pod/perldiag: Document some regex messages

diff --git a/pod/perldiag.pod b/pod/perldiag.pod
index 5761d84f3f3c0f38e6b21c995eb3e935dc8f80fb..8d29e2dca7df2801704f11c714fb1201e9aed1ce 100644 (file)
--- a/pod/perldiag.pod
+++ b/pod/perldiag.pod
@@ -1596,6 +1596,13 @@ valid magic number.
 
 (P) The malloc package that comes with Perl had an internal failure.
 
+=item Corrupted regexp opcode %d > %d
+
+(F)
+This is either an error in Perl, or, if you're using one, your
+L<custom regular expression engine|perlreapi>.  If not the latter,
+report the problem through the L<perlbug> utility.
+
 =item Count after length/code in unpack
 
 (F) You had an unpack template indicating a counted-length string, but
@@ -1783,7 +1790,7 @@ it's ugly.  Your code will be interpreted as an attempt to call a method
 named "elseif" for the class returned by the following block.  This is
 unlikely to be what you want.
 
-=item Empty %s
+=item Empty \%c{} in regex; marked by <-- HERE in m/%s/
 
 (F) C<\p> and C<\P> are used to introduce a named Unicode property, as
 described in L<perlunicode> and L<perlre>.  You used C<\p> or C<\P> in
@@ -1898,6 +1905,17 @@ as a goto, or a loop control statement.
 (W exiting) You are exiting a substitution by unconventional means, such
 as a return, a goto, or a loop control statement.
 
+=item Expecting close bracket in regex; marked by <-- HERE in m/%s/
+
+(F)
+You wrote something like
+
+ (?13
+
+to denote a capturing group of the form
+L<C<(?I<PARNO>)>|perlre/(?PARNO) (?-PARNO) (?+PARNO) (?R) (?0)>,
+but omitted the C<")">.
+
 =item Experimental "%s" subs not enabled
 
 (F) To use lexical subs, you must first enable them:
@@ -2242,6 +2260,17 @@ two from 1 to 32 (or 64, if your platform supports that).
 (W digit) You may have tried to use an 8 or 9 in an octal number.
 Interpretation of the octal number stopped before the 8 or 9.
 
+=item Illegal pattern in regex; marked by <-- HERE in m/%s/
+
+(F)
+You wrote something like
+
+ (?+foo)
+
+The C<"+"> is valid only when followed by digits, indicating a
+capturing group.  See
+L<C<(?I<PARNO>)>|perlre/(?PARNO) (?-PARNO) (?+PARNO) (?R) (?0)>.
+
 =item Illegal switch in PERL5OPT: -%c
 
 (X) The PERL5OPT environment variable may only be used to set the
@@ -2416,6 +2445,18 @@ by Perl or by a user-supplied handler.  See L<attributes>.
 (F) The indicated attributes for a subroutine or variable were not
 recognized by Perl or by a user-supplied handler.  See L<attributes>.
 
+=item Invalid [] range "%*.*s" in regex; marked by <-- HERE in m/%s/
+
+(F)
+You wrote something like
+
+ [z-a]
+
+in a regular expression pattern.  Ranges must be specified with the
+lowest code point first.  Instead write
+
+ [a-z]
+
 =item Invalid character in \N{...}; marked by <-- HERE in \N{%s}
 
 (F) Only certain characters are valid for character names.  The
@@ -5755,6 +5796,12 @@ about.
 
 This is not very useful, and perl currently optimizes this away.
 
+=item Useless use of (?-p) in regex; marked by <-- HERE in m/%s/
+
+(W regexp)
+The C<p> modifier cannot be turned off once set.  Trying to do so is
+futile.
+
 =item Useless use of %s with no values
 
 (W syntax) You used the push() or unshift() function with no arguments
@@ -5764,6 +5811,20 @@ possible in principle that push(@tied_array) could have some effect
 if the array is tied to a class which implements a PUSH method.  If so,
 you can write it as C<push(@tied_array,())> to avoid this warning.
 
+=item Useless (%s%c) - %suse /%c modifier in regex; marked by <-- HERE in m/%s/
+
+(W regexp)
+The C</g> and C</o> regular expression modifiers are global and can't be
+turned off once set; hence things like C<(?g)> or C<(?-o:)> do nothing.
+
+=item Useless (%sc) - %suse /gc modifier in regex; marked by <-- HERE in m/%s/
+
+(W regexp)
+The C</c> regular expression modifier is global, can't be turned off
+once set, and doesn't do anything without the C</g> modifier being
+specified as well; hence things like C<(?c)> or C<(?-c:)> do nothing,
+nor do thing like C<(?gc)> nor C<(?-gc:)> .
+
 =item "use" not allowed in expression
 
 (F) The "use" keyword is recognized and executed at compile time, and

diff --git a/t/porting/diag.t b/t/porting/diag.t
index f09ee6951c151e20a86ff133537bf1086ebc0883..2473e65e45208192d95fcab96ae0434fba91fb50 100644 (file)
--- a/t/porting/diag.t
+++ b/t/porting/diag.t
@@ -490,18 +490,15 @@ chown not implemented!
 clear %s
 Code missing after '/' in pack
 Code missing after '/' in unpack
-Corrupted regexp opcode %d > %d
 '%c' outside of string in pack
 Debug leaking scalars child failed%s with errno %d: %s
 '/' does not take a repeat count in %s
 Don't know how to get file name
 Don't know how to handle magic of type \%o
 -Dp not implemented on this platform
-Empty \%c{} in regex; marked by <-- HERE in m/%s/
 Error reading "%s": %s
 execl not implemented!
 EVAL without pos change exceeded limit in regex
-Expecting close bracket in regex; marked by <-- HERE in m/%s/
 Filehandle opened only for %sput
 Filehandle %s opened only for %sput
 Filehandle STD%s reopened as %s only for input
@@ -523,11 +520,9 @@ Illegal binary digit '%c' ignored
 Illegal character %sin prototype for %s : %s
 Illegal hexadecimal digit '%c' ignored
 Illegal octal digit '%c' ignored
-Illegal pattern in regex; marked by <-- HERE in m/%s/
 Infinite recursion in regex
 internal %<num>p might conflict with future printf extensions
 Invalid argument to sv_cat_decode
-Invalid [] range "%*.*s" in regex; marked by <-- HERE in m/%s/
 Invalid range "%c-%c" in transliteration operator
 Invalid separator character %c%c%c in PerlIO layer specification %s
 Invalid TOKEN object ignored
@@ -634,9 +629,6 @@ Within []-length '%c' not allowed in %s
 Wrong syntax (suid) fd script name "%s"
 'X' outside of string in %s
 'X' outside of string in unpack
-Useless (%s%c) - %suse /%c modifier in regex; marked by <-- HERE in m/%s/
-Useless (%sc) - %suse /gc modifier in regex; marked by <-- HERE in m/%s/
-Useless use of (?-p) in regex; marked by <-- HERE in m/%s/
 
 __CATEGORIES__
 Code point 0x%X is not Unicode, all \p{} matches fail; all \P{} matches succeed