "experimental::script_run";
19
526c4163
KW
20# Since there's so few tests currently, we can afford to try each syntax on
21# all of them
22foreach my $type ('script_run', 'sr') {
23 my $script_run;
24 eval '$script_run = qr/ ^ (*$type: .* ) $ /x;';
034602eb 25
e54417e0
KW
26 unlike("\N{CYRILLIC SMALL LETTER ER}\N{CYRILLIC SMALL LETTER A}\N{CYRILLIC SMALL LETTER U}}\N{CYRILLIC SMALL LETTER ER}\N{CYRILLIC SMALL LETTER A}l", $script_run, "Cyrillic 'paypal' with a Latin 'l' is not a script run");
27 unlike("A\N{GREEK CAPITAL LETTER GAMMA}", $script_run, "Latin followed by Greek isn't a script run");
034602eb 28
e54417e0
KW
29 like("\N{CYRILLIC THOUSANDS SIGN}\N{COMBINING CYRILLIC TITLO}", $script_run, "Cyrillic followed by Permic-Arabic is Arabic");
30 like("\N{OLD PERMIC LETTER AN}\N{COMBINING CYRILLIC TITLO}", $script_run, "Permic followed by Permic-Arabic is Permic");
31 unlike("\N{GLAGOLITIC CAPITAL LETTER AZU}\N{COMBINING CYRILLIC TITLO}", $script_run, "Glagolithic followed by Permic-Arabic isn't a script run");
034602eb 32
e54417e0
KW
33 like("\N{CYRILLIC THOUSANDS SIGN}\N{COMBINING CYRILLIC PALATALIZATION}", $script_run, "Cyrillic followed by Glagolithic-Arabic is Arabic");
34 like("\N{GLAGOLITIC CAPITAL LETTER AZU}\N{COMBINING CYRILLIC PALATALIZATION}", $script_run, "Glagolithic followed by Glagolithic-Arabic is Glagolithic");
35 unlike("\N{OLD PERMIC LETTER AN}\N{COMBINING CYRILLIC PALATALIZATION}", $script_run, "Permic followed by Glagolithic-Arabic isn't a script run");
034602eb 36
e54417e0
KW
37 like("\N{ARABIC-INDIC DIGIT ZERO}\N{ARABIC-INDIC DIGIT ONE}\N{ARABIC-INDIC DIGIT TWO}\N{ARABIC-INDIC DIGIT THREE}\N{ARABIC COMMA}\N{ARABIC-INDIC DIGIT FOUR}\N{THAANA LETTER HAA}", $script_run, "Arabic-Thaana chars followed by Thaana is Thaana");
38 unlike("\N{ARABIC-INDIC DIGIT ZERO}\N{ARABIC-INDIC DIGIT ONE}A", $script_run, "Arabic-Thaana chars followed by Latin isn't a script run");
39 like("\N{ARABIC-INDIC DIGIT ZERO}\N{ARABIC-INDIC DIGIT ONE}\N{ARABIC-INDIC DIGIT TWO}\N{ARABIC-INDIC DIGIT THREE}\N{ARABIC COMMA}\N{ARABIC-INDIC DIGIT FOUR}\N{ARABIC NUMBER SIGN}", $script_run, "Arabic-Thaana chars followed by Arabic is Arabic");
40 unlike("\N{ARABIC-INDIC DIGIT ZERO}\N{ARABIC-INDIC DIGIT ONE}\N{ARABIC-INDIC DIGIT TWO}\N{ARABIC-INDIC DIGIT THREE}\N{EXTENDED ARABIC-INDIC DIGIT NINE}", $script_run, "Arabic-Thaana digits followed by an Arabic digit from a different sequence isn't a script run");
41 like("\N{ARABIC-INDIC DIGIT ZERO}\N{ARABIC-INDIC DIGIT ONE}\N{ARABIC-INDIC DIGIT TWO}\N{ARABIC-INDIC DIGIT THREE}\N{THAANA LETTER HAA}", $script_run, "Arabic-Thaana digits followed by a Thaana leter is a script run");
034602eb 42
e54417e0
KW
43 # The next tests are at a hard-coded boundary in regexec.c at the time of this
44 # writing (U+02B9/02BA).
45 like("abc\N{MODIFIER LETTER SMALL Y}", $script_run, "All Latin is a script run");
46 like("abc\N{MODIFIER LETTER PRIME}", $script_run, "Latin then Common is a script run");
47 like(":a", $script_run, "Common then Latin is a script run");
48 like("-\N{SINHALA LETTER RAYANNA}", $script_run, "Common then Sinhala (which has its own 0) is a script run");
81eabee7 49
e54417e0
KW
50 like("\N{HEBREW LETTER ALEF}\N{HEBREW LETTER TAV}\N{MODIFIER LETTER PRIME}", $script_run, "Hebrew then Common is a script run");
51 unlike("\N{HEBREW LETTER ALEF}\N{HEBREW LETTER TAV}\N{MODIFIER LETTER SMALL Y}", $script_run, "Hebrew then Latin isn't a script run");
52 like("9876543210\N{DESERET SMALL LETTER WU}", $script_run, "0-9 are the digits for Deseret");
53 like("\N{DESERET SMALL LETTER WU}9876543210", $script_run, "Also when they aren't in the initial position");
54 unlike("\N{DESERET SMALL LETTER WU}\N{FULLWIDTH DIGIT FIVE}", $script_run, "Fullwidth digits aren't the digits for Deseret");
55 unlike("\N{FULLWIDTH DIGIT SIX}\N{DESERET SMALL LETTER LONG I}", $script_run, "... likewise if the digits come first");
034602eb 56
e54417e0
KW
57 like("1234567890\N{ARABIC LETTER ALEF}", $script_run, "[0-9] work for Arabic");
58 unlike("1234567890\N{ARABIC LETTER ALEF}\N{ARABIC-INDIC DIGIT FOUR}\N{ARABIC-INDIC DIGIT FIVE}", $script_run, "... but not in combination with real ARABIC digits");
59 unlike("\N{ARABIC LETTER ALEF}\N{ARABIC-INDIC DIGIT SIX}\N{ARABIC-INDIC DIGIT SEVEN}1", $script_run, "... nor when the ARABIC digits come before them");
034602eb 60
e54417e0
KW
61 # This exercises the case where the script zero but not the script is
62 # ambiguous until a non-ambiguous digit is found.
63 like("\N{ARABIC LETTER ALEF}\N{EXTENDED ARABIC-INDIC DIGIT EIGHT}", $script_run, "ARABIC with a Shia digit is a script run");
034602eb 64
e54417e0
KW
65 like("\N{U+03A2}", $script_run, "A single unassigned code point is a script run");
66 unlike("\N{U+03A2}\N{U+03A2}", $script_run, "But not more than one");
67 unlike("A\N{U+03A2}", $script_run, "... and not in combination with an assigned one");
68 unlike("\N{U+03A2}A", $script_run, "... in either order");
69 unlike("\N{U+03A2}0", $script_run, "... nor with a digit following");
034602eb 70
e54417e0
KW
71 like("A\N{COMBINING GRAVE ACCENT}", $script_run, "An inherited script matches others");
72 like("\N{COMBINING GRAVE ACCENT}A", $script_run, "... even if first in the sequence");
034602eb 73
e54417e0 74 like("\N{COMBINING TILDE}\N{COMBINING GRAVE ACCENT}", $script_run, "A script containing only inherited characters matches");
b548543c 75
e54417e0 76 like("\N{DEVANAGARI DOUBLE DANDA}\N{DEVANAGARI DANDA}\N{DEVANAGARI STRESS SIGN UDATTA}\N{DEVANAGARI STRESS SIGN ANUDATTA}\N{NORTH INDIC FRACTION ONE QUARTER}\N{NORTH INDIC QUANTITY MARK}", $script_run, "A bunch of narrowing down of multiple possible scripts");
034602eb 77
e54417e0
KW
78 unlike("\N{JAVANESE PANGRANGKEP}\N{GEORGIAN PARAGRAPH SEPARATOR}", $script_run, "Two code points each in multiple scripts that don't intersect aren't a script run");
79 like("\N{DEVANAGARI SIGN CANDRABINDU VIRAMA}\N{VEDIC TONE YAJURVEDIC KATHAKA INDEPENDENT SVARITA}", $script_run, "Two code points each in multiple scripts that 't intersect singly are a script run");
034602eb 80
e54417e0 81 like("", $script_run, "An empty string is a script run");
ad85be64 82
e54417e0 83 use utf8;
034602eb 84
e54417e0
KW
85 # From UTS 39
86 like("写真だけの結婚式", $script_run, "Mixed Hiragana and Han");
526c4163 87}
034602eb 88
8638266f
KW
89 # Until fixed, this was skipping the '['
90 unlike("abc]c", qr/^ (*sr:a(*sr:[bc]*)c) $/x, "Doesn't skip parts of exact matches");
91
034602eb 92done_testing();