3 $|=1; # outherwise things get mixed up in output
8 set_up_inc( qw '../lib ../ext/re' );
9 eval 'require Config'; # assume defaults if this fails
12 skip_all_without_unicode_tables();
15 use open qw(:utf8 :std);
17 # Kind of a kludge to mark warnings to be expected only if we are testing
18 # under "use re 'strict'"
19 my $only_strict_marker = ':expected_only_under_strict';
21 ## If the markers used are changed (search for "MARKER1" in regcomp.c),
22 ## update only these two regexs, and leave the {#} in the @death/@warning
23 ## arrays below. The {#} is a meta-marker -- it marks where the marker should
26 sub fixup_expect ($$) {
28 # Fixes up the expected results by inserting the boiler plate text.
29 # Returns empty string if that is what is expected. Otherwise, handles
30 # either a scalar, turning it into a single element array; or a ref to an
31 # array, adjusting each element. If called in array context, returns an
32 # array, otherwise the join of all elements.
34 # The string $only_strict_marker will be removed from any expect line it
35 # begins, and if $strict is not true, that expect line will be removed
36 # from the output (hence won't be expected)
38 my ($expect_ref, $strict) = @_;
39 return "" if $expect_ref eq "";
42 if (ref $expect_ref) {
43 @expect = @$expect_ref;
46 @expect = $expect_ref;
50 foreach my $element (@expect) {
51 $element =~ s/\{\#\}/in regex; marked by <-- HERE in/;
52 $element =~ s/\{\#\}/ <-- HERE /;
54 next if $element =~ s/ ^ $only_strict_marker \s* //x && ! $strict;
55 push @new_expect, $element;
57 return wantarray ? @new_expect : join "", @new_expect;
62 $element =~ s/ at .* line \d+\.?\n$//;
63 $element =~ s/in regex; marked by <-- HERE in/{#}/;
64 $element =~ s/ <-- HERE /{#}/;
68 ## Because we don't "use utf8" in this file, we need to do some extra legwork
69 ## for the utf8 tests: Prepend 'use utf8' to the pattern, and mark the strings
70 ## to check against as UTF-8, but for this all to work properly, the character
71 ## 'ネ' (U+30CD) is required in each pattern somewhere as a marker.
73 ## This also creates a second variant of the tests to check if the
74 ## latin1 error messages are working correctly. Because we don't 'use utf8',
75 ## we can't tell if something is UTF-8 or Latin1, so you need the suffix
76 ## '; no latin1' to not have the second variant.
78 my $utf8 = "\x{30cd}";
83 for (my $i = 0; $i < @_; $i += 2) {
86 my $l1_pat = $pat =~ s/$utf8/$l1/gr;
88 $pat = "use utf8; $pat";
91 $l1_msg = [ map { s/$utf8/$l1/gr } @$msg ];
92 @$msg = map { my $c = $_; utf8::decode($c); $c } @$msg;
95 $l1_msg = $msg =~ s/$utf8/$l1/gr;
98 push @ret, $pat => $msg;
100 push @ret, $l1_pat => $l1_msg unless $l1_pat =~ /#no latin1/;
105 my $inf_m1 = ($Config::Config{reg_infty} || 65535) - 1;
106 my $inf_p1 = $inf_m1 + 2;
108 my $B_hex = sprintf("\\x%02X", ord "B");
109 my $low_mixed_alpha = ('A' lt 'a') ? 'A' : 'a';
110 my $high_mixed_alpha = ('A' lt 'a') ? 'a' : 'A';
111 my $low_mixed_digit = ('A' lt '0') ? 'A' : '0';
112 my $high_mixed_digit = ('A' lt '0') ? '0' : 'A';
114 my $colon_hex = sprintf "%02X", ord(":");
115 my $tab_hex = sprintf "%02X", ord("\t");
117 # Key-value pairs of strings eval'd as patterns => warn/error messages that
118 # they should generate. In some cases, the value is an array of multiple
119 # messages. Some groups have the message(s) be default on; others, default
120 # off. This can be overridden on an individual key basis by preceding the
121 # pattern string with either 'default_on' or 'default_off'
123 # The first set are those that should be fatal errors.
125 my $bug133423 = "(?[(?^:(?[\\\x00]))\\]\x00|2[^^]\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80])R.\\670";
129 '/[[=foo=]]/' => 'POSIX syntax [= =] is reserved for future extensions {#} m/[[=foo=]{#}]/',
131 '/(?<= .*)/' => 'Lookbehind longer than 255 not implemented in regex m/(?<= .*)/',
133 '/(?<= x{1000})/' => 'Lookbehind longer than 255 not implemented in regex m/(?<= x{1000})/',
135 '/(?@)/' => 'Sequence (?@...) not implemented {#} m/(?@{#})/',
137 '/(?{ 1/' => 'Missing right curly or square bracket',
139 '/(?(1x))/' => 'Switch condition not recognized {#} m/(?(1x{#}))/',
140 '/(?(1x(?#)))/'=> 'Switch condition not recognized {#} m/(?(1x{#}(?#)))/',
142 '/(?(1)/' => 'Switch (?(condition)... not terminated {#} m/(?(1){#}/',
143 '/(?(1)x/' => 'Switch (?(condition)... not terminated {#} m/(?(1)x{#}/',
144 '/(?(1)x|y/' => 'Switch (?(condition)... not terminated {#} m/(?(1)x|y{#}/',
145 '/(?(1)x|y|z)/' => 'Switch (?(condition)... contains too many branches {#} m/(?(1)x|y|{#}z)/',
147 '/(?(x)y|x)/' => 'Unknown switch condition (?(...)) {#} m/(?(x{#})y|x)/',
148 '/(?(??{}))/' => 'Unknown switch condition (?(...)) {#} m/(?(?{#}?{}))/',
149 '/(?(?[]))/' => 'Unknown switch condition (?(...)) {#} m/(?(?{#}[]))/',
151 '/(?/' => 'Sequence (? incomplete {#} m/(?{#}/',
153 '/(?;x/' => 'Sequence (?;...) not recognized {#} m/(?;{#}x/',
154 '/(?<;x/' => 'Group name must start with a non-digit word character {#} m/(?<;{#}x/',
155 '/(?\ix/' => 'Sequence (?\...) not recognized {#} m/(?\{#}ix/',
156 '/(?\mx/' => 'Sequence (?\...) not recognized {#} m/(?\{#}mx/',
157 '/(?\:x/' => 'Sequence (?\...) not recognized {#} m/(?\{#}:x/',
158 '/(?\=x/' => 'Sequence (?\...) not recognized {#} m/(?\{#}=x/',
159 '/(?\!x/' => 'Sequence (?\...) not recognized {#} m/(?\{#}!x/',
160 '/(?\<=x/' => 'Sequence (?\...) not recognized {#} m/(?\{#}<=x/',
161 '/(?\<!x/' => 'Sequence (?\...) not recognized {#} m/(?\{#}<!x/',
162 '/(?\>x/' => 'Sequence (?\...) not recognized {#} m/(?\{#}>x/',
163 '/(?^-i:foo)/' => 'Sequence (?^-...) not recognized {#} m/(?^-{#}i:foo)/',
164 '/(?^-i)foo/' => 'Sequence (?^-...) not recognized {#} m/(?^-{#}i)foo/',
165 '/(?^d:foo)/' => 'Sequence (?^d...) not recognized {#} m/(?^d{#}:foo)/',
166 '/(?^d)foo/' => 'Sequence (?^d...) not recognized {#} m/(?^d{#})foo/',
167 '/(?^lu:foo)/' => 'Regexp modifiers "l" and "u" are mutually exclusive {#} m/(?^lu{#}:foo)/',
168 '/(?^lu)foo/' => 'Regexp modifiers "l" and "u" are mutually exclusive {#} m/(?^lu{#})foo/',
169 '/(?da:foo)/' => 'Regexp modifiers "d" and "a" are mutually exclusive {#} m/(?da{#}:foo)/',
170 '/(?lil:foo)/' => 'Regexp modifier "l" may not appear twice {#} m/(?lil{#}:foo)/',
171 '/(?aaia:foo)/' => 'Regexp modifier "a" may appear a maximum of twice {#} m/(?aaia{#}:foo)/',
172 '/(?i-l:foo)/' => 'Regexp modifier "l" may not appear after the "-" {#} m/(?i-l{#}:foo)/',
174 '/((x)/' => 'Unmatched ( {#} m/({#}(x)/',
175 '/{(}/' => 'Unmatched ( {#} m/{({#}}/', # [perl #127599]
177 "/x{$inf_p1}/" => "Quantifier in {,} bigger than $inf_m1 {#} m/x{{#}$inf_p1}/",
180 '/x**/' => 'Nested quantifiers {#} m/x**{#}/',
182 '/x[/' => 'Unmatched [ {#} m/x[{#}/',
184 '/*/', => 'Quantifier follows nothing {#} m/*{#}/',
186 '/\p{x/' => 'Missing right brace on \p{} {#} m/\p{{#}x/',
188 '/[\p{x]/' => 'Missing right brace on \p{} {#} m/[\p{{#}x]/',
190 '/(x)\2/' => 'Reference to nonexistent group {#} m/(x)\2{#}/',
192 '/\g/' => 'Unterminated \g... pattern {#} m/\g{#}/',
193 '/\g{1/' => 'Unterminated \g{...} pattern {#} m/\g{1{#}/',
195 'my $m = "\\\"; $m =~ $m', => 'Trailing \ in regex m/\/',
197 '/\x{1/' => 'Missing right brace on \x{} {#} m/\x{1{#}/',
198 '/\x{X/' => 'Missing right brace on \x{} {#} m/\x{{#}X/',
200 '/[\x{X]/' => 'Missing right brace on \x{} {#} m/[\x{{#}X]/',
201 '/[\x{A]/' => 'Missing right brace on \x{} {#} m/[\x{A{#}]/',
203 '/\o{1/' => 'Missing right brace on \o{ {#} m/\o{1{#}/',
204 '/\o{X/' => 'Missing right brace on \o{ {#} m/\o{{#}X/',
206 '/[\o{X]/' => 'Missing right brace on \o{ {#} m/[\o{{#}X]/',
207 '/[\o{7]/' => 'Missing right brace on \o{ {#} m/[\o{7{#}]/',
209 '/[[:barf:]]/' => 'POSIX class [:barf:] unknown {#} m/[[:barf:]{#}]/',
211 '/[[=barf=]]/' => 'POSIX syntax [= =] is reserved for future extensions {#} m/[[=barf=]{#}]/',
213 '/[[.barf.]]/' => 'POSIX syntax [. .] is reserved for future extensions {#} m/[[.barf.]{#}]/',
215 '/[z-a]/' => 'Invalid [] range "z-a" {#} m/[z-a{#}]/',
217 '/\p/' => 'Empty \p {#} m/\p{#}/',
218 '/\P/' => 'Empty \P {#} m/\P{#}/',
219 '/\p{}/' => 'Empty \p{} {#} m/\p{{#}}/',
220 '/\P{}/' => 'Empty \P{} {#} m/\P{{#}}/',
222 '/a\b{cde/' => 'Missing right brace on \b{} {#} m/a\b{{#}cde/',
223 '/a\B{cde/' => 'Missing right brace on \B{} {#} m/a\B{{#}cde/',
225 '/\b{}/' => 'Empty \b{} {#} m/\b{}{#}/',
226 '/\B{}/' => 'Empty \B{} {#} m/\B{}{#}/',
228 '/\b{gc}/' => "'gc' is an unknown bound type {#} m/\\b{gc{#}}/",
229 '/\B{gc}/' => "'gc' is an unknown bound type {#} m/\\B{gc{#}}/",
231 '/(?[[[::]]])/' => "Unexpected ']' with no following ')' in (?[... {#} m/(?[[[::]]{#}])/",
232 '/(?[[[:w:]]])/' => "Unexpected ']' with no following ')' in (?[... {#} m/(?[[[:w:]]{#}])/",
233 '/(?[a])/' => 'Unexpected character {#} m/(?[a{#}])/',
234 '/(?[ + \t ])/' => 'Unexpected binary operator \'+\' with no preceding operand {#} m/(?[ +{#} \t ])/',
235 '/(?[ \cK - ( + \t ) ])/' => 'Unexpected binary operator \'+\' with no preceding operand {#} m/(?[ \cK - ( +{#} \t ) ])/',
236 '/(?[ \cK ( \t ) ])/' => 'Unexpected \'(\' with no preceding operator {#} m/(?[ \cK ({#} \t ) ])/',
237 '/(?[ \cK \t ])/' => 'Operand with no preceding operator {#} m/(?[ \cK \t{#} ])/',
238 '/(?[ \0004 ])/' => 'Need exactly 3 octal digits {#} m/(?[ \0004 {#}])/',
239 '/(?[ \05 ])/' => 'Need exactly 3 octal digits {#} m/(?[ \05 {#}])/',
240 '/(?[ \o{1038} ])/' => 'Non-octal character {#} m/(?[ \o{1038{#}} ])/',
241 '/(?[ \o{} ])/' => 'Empty \o{} {#} m/(?[ \o{}{#} ])/',
242 '/(?[ \x{defg} ])/' => 'Non-hex character {#} m/(?[ \x{defg{#}} ])/',
243 '/(?[ \xabcdef ])/' => 'Use \\x{...} for more than two hex characters {#} m/(?[ \xabc{#}def ])/',
244 '/(?[ \x{} ])/' => 'Empty \x{} {#} m/(?[ \x{}{#} ])/',
245 '/(?[ \cK + ) ])/' => 'Unexpected \')\' {#} m/(?[ \cK + ){#} ])/',
246 '/(?[ \cK + ])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[ \cK + {#}])/',
247 '/(?[ ( ) ])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[ ( ){#} ])/',
248 '/(?[[0]+()+])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[[0]+(){#}+])/',
249 '/(?[ \p{foo} ])/' => 'Can\'t find Unicode property definition "foo" {#} m/(?[ \p{foo}{#} ])/',
250 '/(?[ \p{ foo = bar } ])/' => 'Can\'t find Unicode property definition "foo = bar" {#} m/(?[ \p{ foo = bar }{#} ])/',
251 '/(?[ \8 ])/' => 'Unrecognized escape \8 in character class {#} m/(?[ \8{#} ])/',
252 '/(?[ \t ]/' => "Unexpected ']' with no following ')' in (?[... {#} m/(?[ \\t ]{#}/",
253 '/(?[ [ \t ]/' => "Syntax error in (?[...]) {#} m/(?[ [ \\t ]{#}/",
254 '/(?[ \t ] ]/' => "Unexpected ']' with no following ')' in (?[... {#} m/(?[ \\t ]{#} ]/",
255 '/(?[ [ ] ]/' => "Syntax error in (?[...]) {#} m/(?[ [ ] ]{#}/",
256 '/(?[ \t + \e # This was supposed to be a comment ])/' =>
257 "Syntax error in (?[...]) {#} m/(?[ \\t + \\e # This was supposed to be a comment ]){#}/",
258 '/(?[ ])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[ {#}])/',
259 'm/(?[[a-\d]])/' => 'False [] range "a-\d" {#} m/(?[[a-\d{#}]])/',
260 'm/(?[[\w-x]])/' => 'False [] range "\w-" {#} m/(?[[\w-{#}x]])/',
261 'm/(?[[a-\pM]])/' => 'False [] range "a-\pM" {#} m/(?[[a-\pM{#}]])/',
262 'm/(?[[\pM-x]])/' => 'False [] range "\pM-" {#} m/(?[[\pM-{#}x]])/',
263 'm/(?[[^\N{LATIN CAPITAL LETTER A WITH MACRON AND GRAVE}]])/' => '\N{} here is restricted to one character {#} m/(?[[^\N{U+100.300{#}}]])/',
264 'm/(?[ \p{Digit} & (?^(?[ \p{Thai} | \p{Lao} ]))])/' => 'Sequence (?^(...) not recognized {#} m/(?[ \p{Digit} & (?^({#}?[ \p{Thai} | \p{Lao} ]))])/',
265 'm/(?[ \p{Digit} & (?(?[ \p{Thai} | \p{Lao} ]))])/' => 'Unexpected character {#} m/(?[ \p{Digit} & (?{#}(?[ \p{Thai} | \p{Lao} ]))])/',
266 'm/\o{/' => 'Missing right brace on \o{ {#} m/\o{{#}/',
267 'm/\o/' => 'Missing braces on \o{} {#} m/\o{#}/',
268 'm/\o{}/' => 'Empty \o{} {#} m/\o{}{#}/',
269 'm/[\o{]/' => 'Missing right brace on \o{ {#} m/[\o{{#}]/',
270 'm/[\o]/' => 'Missing braces on \o{} {#} m/[\o{#}]/',
271 'm/[\o{}]/' => 'Empty \o{} {#} m/[\o{}{#}]/',
272 'm/(?^-i:foo)/' => 'Sequence (?^-...) not recognized {#} m/(?^-{#}i:foo)/',
273 'm/\87/' => 'Reference to nonexistent group {#} m/\87{#}/',
274 'm/a\87/' => 'Reference to nonexistent group {#} m/a\87{#}/',
275 'm/a\97/' => 'Reference to nonexistent group {#} m/a\97{#}/',
276 'm/(*DOOF)/' => 'Unknown verb pattern \'DOOF\' {#} m/(*DOOF){#}/',
277 'm/(?&a/' => 'Sequence (?&... not terminated {#} m/(?&a{#}/',
278 'm/(?P=/' => 'Sequence ?P=... not terminated {#} m/(?P={#}/',
279 "m/(?'/" => "Sequence (?'... not terminated {#} m/(?'{#}/",
280 "m/(?</" => "Sequence (?<... not terminated {#} m/(?<{#}/",
281 'm/(?&/' => 'Sequence (?&... not terminated {#} m/(?&{#}/',
282 'm/(?(</' => 'Sequence (?(<... not terminated {#} m/(?(<{#}/',
283 "m/(?('/" => "Sequence (?('... not terminated {#} m/(?('{#}/",
284 'm/\g{/' => 'Sequence \g{... not terminated {#} m/\g{{#}/',
285 'm/\k</' => 'Sequence \k<... not terminated {#} m/\k<{#}/',
286 'm/\cß/' => "Character following \"\\c\" must be printable ASCII",
287 '/((?# This is a comment in the middle of a token)?:foo)/' => 'In \'(?...)\', the \'(\' and \'?\' must be adjacent {#} m/((?# This is a comment in the middle of a token)?{#}:foo)/',
288 '/((?# This is a comment in the middle of a token)*FAIL)/' => 'In \'(*VERB...)\', the \'(\' and \'*\' must be adjacent {#} m/((?# This is a comment in the middle of a token)*{#}FAIL)/',
289 '/((?# This is a comment in the middle of a token)*script_run:foo)/' => 'In \'(*...)\', the \'(\' and \'*\' must be adjacent {#} m/((?# This is a comment in the middle of a token)*{#}script_run:foo)/',
291 '/(*script_runfoo)/' => 'Unknown \'(*...)\' construct \'script_runfoo\' {#} m/(*script_runfoo){#}/',
292 '/(*srfoo)/' => 'Unknown \'(*...)\' construct \'srfoo\' {#} m/(*srfoo){#}/',
293 '/(*script_run)/' => '\'(*script_run\' requires a terminating \':\' {#} m/(*script_run{#})/',
294 '/(*sr)/' => '\'(*sr\' requires a terminating \':\' {#} m/(*sr{#})/',
295 '/(*pla)/' => '\'(*pla\' requires a terminating \':\' {#} m/(*pla{#})/',
296 '/(*script_run/' => 'Unterminated \'(*...\' construct {#} m/(*script_run{#}/',
297 '/(*sr/' => 'Unterminated \'(*...\' construct {#} m/(*sr{#}/',
298 '/(*script_run:foo/' => 'Unterminated \'(*...\' argument {#} m/(*script_run:foo{#}/',
299 '/(*sr:foo/' => 'Unterminated \'(*...\' argument {#} m/(*sr:foo{#}/',
300 '/(?[\ &!])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[\ &!{#}])/', # [perl #126180]
301 '/(?[\ +!])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[\ +!{#}])/', # [perl #126180]
302 '/(?[\ -!])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[\ -!{#}])/', # [perl #126180]
303 '/(?[\ ^!])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[\ ^!{#}])/', # [perl #126180]
304 '/(?[\ |!])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[\ |!{#}])/', # [perl #126180]
305 '/(?[()-!])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[(){#}-!])/', # [perl #126204]
306 '/(?[!()])/' => 'Incomplete expression within \'(?[ ])\' {#} m/(?[!(){#}])/', # [perl #126404]
307 '/\w{/' => 'Unescaped left brace in regex is illegal here {#} m/\w{{#}/',
308 '/\q{/' => 'Unescaped left brace in regex is illegal here {#} m/\q{{#}/',
309 '/\A{/' => 'Unescaped left brace in regex is illegal here {#} m/\A{{#}/',
310 '/.{, 4 }/' => 'Unescaped left brace in regex is illegal here {#} m/.{{#}, 4 }/',
311 '/[x]{, 4}/' => 'Unescaped left brace in regex is illegal here {#} m/[x]{{#}, 4}/',
312 '/\p{Latin}{,4 }/' => 'Unescaped left brace in regex is illegal here {#} m/\p{Latin}{{#},4 }/',
313 '/(?<=/' => 'Sequence (?... not terminated {#} m/(?<={#}/', # [perl #128170]
314 '/\p{vertical
\v tab}/' => 'Can\'t find Unicode property definition "vertical
\v tab" {#} m/\\p{vertical
\v tab}{#}/', # [perl #132055]
315 "/$bug133423/" => "Operand with no preceding operator {#} m/(?[(?^:(?[\\
\0]))\\{#}]
\0|2[^^]\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80\x80])R.\\670/",
316 '/[^/' => 'Unmatched [ {#} m/[{#}^/', # [perl #133767]
317 '/\p{Is_Other_Alphabetic=F}/ ' => 'Can\'t find Unicode property definition "Is_Other_Alphabetic=F" {#} m/\p{Is_Other_Alphabetic=F}{#}/',
318 '/\p{Is_Other_Alphabetic=F}/ ' => 'Can\'t find Unicode property definition "Is_Other_Alphabetic=F" {#} m/\p{Is_Other_Alphabetic=F}{#}/',
319 '/\x{100}(?(/' => 'Unknown switch condition (?(...)) {#} m/\\x{100}(?({#}/', # [perl #133896]
320 '/(?[\N{KEYCAP DIGIT NINE}/' => '\N{} here is restricted to one character {#} m/(?[\\N{U+39.FE0F.20E3{#}}/', # [perl #133988]
321 '/0000000000000000[\N{U+0.00}0000/' => 'Unmatched [ {#} m/0000000000000000[{#}\N{U+0.00}0000/', # [perl #134059]
324 # These are messages that are death under 'use re "strict"', and may or may
325 # not warn otherwise. See comment before @warning as to why some have a
326 # \x{100} in them. This array has 3 elements per construct. [0] is the regex
327 # to use; [1] is the message under no strict (empty to not warn), and [2] is
329 my @death_only_under_strict = (
331 => 'Use \x{...} for more than two hex characters {#} m/\xABC{#}/',
333 => 'Use \x{...} for more than two hex characters {#} m/[\xABC{#}]/',
335 # XXX This is a confusing error message. The G isn't ignored; it just
336 # terminates the \x. Also some messages below are missing the <-- HERE,
337 # aren't all category 'regexp'. (Hence we have to turn off 'digit'
338 # messages as well below)
339 'm/\xAG/' => 'Illegal hexadecimal digit \'G\' ignored',
340 => 'Non-hex character {#} m/\xAG{#}/',
341 'm/[\xAG]/' => 'Illegal hexadecimal digit \'G\' ignored',
342 => 'Non-hex character {#} m/[\xAG{#}]/',
343 'm/\o{789}/' => 'Non-octal character \'8\'. Resolved as "\o{7}"',
344 => 'Non-octal character {#} m/\o{78{#}9}/',
345 'm/[\o{789}]/' => 'Non-octal character \'8\'. Resolved as "\o{7}"',
346 => 'Non-octal character {#} m/[\o{78{#}9}]/',
348 => 'Empty \x{} {#} m/\x{}{#}/',
350 => 'Empty \x{} {#} m/[\x{}{#}]/',
351 'm/\x{ABCDEFG}/' => 'Illegal hexadecimal digit \'G\' ignored',
352 => 'Non-hex character {#} m/\x{ABCDEFG{#}}/',
353 'm/[\x{ABCDEFG}]/' => 'Illegal hexadecimal digit \'G\' ignored',
354 => 'Non-hex character {#} m/[\x{ABCDEFG{#}}]/',
355 "m'[\\y]\\x{100}'" => 'Unrecognized escape \y in character class passed through {#} m/[\y{#}]\x{100}/',
356 => 'Unrecognized escape \y in character class {#} m/[\y{#}]\x{100}/',
357 'm/[a-\d]\x{100}/' => 'False [] range "a-\d" {#} m/[a-\d{#}]\x{100}/',
358 => 'False [] range "a-\d" {#} m/[a-\d{#}]\x{100}/',
359 'm/[\w-x]\x{100}/' => 'False [] range "\w-" {#} m/[\w-{#}x]\x{100}/',
360 => 'False [] range "\w-" {#} m/[\w-{#}x]\x{100}/',
361 'm/[a-\pM]\x{100}/' => 'False [] range "a-\pM" {#} m/[a-\pM{#}]\x{100}/',
362 => 'False [] range "a-\pM" {#} m/[a-\pM{#}]\x{100}/',
363 'm/[\pM-x]\x{100}/' => 'False [] range "\pM-" {#} m/[\pM-{#}x]\x{100}/',
364 => 'False [] range "\pM-" {#} m/[\pM-{#}x]\x{100}/',
365 'm/[^\N{LATIN CAPITAL LETTER A WITH MACRON AND GRAVE}]/' => 'Using just the first character returned by \N{} in character class {#} m/[^\N{U+100.300}{#}]/',
366 => '\N{} here is restricted to one character {#} m/[^\N{U+100.300{#}}]/',
367 'm/[\x03-\N{LATIN CAPITAL LETTER A WITH MACRON AND GRAVE}]/' => 'Using just the first character returned by \N{} in character class {#} m/[\x03-\N{U+100.300}{#}]/',
368 => '\N{} here is restricted to one character {#} m/[\x03-\N{U+100.300{#}}]/',
369 'm/[\N{LATIN CAPITAL LETTER A WITH MACRON AND GRAVE}-\x{10FFFF}]/' => 'Using just the first character returned by \N{} in character class {#} m/[\N{U+100.300}{#}-\x{10FFFF}]/',
370 => '\N{} here is restricted to one character {#} m/[\N{U+100.300{#}}-\x{10FFFF}]/',
371 '/[\08]/' => '\'\08\' resolved to \'\o{0}8\' {#} m/[\08{#}]/',
372 => 'Need exactly 3 octal digits {#} m/[\08{#}]/',
373 '/[\018]/' => '\'\018\' resolved to \'\o{1}8\' {#} m/[\018{#}]/',
374 => 'Need exactly 3 octal digits {#} m/[\018{#}]/',
376 => 'Need exactly 3 octal digits {#} m/[\_\0]{#}/',
378 => 'Need exactly 3 octal digits {#} m/[\07]{#}/',
380 => 'Need exactly 3 octal digits {#} m/[\0005]{#}/',
381 '/[\8\9]\x{100}/' => ['Unrecognized escape \8 in character class passed through {#} m/[\8{#}\9]\x{100}/',
382 'Unrecognized escape \9 in character class passed through {#} m/[\8\9{#}]\x{100}/',
384 => 'Unrecognized escape \8 in character class {#} m/[\8{#}\9]\x{100}/',
385 '/[a-\d]\x{100}/' => 'False [] range "a-\d" {#} m/[a-\d{#}]\x{100}/',
386 => 'False [] range "a-\d" {#} m/[a-\d{#}]\x{100}/',
387 '/[\d-b]\x{100}/' => 'False [] range "\d-" {#} m/[\d-{#}b]\x{100}/',
388 => 'False [] range "\d-" {#} m/[\d-{#}b]\x{100}/',
389 '/[\s-\d]\x{100}/' => 'False [] range "\s-" {#} m/[\s-{#}\d]\x{100}/',
390 => 'False [] range "\s-" {#} m/[\s-{#}\d]\x{100}/',
391 '/[\d-\s]\x{100}/' => 'False [] range "\d-" {#} m/[\d-{#}\s]\x{100}/',
392 => 'False [] range "\d-" {#} m/[\d-{#}\s]\x{100}/',
393 '/[a-[:digit:]]\x{100}/' => 'False [] range "a-[:digit:]" {#} m/[a-[:digit:]{#}]\x{100}/',
394 => 'False [] range "a-[:digit:]" {#} m/[a-[:digit:]{#}]\x{100}/',
395 '/[[:digit:]-b]\x{100}/' => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}b]\x{100}/',
396 => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}b]\x{100}/',
397 '/[[:alpha:]-[:digit:]]\x{100}/' => 'False [] range "[:alpha:]-" {#} m/[[:alpha:]-{#}[:digit:]]\x{100}/',
398 => 'False [] range "[:alpha:]-" {#} m/[[:alpha:]-{#}[:digit:]]\x{100}/',
399 '/[[:digit:]-[:alpha:]]\x{100}/' => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}[:alpha:]]\x{100}/',
400 => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}[:alpha:]]\x{100}/',
401 '/[a\zb]\x{100}/' => 'Unrecognized escape \z in character class passed through {#} m/[a\z{#}b]\x{100}/',
402 => 'Unrecognized escape \z in character class {#} m/[a\z{#}b]\x{100}/',
404 => 'Literal vertical space in [] is illegal except under /x {#} m/[a
\f{#}b]/',
405 '/:{4,a}/' => 'Unescaped left brace in regex is passed through {#} m/:{{#}4,a}/',
406 => 'Unescaped left brace in regex is illegal here {#} m/:{{#}4,a}/',
407 '/xa{3\,4}y/' => 'Unescaped left brace in regex is passed through {#} m/xa{{#}3\,4}y/',
408 => 'Unescaped left brace in regex is illegal here {#} m/xa{{#}3\,4}y/',
409 '/\\${[^\\}]*}/' => 'Unescaped left brace in regex is passed through {#} m/\\${{#}[^\\}]*}/',
410 => 'Unescaped left brace in regex is illegal here {#} m/\\${{#}[^\\}]*}/',
411 '/.{/' => 'Unescaped left brace in regex is passed through {#} m/.{{#}/',
412 => 'Unescaped left brace in regex is illegal here {#} m/.{{#}/',
413 '/[x]{/' => 'Unescaped left brace in regex is passed through {#} m/[x]{{#}/',
414 => 'Unescaped left brace in regex is illegal here {#} m/[x]{{#}/',
415 '/\p{Latin}{/' => 'Unescaped left brace in regex is passed through {#} m/\p{Latin}{{#}/',
416 => 'Unescaped left brace in regex is illegal here {#} m/\p{Latin}{{#}/',
418 => "Empty \\x {#} m/\\x{100}\\x{#}/",
421 # These need the character 'ネ' as a marker for mark_as_utf8()
422 my @death_utf8 = mark_as_utf8(
423 '/ネ(?<= .*)/' => 'Lookbehind longer than 255 not implemented in regex m/ネ(?<= .*)/',
425 '/(?<= ネ{1000})/' => 'Lookbehind longer than 255 not implemented in regex m/(?<= ネ{1000})/',
427 '/ネ(?ネ)ネ/' => 'Sequence (?ネ...) not recognized {#} m/ネ(?ネ{#})ネ/',
429 '/ネ(?(1ネ))ネ/' => 'Switch condition not recognized {#} m/ネ(?(1ネ{#}))ネ/',
431 '/(?(1)ネ|y|ヌ)/' => 'Switch (?(condition)... contains too many branches {#} m/(?(1)ネ|y|{#}ヌ)/',
433 '/(?(ネ)y|ネ)/' => 'Unknown switch condition (?(...)) {#} m/(?(ネ{#})y|ネ)/',
435 '/ネ(?/' => 'Sequence (? incomplete {#} m/ネ(?{#}/',
437 '/ネ(?;ネ/' => 'Sequence (?;...) not recognized {#} m/ネ(?;{#}ネ/',
438 '/ネ(?<;ネ/' => 'Group name must start with a non-digit word character {#} m/ネ(?<;{#}ネ/',
439 '/ネ(?\ixネ/' => 'Sequence (?\...) not recognized {#} m/ネ(?\{#}ixネ/',
440 '/ネ(?^lu:ネ)/' => 'Regexp modifiers "l" and "u" are mutually exclusive {#} m/ネ(?^lu{#}:ネ)/',
441 '/ネ(?lil:ネ)/' => 'Regexp modifier "l" may not appear twice {#} m/ネ(?lil{#}:ネ)/',
442 '/ネ(?aaia:ネ)/' => 'Regexp modifier "a" may appear a maximum of twice {#} m/ネ(?aaia{#}:ネ)/',
443 '/ネ(?i-l:ネ)/' => 'Regexp modifier "l" may not appear after the "-" {#} m/ネ(?i-l{#}:ネ)/',
445 '/ネ((ネ)/' => 'Unmatched ( {#} m/ネ({#}(ネ)/',
447 "/ネ{$inf_p1}ネ/" => "Quantifier in {,} bigger than $inf_m1 {#} m/ネ{{#}$inf_p1}ネ/",
450 '/ネ**ネ/' => 'Nested quantifiers {#} m/ネ**{#}ネ/',
452 '/ネ[ネ/' => 'Unmatched [ {#} m/ネ[{#}ネ/',
454 '/*ネ/', => 'Quantifier follows nothing {#} m/*{#}ネ/',
456 '/ネ\p{ネ/' => 'Missing right brace on \p{} {#} m/ネ\p{{#}ネ/',
458 '/(ネ)\2ネ/' => 'Reference to nonexistent group {#} m/(ネ)\2{#}ネ/',
460 '/\g{ネ/; #no latin1' => 'Sequence \g{... not terminated {#} m/\g{ネ{#}/',
462 'my $m = "ネ\\\"; $m =~ $m', => 'Trailing \ in regex m/ネ\/',
464 '/\x{ネ/' => 'Missing right brace on \x{} {#} m/\x{{#}ネ/',
465 '/ネ[\x{ネ]ネ/' => 'Missing right brace on \x{} {#} m/ネ[\x{{#}ネ]ネ/',
466 '/ネ[\x{ネ]/' => 'Missing right brace on \x{} {#} m/ネ[\x{{#}ネ]/',
468 '/ネ\o{ネ/' => 'Missing right brace on \o{ {#} m/ネ\o{{#}ネ/',
470 '/[ネ-a]ネ/' => 'Invalid [] range "ネ-a" {#} m/[ネ-a{#}]ネ/',
472 '/ネ\p{}ネ/' => 'Empty \p{} {#} m/ネ\p{{#}}ネ/',
474 '/ネ(?[[[:ネ]]])ネ/' => "Unexpected ']' with no following ')' in (?[... {#} m/ネ(?[[[:ネ]]{#}])ネ/",
475 '/ネ(?[[[:ネ: ])ネ/' => "Syntax error in (?[...]) {#} m/ネ(?[[[:ネ: ])ネ{#}/",
476 '/ネ(?[[[::]]])ネ/' => "Unexpected ']' with no following ')' in (?[... {#} m/ネ(?[[[::]]{#}])ネ/",
477 '/ネ(?[[[:ネ:]]])ネ/' => "Unexpected ']' with no following ')' in (?[... {#} m/ネ(?[[[:ネ:]]{#}])ネ/",
478 '/ネ(?[ネ])ネ/' => 'Unexpected character {#} m/ネ(?[ネ{#}])ネ/',
479 '/ネ(?[ + [ネ] ])/' => 'Unexpected binary operator \'+\' with no preceding operand {#} m/ネ(?[ +{#} [ネ] ])/',
480 '/ネ(?[ \cK - ( + [ネ] ) ])/' => 'Unexpected binary operator \'+\' with no preceding operand {#} m/ネ(?[ \cK - ( +{#} [ネ] ) ])/',
481 '/ネ(?[ \cK ( [ネ] ) ])/' => 'Unexpected \'(\' with no preceding operator {#} m/ネ(?[ \cK ({#} [ネ] ) ])/',
482 '/ネ(?[ \cK [ネ] ])ネ/' => 'Operand with no preceding operator {#} m/ネ(?[ \cK [ネ{#}] ])ネ/',
483 '/ネ(?[ \0004 ])ネ/' => 'Need exactly 3 octal digits {#} m/ネ(?[ \0004 {#}])ネ/',
484 '/(?[ \o{ネ} ])ネ/' => 'Non-octal character {#} m/(?[ \o{ネ{#}} ])ネ/',
485 '/ネ(?[ \o{} ])ネ/' => 'Empty \o{} {#} m/ネ(?[ \o{}{#} ])ネ/',
486 '/(?[ \x{ネ} ])ネ/' => 'Non-hex character {#} m/(?[ \x{ネ{#}} ])ネ/',
487 '/(?[ \p{ネ} ])/' => 'Can\'t find Unicode property definition "ネ" {#} m/(?[ \p{ネ}{#} ])/',
488 '/(?[ \p{ ネ = bar } ])/' => 'Can\'t find Unicode property definition "ネ = bar" {#} m/(?[ \p{ ネ = bar }{#} ])/',
489 '/ネ(?[ \t ]/' => "Unexpected ']' with no following ')' in (?[... {#} m/ネ(?[ \\t ]{#}/",
490 '/(?[ \t + \e # ネ This was supposed to be a comment ])/' =>
491 "Syntax error in (?[...]) {#} m/(?[ \\t + \\e # ネ This was supposed to be a comment ]){#}/",
492 'm/(*ネ)ネ/' => q<Unknown '(*...)' construct 'ネ' {#} m/(*ネ){#}ネ/>,
493 '/\cネ/' => "Character following \"\\c\" must be printable ASCII",
494 '/\b{ネ}/' => "'ネ' is an unknown bound type {#} m/\\b{ネ{#}}/",
495 '/\B{ネ}/' => "'ネ' is an unknown bound type {#} m/\\B{ネ{#}}/",
497 push @death, @death_utf8;
499 my @death_utf8_only_under_strict = (
500 "m'ネ[\\y]ネ'" => 'Unrecognized escape \y in character class passed through {#} m/ネ[\y{#}]ネ/',
501 => 'Unrecognized escape \y in character class {#} m/ネ[\y{#}]ネ/',
502 'm/ネ[ネ-\d]ネ/' => 'False [] range "ネ-\d" {#} m/ネ[ネ-\d{#}]ネ/',
503 => 'False [] range "ネ-\d" {#} m/ネ[ネ-\d{#}]ネ/',
504 'm/ネ[\w-ネ]ネ/' => 'False [] range "\w-" {#} m/ネ[\w-{#}ネ]ネ/',
505 => 'False [] range "\w-" {#} m/ネ[\w-{#}ネ]ネ/',
506 'm/ネ[ネ-\pM]ネ/' => 'False [] range "ネ-\pM" {#} m/ネ[ネ-\pM{#}]ネ/',
507 => 'False [] range "ネ-\pM" {#} m/ネ[ネ-\pM{#}]ネ/',
508 '/ネ[ネ-[:digit:]]ネ/' => 'False [] range "ネ-[:digit:]" {#} m/ネ[ネ-[:digit:]{#}]ネ/',
509 => 'False [] range "ネ-[:digit:]" {#} m/ネ[ネ-[:digit:]{#}]ネ/',
510 '/ネ[\d-\s]ネ/' => 'False [] range "\d-" {#} m/ネ[\d-{#}\s]ネ/',
511 => 'False [] range "\d-" {#} m/ネ[\d-{#}\s]ネ/',
512 '/ネ[a\zb]ネ/' => 'Unrecognized escape \z in character class passed through {#} m/ネ[a\z{#}b]ネ/',
513 => 'Unrecognized escape \z in character class {#} m/ネ[a\z{#}b]ネ/',
515 # Tests involving a user-defined charnames translator are in pat_advanced.t
517 # In the following arrays of warnings, the value can be an array of things to
518 # expect. If the empty string, it means no warning should be raised.
521 # Key-value pairs of code/error of code that should have non-fatal regexp
522 # warnings. Most currently have \x{100} appended to them to force them to be
523 # upgraded to UTF-8, and the first pass restarted. Previously this would
524 # cause some warnings to be output twice. This tests that that behavior has
528 'm/\b*\x{100}/' => '\b* matches null string many times {#} m/\b*{#}\x{100}/',
529 '/\b{g}/a' => "Using /u for '\\b{g}' instead of /a {#} m/\\b{g}{#}/",
530 '/\B{gcb}/a' => "Using /u for '\\B{gcb}' instead of /a {#} m/\\B{gcb}{#}/",
531 'm/[:blank:]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:blank:]{#}\x{100}/',
532 'm/[[:cntrl:]][:^ascii:]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[[:cntrl:]][:^ascii:]{#}\x{100}/',
533 'm/[[:ascii]]\x{100}/' => "Assuming NOT a POSIX class since there is no terminating ':' {#} m/[[:ascii{#}]]\\x{100}/",
534 'm/(?[[:word]])\x{100}/' => "Assuming NOT a POSIX class since there is no terminating ':' {#} m/(?[[:word{#}]])\\x{100}/",
535 "m'\\y\\x{100}'" => 'Unrecognized escape \y passed through {#} m/\y{#}\x{100}/',
536 '/x{3,1}/' => 'Quantifier {n,m} with n > m can\'t match {#} m/x{3,1}{#}/',
537 '/\08/' => '\'\08\' resolved to \'\o{0}8\' {#} m/\08{#}/',
538 '/\018/' => '\'\018\' resolved to \'\o{1}8\' {#} m/\018{#}/',
539 '/(?=a)*/' => '(?=a)* matches null string many times {#} m/(?=a)*{#}/',
540 'my $x = \'\m\'; qr/a$x/' => 'Unrecognized escape \m passed through {#} m/a\m{#}/',
541 '/\q/' => 'Unrecognized escape \q passed through {#} m/\q{#}/',
542 '/\q\p{Any}/' => 'Unrecognized escape \q passed through {#} m/\q{#}\p{Any}/',
544 # These two tests do not include the marker, because regcomp.c no
545 # longer knows where it goes by the time this warning is emitted.
546 # See [perl #122680] regcomp warning gives wrong position of
548 '/(?=a){1,3}\x{100}/' => 'Quantifier unexpected on zero-length expression in regex m/(?=a){1,3}\x{100}/',
549 '/(a|b)(?=a){3}\x{100}/' => 'Quantifier unexpected on zero-length expression in regex m/(a|b)(?=a){3}\x{100}/',
553 '/[:alpha:]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:alpha:]{#}\x{100}/',
554 '/[:zog:]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes (but this one isn\'t fully valid) {#} m/[:zog:]{#}\x{100}/',
555 '/[.zog.]\x{100}/' => 'POSIX syntax [. .] belongs inside character classes (but this one isn\'t implemented) {#} m/[.zog.]{#}\x{100}/',
557 '/(?c)\x{100}/' => 'Useless (?c) - use /gc modifier {#} m/(?c{#})\x{100}/',
558 '/(?-c)\x{100}/' => 'Useless (?-c) - don\'t use /gc modifier {#} m/(?-c{#})\x{100}/',
559 '/(?g)\x{100}/' => 'Useless (?g) - use /g modifier {#} m/(?g{#})\x{100}/',
560 '/(?-g)\x{100}/' => 'Useless (?-g) - don\'t use /g modifier {#} m/(?-g{#})\x{100}/',
561 '/(?o)\x{100}/' => 'Useless (?o) - use /o modifier {#} m/(?o{#})\x{100}/',
562 '/(?-o)\x{100}/' => 'Useless (?-o) - don\'t use /o modifier {#} m/(?-o{#})\x{100}/',
563 '/(?g-o)\x{100}/' => [ 'Useless (?g) - use /g modifier {#} m/(?g{#}-o)\x{100}/',
564 'Useless (?-o) - don\'t use /o modifier {#} m/(?g-o{#})\x{100}/',
566 '/(?g-c)\x{100}/' => [ 'Useless (?g) - use /g modifier {#} m/(?g{#}-c)\x{100}/',
567 'Useless (?-c) - don\'t use /gc modifier {#} m/(?g-c{#})\x{100}/',
569 # (?c) means (?g) error won't be thrown
570 '/(?o-cg)\x{100}/' => [ 'Useless (?o) - use /o modifier {#} m/(?o{#}-cg)\x{100}/',
571 'Useless (?-c) - don\'t use /gc modifier {#} m/(?o-c{#}g)\x{100}/',
573 '/(?ogc)\x{100}/' => [ 'Useless (?o) - use /o modifier {#} m/(?o{#}gc)\x{100}/',
574 'Useless (?g) - use /g modifier {#} m/(?og{#}c)\x{100}/',
575 'Useless (?c) - use /gc modifier {#} m/(?ogc{#})\x{100}/',
577 '/a{1,1}?\x{100}/' => 'Useless use of greediness modifier \'?\' {#} m/a{1,1}?{#}\x{100}/',
578 "/(?[ [ % - % ] ])/" => "",
579 "/(?[ [ : - \\x$colon_hex ] ])\\x{100}/" => "\": - \\x$colon_hex \" is more clearly written simply as \":\" {#} m/(?[ [ : - \\x$colon_hex {#}] ])\\x{100}/",
580 "/(?[ [ \\x$colon_hex - : ] ])\\x{100}/" => "\"\\x$colon_hex\ - : \" is more clearly written simply as \":\" {#} m/(?[ [ \\x$colon_hex - : {#}] ])\\x{100}/",
581 "/(?[ [ \\t - \\x$tab_hex ] ])\\x{100}/" => "\"\\t - \\x$tab_hex \" is more clearly written simply as \"\\t\" {#} m/(?[ [ \\t - \\x$tab_hex {#}] ])\\x{100}/",
582 "/(?[ [ \\x$tab_hex - \\t ] ])\\x{100}/" => "\"\\x$tab_hex\ - \\t \" is more clearly written simply as \"\\t\" {#} m/(?[ [ \\x$tab_hex - \\t {#}] ])\\x{100}/",
583 "/(?[ [ $B_hex - C ] ])/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/(?[ [ $B_hex - C {#}] ])/",
584 "/(?[ [ A - $B_hex ] ])/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/(?[ [ A - $B_hex {#}] ])/",
585 "/(?[ [ $low_mixed_alpha - $high_mixed_alpha ] ])/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/(?[ [ $low_mixed_alpha - $high_mixed_alpha {#}] ])/",
586 "/(?[ [ $low_mixed_digit - $high_mixed_digit ] ])/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/(?[ [ $low_mixed_digit - $high_mixed_digit {#}] ])/",
589 '/[:blank]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes (but this one isn\'t fully valid) {#} m/[:blank{#}]\x{100}/',
590 '/[[:digit]]\x{100}/' => 'Assuming NOT a POSIX class since there is no terminating \':\' {#} m/[[:digit{#}]]\x{100}/', # [perl # 8904]
591 '/[[:digit:foo]\x{100}/' => 'Assuming NOT a POSIX class since there is no terminating \']\' {#} m/[[:digit:{#}foo]\x{100}/',
592 '/[[:di#it:foo]\x{100}/x' => 'Assuming NOT a POSIX class since there is no terminating \']\' {#} m/[[:di#it:{#}foo]\x{100}/',
593 '/[[:dgit]]\x{100}/' => 'Assuming NOT a POSIX class since there is no terminating \':\' {#} m/[[:dgit{#}]]\x{100}/',
594 '/[[:dgit:foo]\x{100}/' => 'Assuming NOT a POSIX class since there is no terminating \']\' {#} m/[[:dgit:{#}foo]\x{100}/',
595 '/[[:dgt]]\x{100}/' => "", # Far enough away from a real class to not be recognized as one
596 '/[[:dgt:foo]\x{100}/' => "",
597 '/[[:DIGIT]]\x{100}/' => [ 'Assuming NOT a POSIX class since the name must be all lowercase letters {#} m/[[:DIGIT{#}]]\x{100}/',
598 'Assuming NOT a POSIX class since there is no terminating \':\' {#} m/[[:DIGIT{#}]]\x{100}/',
600 '/[[digit]\x{100}/' => [ 'Assuming NOT a POSIX class since there must be a starting \':\' {#} m/[[{#}digit]\x{100}/',
601 'Assuming NOT a POSIX class since there is no terminating \':\' {#} m/[[digit{#}]\x{100}/',
603 '/[[alpha]]\x{100}/' => [ 'Assuming NOT a POSIX class since there must be a starting \':\' {#} m/[[{#}alpha]]\x{100}/',
604 'Assuming NOT a POSIX class since there is no terminating \':\' {#} m/[[alpha{#}]]\x{100}/',
606 '/[[^word]\x{100}/' => [ 'Assuming NOT a POSIX class since the \'^\' must come after the colon {#} m/[[^{#}word]\x{100}/',
607 'Assuming NOT a POSIX class since there must be a starting \':\' {#} m/[[^{#}word]\x{100}/',
608 'Assuming NOT a POSIX class since there is no terminating \':\' {#} m/[[^word{#}]\x{100}/',
610 '/[[ ^ : x d i g i t : ] ]\x{100}/' => [ 'Assuming NOT a POSIX class since no blanks are allowed in one {#} m/[[ {#}^ : x d i g i t : ] ]\x{100}/',
611 'Assuming NOT a POSIX class since the \'^\' must come after the colon {#} m/[[ ^{#} : x d i g i t : ] ]\x{100}/',
612 'Assuming NOT a POSIX class since no blanks are allowed in one {#} m/[[ ^ {#}: x d i g i t : ] ]\x{100}/',
613 'Assuming NOT a POSIX class since no blanks are allowed in one {#} m/[[ ^ : {#}x d i g i t : ] ]\x{100}/',
614 'Assuming NOT a POSIX class since no blanks are allowed in one {#} m/[[ ^ : x d i g i t : ]{#} ]\x{100}/',
615 $only_strict_marker . 'Unescaped literal \']\' {#} m/[[ ^ : x d i g i t : ] ]{#}\x{100}/',
617 '/[foo:lower:]]\x{100}/' => 'Assuming NOT a POSIX class since it doesn\'t start with a \'[\' {#} m/[foo{#}:lower:]]\x{100}/',
618 '/[[;upper;]]\x{100}/' => [ 'Assuming NOT a POSIX class since a semi-colon was found instead of a colon {#} m/[[;{#}upper;]]\x{100}/',
619 'Assuming NOT a POSIX class since a semi-colon was found instead of a colon {#} m/[[;upper;]{#}]\x{100}/',
621 '/[foo;punct;]]\x{100}/' => [ 'Assuming NOT a POSIX class since it doesn\'t start with a \'[\' {#} m/[foo{#};punct;]]\x{100}/',
622 'Assuming NOT a POSIX class since a semi-colon was found instead of a colon {#} m/[foo;{#}punct;]]\x{100}/',
623 'Assuming NOT a POSIX class since a semi-colon was found instead of a colon {#} m/[foo;punct;]{#}]\x{100}/',
625 '/[][[:alpha:]]/' => "", # [perl #127581]
626 '/[][[:alpha:]\\@\\\\^_?]/' => "", # [perl #131522]
627 '/(?[[:w:]])/' => "",
628 '/([.].*)[.]/' => "", # [perl #127582]
629 '/[.].*[.]/' => "", # [perl #127604]
631 '/(?xmsixp:abc)/' => "",
632 '/(?xmsixp)abc/' => "",
633 '/(?xxxx:abc)/' => "",
635 ); # See comments before this for why '\x{100}' is generally needed
637 # These need the character 'ネ' as a marker for mark_as_utf8()
638 my @warnings_utf8 = mark_as_utf8(
639 'm/ネ\b*ネ/' => '\b* matches null string many times {#} m/ネ\b*{#}ネ/',
640 '/(?=ネ)*/' => '(?=ネ)* matches null string many times {#} m/(?=ネ)*{#}/',
641 'm/ネ[:foo:]ネ/' => 'POSIX syntax [: :] belongs inside character classes (but this one isn\'t fully valid) {#} m/ネ[:foo:]{#}ネ/',
642 '/ネ(?c)ネ/' => 'Useless (?c) - use /gc modifier {#} m/ネ(?c{#})ネ/',
643 '/utf8 ネ (?ogc) ネ/' => [
644 'Useless (?o) - use /o modifier {#} m/utf8 ネ (?o{#}gc) ネ/',
645 'Useless (?g) - use /g modifier {#} m/utf8 ネ (?og{#}c) ネ/',
646 'Useless (?c) - use /gc modifier {#} m/utf8 ネ (?ogc{#}) ネ/',
649 '/ネ(?[[:ネ:]])ネ/' => "",
653 push @warning, @warnings_utf8;
655 my @warning_only_under_strict = (
656 '/[\N{U+00}-\x01]\x{100}/' => 'Both or neither range ends should be Unicode {#} m/[\N{U+00}-\x01{#}]\x{100}/',
657 '/[\x00-\N{SOH}]\x{100}/' => 'Both or neither range ends should be Unicode {#} m/[\x00-\N{U+01}{#}]\x{100}/',
658 '/[\N{DEL}-\o{377}]\x{100}/' => 'Both or neither range ends should be Unicode {#} m/[\N{U+7F}-\o{377}{#}]\x{100}/',
659 '/[\o{0}-\N{U+01}]\x{100}/' => 'Both or neither range ends should be Unicode {#} m/[\o{0}-\N{U+01}{#}]\x{100}/',
660 '/[\000-\N{U+01}]\x{100}/' => 'Both or neither range ends should be Unicode {#} m/[\000-\N{U+01}{#}]\x{100}/',
661 '/[\N{DEL}-\377]\x{100}/' => 'Both or neither range ends should be Unicode {#} m/[\N{U+7F}-\377{#}]\x{100}/',
662 '/[\N{U+00}-A]\x{100}/' => 'Ranges of ASCII printables should be some subset of "0-9", "A-Z", or "a-z" {#} m/[\N{U+00}-A{#}]\x{100}/',
663 '/[a-\N{U+FF}]\x{100}/' => 'Ranges of ASCII printables should be some subset of "0-9", "A-Z", or "a-z" {#} m/[a-\N{U+FF}{#}]\x{100}/',
664 '/[\N{U+00}-\a]\x{100}/' => "",
665 '/[\a-\N{U+FF}]\x{100}/' => "",
666 '/[\N{U+FF}-\x{100}]/' => 'Both or neither range ends should be Unicode {#} m/[\N{U+FF}-\x{100}{#}]/',
667 '/[\N{U+100}-\x{101}]/' => "",
669 "/[:-\\x$colon_hex]\\x{100}/" => "\":-\\x$colon_hex\" is more clearly written simply as \":\" {#} m/[:-\\x$colon_hex\{#}]\\x{100}/",
670 "/[\\x$colon_hex-:]\\x{100}/" => "\"\\x$colon_hex-:\" is more clearly written simply as \":\" {#} m/[\\x$colon_hex\-:{#}]\\x{100}/",
671 "/[\\t-\\x$tab_hex]\\x{100}/" => "\"\\t-\\x$tab_hex\" is more clearly written simply as \"\\t\" {#} m/[\\t-\\x$tab_hex\{#}]\\x{100}/",
672 "/[\\x$tab_hex-\\t]\\x{100}/" => "\"\\x$tab_hex-\\t\" is more clearly written simply as \"\\t\" {#} m/[\\x$tab_hex\-\\t{#}]\\x{100}/",
673 "/[$B_hex-C]/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/[$B_hex-C{#}]/",
674 "/[A-$B_hex]/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/[A-$B_hex\{#}]/",
675 "/[$low_mixed_alpha-$high_mixed_alpha]/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/[$low_mixed_alpha-$high_mixed_alpha\{#}]/",
676 "/[$low_mixed_digit-$high_mixed_digit]/" => "Ranges of ASCII printables should be some subset of \"0-9\", \"A-Z\", or \"a-z\" {#} m/[$low_mixed_digit-$high_mixed_digit\{#}]/",
677 '/\b<GCB}/' => 'Unescaped literal \'}\' {#} m/\b<GCB}{#}/',
678 '/[ ]def]/' => 'Unescaped literal \']\' {#} m/[ ]def]{#}/',
679 '/(?)/' => 'Empty (?) without any modifiers {#} m/(?){#}/', # [perl #132851]
682 my @warning_utf8_only_under_strict = mark_as_utf8(
683 '/ネ[᪉-᪐]/; #no latin1' => "Ranges of digits should be from the same group of 10 {#} m/ネ[᪉-᪐{#}]/",
684 '/ネ(?[ [ ᪉ - ᪐ ] ])/; #no latin1' => "Ranges of digits should be from the same group of 10 {#} m/ネ(?[ [ ᪉ - ᪐ {#}] ])/",
685 '/ネ[᧙-᧚]/; #no latin1' => "Ranges of digits should be from the same group of 10 {#} m/ネ[᧙-᧚{#}]/",
686 '/ネ(?[ [ ᧙ - ᧚ ] ])/; #no latin1' => "Ranges of digits should be from the same group of 10 {#} m/ネ(?[ [ ᧙ - ᧚ {#}] ])/",
687 '/ネ(?[ [ 𝟘 - 𝟡 ] ])/; #no latin1' => "",
688 '/ネ(?[ [ 𝟧 - 𝟱 ] ])/; #no latin1' => "Ranges of digits should be from the same group of 10 {#} m/ネ(?[ [ 𝟧 - 𝟱 {#}] ])/",
689 '/ネ(?[ [ 𝟧 - 𝟰 ] ])/; #no latin1' => "Ranges of digits should be from the same group of 10 {#} m/ネ(?[ [ 𝟧 - 𝟰 {#}] ])/",
692 push @warning_only_under_strict, @warning_utf8_only_under_strict;
694 my @experimental_regex_sets = (
695 '/(?[ \t ])/' => 'The regex_sets feature is experimental {#} m/(?[{#} \t ])/',
696 'use utf8; /utf8 ネ (?[ [\tネ] ])/' => do { use utf8; 'The regex_sets feature is experimental {#} m/utf8 ネ (?[{#} [\tネ] ])/' },
697 '/noutf8 ネ (?[ [\tネ] ])/' => 'The regex_sets feature is experimental {#} m/noutf8 ネ (?[{#} [\tネ] ])/',
704 '/foo({bar)/' => 'Unescaped left brace in regex is deprecated here (and will be fatal in Perl 5.32), passed through {#} m/foo({{#}bar)/',
705 '/foo(:?{bar)/' => "",
710 for my $strict ("", "use re 'strict';") {
712 # First time just use @death; but under strict we add the things that fail
713 # there. Doing it this way makes sure that 'strict' doesnt change the
714 # things that are already fatal when not under strict.
716 for (my $i = 0; $i < @death_only_under_strict; $i += 3) {
717 push @death, $death_only_under_strict[$i], # The regex
718 $death_only_under_strict[$i+2]; # The fatal msg
720 for (my $i = 0; $i < @death_utf8_only_under_strict; $i += 3) {
722 # Same with the utf8 versions
723 push @death, mark_as_utf8($death_utf8_only_under_strict[$i],
724 $death_utf8_only_under_strict[$i+2]);
727 for (my $i = 0; $i < @death; $i += 2) {
728 my $regex = $death[$i] =~ s/ default_ (on | off) //rx;
729 my $expect = fixup_expect($death[$i+1], $strict);
731 fail("$0: Internal error: '$death[$i]' should have an error message");
734 no warnings 'experimental::regex_sets';
735 no warnings 'experimental::re_strict';
739 my $eval_string = "$strict $regex";
741 eval "$eval_string; \$meaning_of_life = 42";
742 ok (! defined $meaning_of_life, "$eval_string died");
744 if ($error =~ qr/\Q$expect/) {
745 ok(1, "... and gave expected message");
748 diag("Have: " . _q(add_markers($error)));
749 diag("Want: " . _q($death[$i+1]));
751 }, undef, "... and no other warnings");
756 for my $strict ("", "no warnings 'experimental::re_strict'; use re 'strict';") {
757 my @warning_tests = @warning;
759 # Build the tests for @warning. Use the strict/non-strict versions
762 push @warning_tests, @warning_only_under_strict;
765 for (my $i = 0; $i < @warning_only_under_strict; $i += 2) {
767 # (?[ ]) are always under strict
768 if ($warning_only_under_strict[$i] =~ /\Q(?[/) {
769 push @warning_tests, $warning_only_under_strict[$i], # The regex
770 $warning_only_under_strict[$i+1];
773 push @warning_tests, $warning_only_under_strict[$i], # The regex
774 ""; # No warning because not strict
777 for (my $i = 0; $i < @death_only_under_strict; $i += 3) {
778 push @warning_tests, $death_only_under_strict[$i], # The regex
779 $death_only_under_strict[$i+1]; # The warning
781 for (my $i = 0; $i < @death_utf8_only_under_strict; $i += 3) {
782 push @warning_tests, mark_as_utf8($death_utf8_only_under_strict[$i],
783 $death_utf8_only_under_strict[$i+1]);
787 foreach my $ref (\@warning_tests,
788 \@experimental_regex_sets,
792 my $turn_off_warnings = "";
794 if ($ref == \@warning_tests) {
795 $warning_type = 'regexp, digit';
796 $turn_off_warnings = "no warnings 'experimental::regex_sets';";
797 $default_on = $strict;
799 elsif ($ref == \@deprecated) {
800 $warning_type = 'regexp, deprecated';
803 elsif ($ref == \@experimental_regex_sets) {
804 $warning_type = 'experimental::regex_sets';
808 fail("$0: Internal error: Unexpected loop variable");
811 for (my $i = 0; $i < @$ref; $i += 2) {
812 my $this_default_on = $default_on;
813 my $regex = $ref->[$i];
814 if ($regex =~ s/ default_ (on | off) //x) {
815 $this_default_on = $1 eq 'on';
817 my @expect = fixup_expect($ref->[$i+1], $strict);
819 # A length-1 array with an empty warning means no warning gets
821 undef @expect if @expect == 1 && $expect[0] eq "";
825 #use feature 'unicode_eval';
826 #print STDERR __LINE__, ": ", "eval '$strict no warnings; $regex'", "\n";
827 eval "$strict no warnings; $regex";
829 if (is($@, "", "$strict $regex did not die")) {
830 my @got = capture_warnings(sub {
832 eval "$strict $turn_off_warnings $regex" });
834 if (! is(scalar @got, scalar @expect,
835 "... and gave expected number ($count) of warnings"))
837 if (@got < @expect) {
839 note "Expected warnings not gotten:\n\t" . join "\n\t",
840 @expect[$count .. $#expect];
843 note "Unexpected warnings gotten:\n\t" . join("\n\t",
844 @got[$count .. $#got]);
847 foreach my $i (0 .. $count - 1) {
848 if (! like($got[$i], qr/\Q$expect[$i]/,
849 "... and gave expected warning"))
853 diag("GOT\n'$got[$i]'\nEXPECT\n'$expect[$i]'");
856 ok (0 == capture_warnings(sub {
858 eval "$strict no warnings '$warning_type'; $regex;" }
860 "... and turning off '$warning_type' warnings suppressed it");
862 # Test that whether the warning is on by default is
863 # correct. This test relies on the fact that we
864 # are outside the scope of any ‘use warnings’.
866 my @warns = capture_warnings(sub { $_ = "x";
867 eval "$strict $regex" });
868 # Warning should be on as well if is testing
869 # '(?[...])' which turns on strict
870 if ($this_default_on || grep { $_ =~ /\Q(?[/ } @expect ) {
871 ok @warns > 0, "... and the warning is on by default";
874 ok @warns == 0, "... and the warning is off by default";