BEGIN {
chdir 't' if -d 't';
- @INC = '../lib';
+ @INC = qw '../lib ../ext/re';
require './test.pl';
+ skip_all_without_unicode_tables();
eval 'require Config'; # assume defaults if this fails
}
use strict;
+use open qw(:utf8 :std);
##
## If the markers used are changed (search for "MARKER1" in regcomp.c),
return wantarray ? @expect : join "", @expect;
}
+## Because we don't "use utf8" in this file, we need to do some extra legwork
+## for the utf8 tests: Append 'use utf8' to the pattern, and mark the strings
+## to check against as UTF-8
+##
+## This also creates a second variant of the tests to check if the
+## latin1 error messages are working correctly.
+my $l1 = "\x{ef}";
+my $utf8 = "\x{30cd}";
+utf8::encode($utf8);
+
+sub mark_as_utf8 {
+ my @ret;
+ while ( my ($pat, $msg) = splice(@_, 0, 2) ) {
+ my $l1_pat = $pat =~ s/$utf8/$l1/gr;
+ my $l1_msg;
+ $pat = "use utf8; $pat";
+
+ if (ref $msg) {
+ $l1_msg = [ map { s/$utf8/$l1/gr } @$msg ];
+ @$msg = map { my $c = $_; utf8::decode($c); $c } @$msg;
+ }
+ else {
+ $l1_msg = $msg =~ s/$utf8/$l1/gr;
+ utf8::decode($msg);
+ }
+ push @ret, $pat => $msg;
+ push @ret, $l1_pat => $l1_msg unless $l1_pat =~ /#no latin1/;
+ }
+ return @ret;
+}
+
my $inf_m1 = ($Config::Config{reg_infty} || 32767) - 1;
my $inf_p1 = $inf_m1 + 2;
'/(?{ 1/' => 'Missing right curly or square bracket',
'/(?(1x))/' => 'Switch condition not recognized {#} m/(?(1x{#}))/',
+ '/(?(1x(?#)))/'=> 'Switch condition not recognized {#} m/(?(1x{#}(?#)))/',
'/(?(1)x|y|z)/' => 'Switch (?(condition)... contains too many branches {#} m/(?(1)x|y|{#}z)/',
- '/(?(x)y|x)/' => 'Unknown switch condition (?(x) {#} m/(?({#}x)y|x)/',
+ '/(?(x)y|x)/' => 'Unknown switch condition (?(...)) {#} m/(?(x{#})y|x)/',
+ '/(?(??{}))/' => 'Unknown switch condition (?(...)) {#} m/(?(?{#}?{}))/',
+ '/(?(?[]))/' => 'Unknown switch condition (?(...)) {#} m/(?(?{#}[]))/',
'/(?/' => 'Sequence (? incomplete {#} m/(?{#}/',
'/(?lil:foo)/' => 'Regexp modifier "l" may not appear twice {#} m/(?lil{#}:foo)/',
'/(?aaia:foo)/' => 'Regexp modifier "a" may appear a maximum of twice {#} m/(?aaia{#}:foo)/',
'/(?i-l:foo)/' => 'Regexp modifier "l" may not appear after the "-" {#} m/(?i-l{#}:foo)/',
+'/a\b{cde/' => 'Use "\b\{" instead of "\b{" {#} m/a\{#}b{cde/',
+'/a\B{cde/' => 'Use "\B\{" instead of "\B{" {#} m/a\{#}B{cde/',
'/((x)/' => 'Unmatched ( {#} m/({#}(x)/',
'/(x)\2/' => 'Reference to nonexistent group {#} m/(x)\2{#}/',
+ '/\g/' => 'Unterminated \g... pattern {#} m/\g{#}/',
'/\g{1/' => 'Unterminated \g{...} pattern {#} m/\g{1{#}/',
'my $m = "\\\"; $m =~ $m', => 'Trailing \ in regex m/\/',
'm/[\o]/' => 'Missing braces on \o{} {#} m/[\o{#}]/',
'm/[\o{}]/' => 'Number with no digits {#} m/[\o{}{#}]/',
'm/(?^-i:foo)/' => 'Sequence (?^-...) not recognized {#} m/(?^-{#}i:foo)/',
+ 'm/\87/' => 'Reference to nonexistent group {#} m/\87{#}/',
+ 'm/a\87/' => 'Reference to nonexistent group {#} m/a\87{#}/',
+ 'm/a\97/' => 'Reference to nonexistent group {#} m/a\97{#}/',
+ 'm/(*DOOF)/' => 'Unknown verb pattern \'DOOF\' {#} m/(*DOOF){#}/',
+ 'm/(?&a/' => 'Sequence (?&... not terminated {#} m/(?&a{#}/',
+ 'm/(?P=/' => 'Sequence ?P=... not terminated {#} m/(?P={#}/',
+ "m/(?'/" => "Sequence (?'... not terminated {#} m/(?'{#}/",
+ "m/(?</" => "Sequence (?<... not terminated {#} m/(?<{#}/",
+ 'm/(?&/' => 'Sequence (?&... not terminated {#} m/(?&{#}/',
+ 'm/(?(</' => 'Sequence (?(<... not terminated {#} m/(?(<{#}/',
+ "m/(?('/" => "Sequence (?('... not terminated {#} m/(?('{#}/",
+ 'm/\g{/' => 'Sequence \g{... not terminated {#} m/\g{{#}/',
+ 'm/\k</' => 'Sequence \k<... not terminated {#} m/\k<{#}/',
+ 'm/\cß/' => "Character following \"\\c\" must be printable ASCII",
+ '/((?# 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)/',
+ '/((?# 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)/',
+);
+
+my @death_utf8 = mark_as_utf8(
+ '/ネ[[=ネ=]]ネ/' => 'POSIX syntax [= =] is reserved for future extensions {#} m/ネ[[=ネ=]{#}]ネ/',
+ '/ネ(?<= .*)/' => 'Variable length lookbehind not implemented in regex m/ネ(?<= .*)/',
+
+ '/(?<= ネ{1000})/' => 'Lookbehind longer than 255 not implemented in regex m/(?<= ネ{1000})/',
+
+ '/ネ(?ネ)ネ/' => 'Sequence (?ネ...) not recognized {#} m/ネ(?ネ{#})ネ/',
+
+ '/ネ(?(1ネ))ネ/' => 'Switch condition not recognized {#} m/ネ(?(1ネ{#}))ネ/',
+
+ '/(?(1)ネ|y|ヌ)/' => 'Switch (?(condition)... contains too many branches {#} m/(?(1)ネ|y|{#}ヌ)/',
+
+ '/(?(ネ)y|ネ)/' => 'Unknown switch condition (?(...)) {#} m/(?(ネ{#})y|ネ)/',
+
+ '/ネ(?/' => 'Sequence (? incomplete {#} m/ネ(?{#}/',
+
+ '/ネ(?;ネ/' => 'Sequence (?;...) not recognized {#} m/ネ(?;{#}ネ/',
+ '/ネ(?<;ネ/' => 'Group name must start with a non-digit word character {#} m/ネ(?<;{#}ネ/',
+ '/ネ(?\ixネ/' => 'Sequence (?\...) not recognized {#} m/ネ(?\{#}ixネ/',
+ '/ネ(?^lu:ネ)/' => 'Regexp modifiers "l" and "u" are mutually exclusive {#} m/ネ(?^lu{#}:ネ)/',
+'/ネ(?lil:ネ)/' => 'Regexp modifier "l" may not appear twice {#} m/ネ(?lil{#}:ネ)/',
+'/ネ(?aaia:ネ)/' => 'Regexp modifier "a" may appear a maximum of twice {#} m/ネ(?aaia{#}:ネ)/',
+'/ネ(?i-l:ネ)/' => 'Regexp modifier "l" may not appear after the "-" {#} m/ネ(?i-l{#}:ネ)/',
+
+ '/ネ((ネ)/' => 'Unmatched ( {#} m/ネ({#}(ネ)/',
+
+ "/ネ{$inf_p1}ネ/" => "Quantifier in {,} bigger than $inf_m1 {#} m/ネ{{#}$inf_p1}ネ/",
+
+
+ '/ネ**ネ/' => 'Nested quantifiers {#} m/ネ**{#}ネ/',
+
+ '/ネ[ネ/' => 'Unmatched [ {#} m/ネ[{#}ネ/',
+
+ '/*ネ/', => 'Quantifier follows nothing {#} m/*{#}ネ/',
+
+ '/ネ\p{ネ/' => 'Missing right brace on \p{} {#} m/ネ\p{{#}ネ/',
+
+ '/(ネ)\2ネ/' => 'Reference to nonexistent group {#} m/(ネ)\2{#}ネ/',
+
+ '/\g{ネ/; #no latin1' => 'Sequence \g{... not terminated {#} m/\g{ネ{#}/',
+
+ 'my $m = "ネ\\\"; $m =~ $m', => 'Trailing \ in regex m/ネ\/',
+
+ '/\x{ネ/' => 'Missing right brace on \x{} {#} m/\x{{#}ネ/',
+ '/ネ[\x{ネ]ネ/' => 'Missing right brace on \x{} {#} m/ネ[\x{{#}ネ]ネ/',
+ '/ネ[\x{ネ]/' => 'Missing right brace on \x{} {#} m/ネ[\x{{#}ネ]/',
+
+ '/ネ\o{ネ/' => 'Missing right brace on \o{ {#} m/ネ\o{{#}ネ/',
+ '/ネ[[:ネ:]]ネ/' => 'POSIX class [:ネ:] unknown {#} m/ネ[[:ネ:]{#}]ネ/',
+
+ '/ネ[[=ネ=]]ネ/' => 'POSIX syntax [= =] is reserved for future extensions {#} m/ネ[[=ネ=]{#}]ネ/',
+
+ '/ネ[[.ネ.]]ネ/' => 'POSIX syntax [. .] is reserved for future extensions {#} m/ネ[[.ネ.]{#}]ネ/',
+
+ '/[ネ-a]ネ/' => 'Invalid [] range "ネ-a" {#} m/[ネ-a{#}]ネ/',
+
+ '/ネ\p{}ネ/' => 'Empty \p{} {#} m/ネ\p{{#}}ネ/',
+
+ '/ネ(?[[[:ネ]]])ネ/' => "Unmatched ':' in POSIX class {#} m/ネ(?[[[:ネ{#}]]])ネ/",
+ '/ネ(?[[[:ネ: ])ネ/' => "Unmatched '[' in POSIX class {#} m/ネ(?[[[:ネ:{#} ])ネ/",
+ '/ネ(?[[[::]]])ネ/' => "POSIX class [::] unknown {#} m/ネ(?[[[::]{#}]])ネ/",
+ '/ネ(?[[[:ネ:]]])ネ/' => "POSIX class [:ネ:] unknown {#} m/ネ(?[[[:ネ:]{#}]])ネ/",
+ '/ネ(?[[:ネ:]])ネ/' => "POSIX class [:ネ:] unknown {#} m/ネ(?[[:ネ:]{#}])ネ/",
+ '/ネ(?[ネ])ネ/' => 'Unexpected character {#} m/ネ(?[ネ{#}])ネ/',
+ '/ネ(?[ネ])/l' => '(?[...]) not valid in locale {#} m/ネ(?[{#}ネ])/',
+ '/ネ(?[ + [ネ] ])/' => 'Unexpected binary operator \'+\' with no preceding operand {#} m/ネ(?[ +{#} [ネ] ])/',
+ '/ネ(?[ \cK - ( + [ネ] ) ])/' => 'Unexpected binary operator \'+\' with no preceding operand {#} m/ネ(?[ \cK - ( +{#} [ネ] ) ])/',
+ '/ネ(?[ \cK ( [ネ] ) ])/' => 'Unexpected \'(\' with no preceding operator {#} m/ネ(?[ \cK ({#} [ネ] ) ])/',
+ '/ネ(?[ \cK [ネ] ])ネ/' => 'Operand with no preceding operator {#} m/ネ(?[ \cK [ネ{#}] ])ネ/',
+ '/ネ(?[ \0004 ])ネ/' => 'Need exactly 3 octal digits {#} m/ネ(?[ \0004 {#}])ネ/',
+ '/(?[ \o{ネ} ])ネ/' => 'Non-octal character {#} m/(?[ \o{ネ{#}} ])ネ/',
+ '/ネ(?[ \o{} ])ネ/' => 'Number with no digits {#} m/ネ(?[ \o{}{#} ])ネ/',
+ '/(?[ \x{ネ} ])ネ/' => 'Non-hex character {#} m/(?[ \x{ネ{#}} ])ネ/',
+ '/(?[ \p{ネ} ])/' => 'Property \'ネ\' is unknown {#} m/(?[ \p{ネ}{#} ])/',
+ '/(?[ \p{ ネ = bar } ])/' => 'Property \'ネ = bar\' is unknown {#} m/(?[ \p{ ネ = bar }{#} ])/',
+ '/ネ(?[ \t ]/' => 'Syntax error in (?[...]) in regex m/ネ(?[ \t ]/',
+ '/(?[ \t + \e # ネ This was supposed to be a comment ])/' => 'Syntax error in (?[...]) in regex m/(?[ \t + \e # ネ This was supposed to be a comment ])/',
+ 'm/(*ネ)ネ/' => q<Unknown verb pattern 'ネ' {#} m/(*ネ){#}ネ/>,
+ '/\cネ/' => "Character following \"\\c\" must be printable ASCII",
);
+push @death, @death_utf8;
+
# Tests involving a user-defined charnames translator are in pat_advanced.t
# In the following arrays of warnings, the value can be an array of things to
# expect. If the empty string, it means no warning should be raised.
-##
-## Key-value pairs of code/error of code that should have non-fatal regexp warnings.
-##
-my @warning = (
- 'm/\b*/' => '\b* matches null string many times {#} m/\b*{#}/',
- 'm/[:blank:]/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:blank:]{#}/',
+# Key-value pairs of code/error of code that should have non-fatal regexp
+# warnings. Most currently have \x{100} appended to them to force them to be
+# upgraded to UTF-8, and the first pass restarted. Previously this would
+# cause some warnings to be output twice. This tests that that behavior has
+# been fixed.
- "m'[\\y]'" => 'Unrecognized escape \y in character class passed through {#} m/[\y{#}]/',
-
- 'm/[a-\d]/' => 'False [] range "a-\d" {#} m/[a-\d{#}]/',
- 'm/[\w-x]/' => 'False [] range "\w-" {#} m/[\w-{#}x]/',
- 'm/[a-\pM]/' => 'False [] range "a-\pM" {#} m/[a-\pM{#}]/',
- 'm/[\pM-x]/' => 'False [] range "\pM-" {#} m/[\pM-{#}x]/',
- "m'\\y'" => 'Unrecognized escape \y passed through {#} m/\y{#}/',
+my @warning = (
+ 'm/\b*\x{100}/' => '\b* matches null string many times {#} m/\b*{#}\x{100}/',
+ 'm/[:blank:]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:blank:]{#}\x{100}/',
+ "m'[\\y]\\x{100}'" => 'Unrecognized escape \y in character class passed through {#} m/[\y{#}]\x{100}/',
+ 'm/[a-\d]\x{100}/' => 'False [] range "a-\d" {#} m/[a-\d{#}]\x{100}/',
+ 'm/[\w-x]\x{100}/' => 'False [] range "\w-" {#} m/[\w-{#}x]\x{100}/',
+ 'm/[a-\pM]\x{100}/' => 'False [] range "a-\pM" {#} m/[a-\pM{#}]\x{100}/',
+ 'm/[\pM-x]\x{100}/' => 'False [] range "\pM-" {#} m/[\pM-{#}x]\x{100}/',
+ '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}{#}]/',
+ "m'\\y\\x{100}'" => 'Unrecognized escape \y passed through {#} m/\y{#}\x{100}/',
'/x{3,1}/' => 'Quantifier {n,m} with n > m can\'t match {#} m/x{3,1}{#}/',
'/\08/' => '\'\08\' resolved to \'\o{0}8\' {#} m/\08{#}/',
'/\018/' => '\'\018\' resolved to \'\o{1}8\' {#} m/\018{#}/',
'/[\08]/' => '\'\08\' resolved to \'\o{0}8\' {#} m/[\08{#}]/',
'/[\018]/' => '\'\018\' resolved to \'\o{1}8\' {#} m/[\018{#}]/',
- '/\87/' => 'Unrecognized escape \8 passed through {#} m/\8{#}7/',
- '/a\87/' => 'Unrecognized escape \8 passed through {#} m/a\8{#}7/',
- '/a\97/' => 'Unrecognized escape \9 passed through {#} m/a\9{#}7/',
'/(?=a)*/' => '(?=a)* matches null string many times {#} m/(?=a)*{#}/',
'my $x = \'\m\'; qr/a$x/' => 'Unrecognized escape \m passed through {#} m/a\m{#}/',
'/\q/' => 'Unrecognized escape \q passed through {#} m/\q{#}/',
- '/\q{/' => 'Unrecognized escape \q{ passed through {#} m/\q{{#}/',
- '/(?=a){1,3}/' => 'Quantifier unexpected on zero-length expression {#} m/(?=a){1,3}{#}/',
+
+ # Feel free to modify these 2 tests, should they start failing because the
+ # marker of where the problem is becomes wrong. The current behavior is
+ # bad, always marking at the very end of the regex instead of where the
+ # problem is. See [perl #122680] regcomp warning gives wrong position of
+ # problem.
+ '/(?=a){1,3}\x{100}/' => 'Quantifier unexpected on zero-length expression {#} m/(?=a){1,3}\x{100}{#}/',
+ '/(a|b)(?=a){3}\x{100}/' => 'Quantifier unexpected on zero-length expression {#} m/(a|b)(?=a){3}\x{100}{#}/',
+
'/\_/' => "",
'/[\_\0]/' => "",
'/[\07]/' => "",
'/[\006]/' => "",
'/[\0005]/' => "",
- '/[\8\9]/' => ['Unrecognized escape \8 in character class passed through {#} m/[\8{#}\9]/',
- 'Unrecognized escape \9 in character class passed through {#} m/[\8\9{#}]/',
+ '/[\8\9]\x{100}/' => ['Unrecognized escape \8 in character class passed through {#} m/[\8{#}\9]\x{100}/',
+ 'Unrecognized escape \9 in character class passed through {#} m/[\8\9{#}]\x{100}/',
],
- '/[:alpha:]/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:alpha:]{#}/',
- '/[:zog:]/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:zog:]{#}/',
- '/[.zog.]/' => 'POSIX syntax [. .] belongs inside character classes {#} m/[.zog.]{#}/',
+ '/[:alpha:]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:alpha:]{#}\x{100}/',
+ '/[:zog:]\x{100}/' => 'POSIX syntax [: :] belongs inside character classes {#} m/[:zog:]{#}\x{100}/',
+ '/[.zog.]\x{100}/' => 'POSIX syntax [. .] belongs inside character classes {#} m/[.zog.]{#}\x{100}/',
'/[a-b]/' => "",
- '/[a-\d]/' => 'False [] range "a-\d" {#} m/[a-\d{#}]/',
- '/[\d-b]/' => 'False [] range "\d-" {#} m/[\d-{#}b]/',
- '/[\s-\d]/' => 'False [] range "\s-" {#} m/[\s-{#}\d]/',
- '/[\d-\s]/' => 'False [] range "\d-" {#} m/[\d-{#}\s]/',
- '/[a-[:digit:]]/' => 'False [] range "a-[:digit:]" {#} m/[a-[:digit:]{#}]/',
- '/[[:digit:]-b]/' => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}b]/',
- '/[[:alpha:]-[:digit:]]/' => 'False [] range "[:alpha:]-" {#} m/[[:alpha:]-{#}[:digit:]]/',
- '/[[:digit:]-[:alpha:]]/' => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}[:alpha:]]/',
- '/[a\zb]/' => 'Unrecognized escape \z in character class passed through {#} m/[a\z{#}b]/',
- '/(?c)/' => 'Useless (?c) - use /gc modifier {#} m/(?c{#})/',
- '/(?-c)/' => 'Useless (?-c) - don\'t use /gc modifier {#} m/(?-c{#})/',
- '/(?g)/' => 'Useless (?g) - use /g modifier {#} m/(?g{#})/',
- '/(?-g)/' => 'Useless (?-g) - don\'t use /g modifier {#} m/(?-g{#})/',
- '/(?o)/' => 'Useless (?o) - use /o modifier {#} m/(?o{#})/',
- '/(?-o)/' => 'Useless (?-o) - don\'t use /o modifier {#} m/(?-o{#})/',
- '/(?g-o)/' => [ 'Useless (?g) - use /g modifier {#} m/(?g{#}-o)/',
- 'Useless (?-o) - don\'t use /o modifier {#} m/(?g-o{#})/',
+ '/[a-\d]\x{100}/' => 'False [] range "a-\d" {#} m/[a-\d{#}]\x{100}/',
+ '/[\d-b]\x{100}/' => 'False [] range "\d-" {#} m/[\d-{#}b]\x{100}/',
+ '/[\s-\d]\x{100}/' => 'False [] range "\s-" {#} m/[\s-{#}\d]\x{100}/',
+ '/[\d-\s]\x{100}/' => 'False [] range "\d-" {#} m/[\d-{#}\s]\x{100}/',
+ '/[a-[:digit:]]\x{100}/' => 'False [] range "a-[:digit:]" {#} m/[a-[:digit:]{#}]\x{100}/',
+ '/[[:digit:]-b]\x{100}/' => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}b]\x{100}/',
+ '/[[:alpha:]-[:digit:]]\x{100}/' => 'False [] range "[:alpha:]-" {#} m/[[:alpha:]-{#}[:digit:]]\x{100}/',
+ '/[[:digit:]-[:alpha:]]\x{100}/' => 'False [] range "[:digit:]-" {#} m/[[:digit:]-{#}[:alpha:]]\x{100}/',
+ '/[a\zb]\x{100}/' => 'Unrecognized escape \z in character class passed through {#} m/[a\z{#}b]\x{100}/',
+ '/(?c)\x{100}/' => 'Useless (?c) - use /gc modifier {#} m/(?c{#})\x{100}/',
+ '/(?-c)\x{100}/' => 'Useless (?-c) - don\'t use /gc modifier {#} m/(?-c{#})\x{100}/',
+ '/(?g)\x{100}/' => 'Useless (?g) - use /g modifier {#} m/(?g{#})\x{100}/',
+ '/(?-g)\x{100}/' => 'Useless (?-g) - don\'t use /g modifier {#} m/(?-g{#})\x{100}/',
+ '/(?o)\x{100}/' => 'Useless (?o) - use /o modifier {#} m/(?o{#})\x{100}/',
+ '/(?-o)\x{100}/' => 'Useless (?-o) - don\'t use /o modifier {#} m/(?-o{#})\x{100}/',
+ '/(?g-o)\x{100}/' => [ 'Useless (?g) - use /g modifier {#} m/(?g{#}-o)\x{100}/',
+ 'Useless (?-o) - don\'t use /o modifier {#} m/(?g-o{#})\x{100}/',
],
- '/(?g-c)/' => [ 'Useless (?g) - use /g modifier {#} m/(?g{#}-c)/',
- 'Useless (?-c) - don\'t use /gc modifier {#} m/(?g-c{#})/',
+ '/(?g-c)\x{100}/' => [ 'Useless (?g) - use /g modifier {#} m/(?g{#}-c)\x{100}/',
+ 'Useless (?-c) - don\'t use /gc modifier {#} m/(?g-c{#})\x{100}/',
],
# (?c) means (?g) error won't be thrown
- '/(?o-cg)/' => [ 'Useless (?o) - use /o modifier {#} m/(?o{#}-cg)/',
- 'Useless (?-c) - don\'t use /gc modifier {#} m/(?o-c{#}g)/',
+ '/(?o-cg)\x{100}/' => [ 'Useless (?o) - use /o modifier {#} m/(?o{#}-cg)\x{100}/',
+ 'Useless (?-c) - don\'t use /gc modifier {#} m/(?o-c{#}g)\x{100}/',
],
- '/(?ogc)/' => [ 'Useless (?o) - use /o modifier {#} m/(?o{#}gc)/',
- 'Useless (?g) - use /g modifier {#} m/(?og{#}c)/',
- 'Useless (?c) - use /gc modifier {#} m/(?ogc{#})/',
+ '/(?ogc)\x{100}/' => [ 'Useless (?o) - use /o modifier {#} m/(?o{#}gc)\x{100}/',
+ 'Useless (?g) - use /g modifier {#} m/(?og{#}c)\x{100}/',
+ 'Useless (?c) - use /gc modifier {#} m/(?ogc{#})\x{100}/',
],
+ '/a{1,1}?\x{100}/' => 'Useless use of greediness modifier \'?\' {#} m/a{1,1}?{#}\x{100}/',
+ '/b{3} +\x{100}/x' => 'Useless use of greediness modifier \'+\' {#} m/b{3} +{#}\x{100}/',
+); # See comments before this for why '\x{100}' is generally needed
+
+my @warnings_utf8 = mark_as_utf8(
+ 'm/ネ\b*ネ/' => '\b* matches null string many times {#} m/ネ\b*{#}ネ/',
+ '/(?=ネ)*/' => '(?=ネ)* matches null string many times {#} m/(?=ネ)*{#}/',
+ 'm/ネ[:foo:]ネ/' => 'POSIX syntax [: :] belongs inside character classes {#} m/ネ[:foo:]{#}ネ/',
+ "m'ネ[\\y]ネ'" => 'Unrecognized escape \y in character class passed through {#} m/ネ[\y{#}]ネ/',
+ 'm/ネ[ネ-\d]ネ/' => 'False [] range "ネ-\d" {#} m/ネ[ネ-\d{#}]ネ/',
+ 'm/ネ[\w-ネ]ネ/' => 'False [] range "\w-" {#} m/ネ[\w-{#}ネ]ネ/',
+ 'm/ネ[ネ-\pM]ネ/' => 'False [] range "ネ-\pM" {#} m/ネ[ネ-\pM{#}]ネ/',
+ '/ネ[ネ-[:digit:]]ネ/' => 'False [] range "ネ-[:digit:]" {#} m/ネ[ネ-[:digit:]{#}]ネ/',
+ '/ネ[\d-\s]ネ/' => 'False [] range "\d-" {#} m/ネ[\d-{#}\s]ネ/',
+ '/ネ[a\zb]ネ/' => 'Unrecognized escape \z in character class passed through {#} m/ネ[a\z{#}b]ネ/',
+ '/ネ(?c)ネ/' => 'Useless (?c) - use /gc modifier {#} m/ネ(?c{#})ネ/',
+ '/utf8 ネ (?ogc) ネ/' => [
+ 'Useless (?o) - use /o modifier {#} m/utf8 ネ (?o{#}gc) ネ/',
+ 'Useless (?g) - use /g modifier {#} m/utf8 ネ (?og{#}c) ネ/',
+ 'Useless (?c) - use /gc modifier {#} m/utf8 ネ (?ogc{#}) ネ/',
+ ],
+
);
+push @warning, @warnings_utf8;
+
my @experimental_regex_sets = (
'/(?[ \t ])/' => 'The regex_sets feature is experimental {#} m/(?[{#} \t ])/',
+ 'use utf8; /utf8 ネ (?[ [\tネ] ])/' => do { use utf8; 'The regex_sets feature is experimental {#} m/utf8 ネ (?[{#} [\tネ] ])/' },
+ '/noutf8 ネ (?[ [\tネ] ])/' => 'The regex_sets feature is experimental {#} m/noutf8 ネ (?[{#} [\tネ] ])/',
);
my @deprecated = (
- '/a\b{cde/' => '"\b{" is deprecated; use "\b\{" or "\b[{]" instead {#} m/a\{#}b{cde/',
- '/a\B{cde/' => '"\B{" is deprecated; use "\B\{" or "\B[{]" instead {#} m/a\{#}B{cde/',
- 'use utf8; /(?x)\\85\85\\85/' => 'Escape literal pattern white space under /x {#} m/(?x)\\85\85{#}\\85/',
- '/((?# This is a comment in the middle of a token)?:foo)/' => 'In \'(?...)\', splitting the initial \'(?\' is deprecated {#} m/((?# This is a comment in the middle of a token)?{#}:foo)/',
- '/((?# This is a comment in the middle of a token)*FAIL)/' => 'In \'(*VERB...)\', splitting the initial \'(*\' is deprecated {#} m/((?# This is a comment in the middle of a token)*{#}FAIL)/',
+ '/\w{/' => 'Unescaped left brace in regex is deprecated, passed through {#} m/\w{{#}/',
+ '/\q{/' => [
+ 'Unrecognized escape \q{ passed through {#} m/\q{{#}/',
+ 'Unescaped left brace in regex is deprecated, passed through {#} m/\q{{#}/'
+ ],
+ '/:{4,a}/' => 'Unescaped left brace in regex is deprecated, passed through {#} m/:{{#}4,a}/',
);
while (my ($regex, $expect) = splice @death, 0, 2) {
}
}
foreach my $i (0 .. $count - 1) {
- if (like($got[$i], qr/\Q$expect[$i]/, "... and gave expected warning[$i]")) {
+ if (! like($got[$i], qr/\Q$expect[$i]/, "... and gave expected warning")) {
+ chomp($got[$i]);
+ chomp($expect[$i]);
+ diag("GOT\n'$got[$i]'\nEXPECT\n'$expect[$i]'");
+ }
+ else {
ok (0 == capture_warnings(sub {
$_ = "x";
eval "no warnings '$warning_type'; $regex;" }
),
"... and turning off '$warning_type' warnings suppressed it");
+ # Test that whether the warning is on by default is
+ # correct. Experimental and deprecated warnings are;
+ # others are not. This test relies on the fact that we
+ # are outside the scope of any ‘use warnings’.
+ local $^W;
+ my $on = 'on' x ($warning_type ne 'regexp');
+ ok !!$on ==
+ capture_warnings(sub { $_ = "x"; eval $regex }),
+ "... and the warning is " . ($on||'off')
+ . " by default";
}
}
}