xref: /openbsd/gnu/usr.bin/perl/ext/re/re.pm (revision 3cab2bb3)
1package re;
2
3# pragma for controlling the regexp engine
4use strict;
5use warnings;
6
7our $VERSION     = "0.37";
8our @ISA         = qw(Exporter);
9our @EXPORT_OK   = ('regmust',
10                    qw(is_regexp regexp_pattern
11                       regname regnames regnames_count));
12our %EXPORT_OK = map { $_ => 1 } @EXPORT_OK;
13
14my %bitmask = (
15    taint   => 0x00100000, # HINT_RE_TAINT
16    eval    => 0x00200000, # HINT_RE_EVAL
17);
18
19my $flags_hint = 0x02000000; # HINT_RE_FLAGS
20my $PMMOD_SHIFT = 0;
21my %reflags = (
22    m => 1 << ($PMMOD_SHIFT + 0),
23    s => 1 << ($PMMOD_SHIFT + 1),
24    i => 1 << ($PMMOD_SHIFT + 2),
25    x => 1 << ($PMMOD_SHIFT + 3),
26   xx => 1 << ($PMMOD_SHIFT + 4),
27    n => 1 << ($PMMOD_SHIFT + 5),
28    p => 1 << ($PMMOD_SHIFT + 6),
29    strict => 1 << ($PMMOD_SHIFT + 10),
30# special cases:
31    d => 0,
32    l => 1,
33    u => 2,
34    a => 3,
35    aa => 4,
36);
37
38sub setcolor {
39 eval {				# Ignore errors
40  require Term::Cap;
41
42  my $terminal = Tgetent Term::Cap ({OSPEED => 9600}); # Avoid warning.
43  my $props = $ENV{PERL_RE_TC} || 'md,me,so,se,us,ue';
44  my @props = split /,/, $props;
45  my $colors = join "\t", map {$terminal->Tputs($_,1)} @props;
46
47  $colors =~ s/\0//g;
48  $ENV{PERL_RE_COLORS} = $colors;
49 };
50 if ($@) {
51    $ENV{PERL_RE_COLORS} ||= qq'\t\t> <\t> <\t\t';
52 }
53
54}
55
56my %flags = (
57    COMPILE         => 0x0000FF,
58    PARSE           => 0x000001,
59    OPTIMISE        => 0x000002,
60    TRIEC           => 0x000004,
61    DUMP            => 0x000008,
62    FLAGS           => 0x000010,
63    TEST            => 0x000020,
64
65    EXECUTE         => 0x00FF00,
66    INTUIT          => 0x000100,
67    MATCH           => 0x000200,
68    TRIEE           => 0x000400,
69
70    EXTRA           => 0xFF0000,
71    TRIEM           => 0x010000,
72    OFFSETS         => 0x020000,
73    OFFSETSDBG      => 0x040000,
74    STATE           => 0x080000,
75    OPTIMISEM       => 0x100000,
76    STACK           => 0x280000,
77    BUFFERS         => 0x400000,
78    GPOS            => 0x800000,
79);
80$flags{ALL} = -1 & ~($flags{OFFSETS}|$flags{OFFSETSDBG}|$flags{BUFFERS});
81$flags{All} = $flags{all} = $flags{DUMP} | $flags{EXECUTE};
82$flags{Extra} = $flags{EXECUTE} | $flags{COMPILE} | $flags{GPOS};
83$flags{More} = $flags{MORE} = $flags{All} | $flags{TRIEC} | $flags{TRIEM} | $flags{STATE};
84$flags{State} = $flags{DUMP} | $flags{EXECUTE} | $flags{STATE};
85$flags{TRIE} = $flags{DUMP} | $flags{EXECUTE} | $flags{TRIEC};
86
87if (defined &DynaLoader::boot_DynaLoader) {
88    require XSLoader;
89    XSLoader::load();
90}
91# else we're miniperl
92# We need to work for miniperl, because the XS toolchain uses Text::Wrap, which
93# uses re 'taint'.
94
95sub _load_unload {
96    my ($on)= @_;
97    if ($on) {
98	# We call install() every time, as if we didn't, we wouldn't
99	# "see" any changes to the color environment var since
100	# the last time it was called.
101
102	# install() returns an integer, which if casted properly
103	# in C resolves to a structure containing the regexp
104	# hooks. Setting it to a random integer will guarantee
105	# segfaults.
106	$^H{regcomp} = install();
107    } else {
108        delete $^H{regcomp};
109    }
110}
111
112sub bits {
113    my $on = shift;
114    my $bits = 0;
115    my $turning_all_off = ! @_ && ! $on;
116    if ($turning_all_off) {
117
118        # Pretend were called with certain parameters, which are best dealt
119        # with that way.
120        push @_, keys %bitmask; # taint and eval
121        push @_, 'strict';
122    }
123
124    # Process each subpragma parameter
125   ARG:
126    foreach my $idx (0..$#_){
127        my $s=$_[$idx];
128        if ($s eq 'Debug' or $s eq 'Debugcolor') {
129            setcolor() if $s =~/color/i;
130            ${^RE_DEBUG_FLAGS} = 0 unless defined ${^RE_DEBUG_FLAGS};
131            for my $idx ($idx+1..$#_) {
132                if ($flags{$_[$idx]}) {
133                    if ($on) {
134                        ${^RE_DEBUG_FLAGS} |= $flags{$_[$idx]};
135                    } else {
136                        ${^RE_DEBUG_FLAGS} &= ~ $flags{$_[$idx]};
137                    }
138                } else {
139                    require Carp;
140                    Carp::carp("Unknown \"re\" Debug flag '$_[$idx]', possible flags: ",
141                               join(", ",sort keys %flags ) );
142                }
143            }
144            _load_unload($on ? 1 : ${^RE_DEBUG_FLAGS});
145            last;
146        } elsif ($s eq 'debug' or $s eq 'debugcolor') {
147	    setcolor() if $s =~/color/i;
148	    _load_unload($on);
149	    last;
150        } elsif (exists $bitmask{$s}) {
151	    $bits |= $bitmask{$s};
152	} elsif ($EXPORT_OK{$s}) {
153	    require Exporter;
154	    re->export_to_level(2, 're', $s);
155        } elsif ($s eq 'strict') {
156            if ($on) {
157                $^H{reflags} |= $reflags{$s};
158                warnings::warnif('experimental::re_strict',
159                                 "\"use re 'strict'\" is experimental");
160
161                # Turn on warnings if not already done.
162                if (! warnings::enabled('regexp')) {
163                    require warnings;
164                    warnings->import('regexp');
165                    $^H{re_strict} = 1;
166                }
167            }
168            else {
169                $^H{reflags} &= ~$reflags{$s} if $^H{reflags};
170
171                # Turn off warnings if we turned them on.
172                warnings->unimport('regexp') if $^H{re_strict};
173            }
174	    if ($^H{reflags}) {
175                $^H |= $flags_hint;
176            }
177            else {
178                $^H &= ~$flags_hint;
179            }
180	} elsif ($s =~ s/^\///) {
181	    my $reflags = $^H{reflags} || 0;
182	    my $seen_charset;
183            my $x_count = 0;
184	    while ($s =~ m/( . )/gx) {
185                local $_ = $1;
186		if (/[adul]/) {
187                    # The 'a' may be repeated; hide this from the rest of the
188                    # code by counting and getting rid of all of them, then
189                    # changing to 'aa' if there is a repeat.
190                    if ($_ eq 'a') {
191                        my $sav_pos = pos $s;
192                        my $a_count = $s =~ s/a//g;
193                        pos $s = $sav_pos - 1;  # -1 because got rid of the 'a'
194                        if ($a_count > 2) {
195			    require Carp;
196                            Carp::carp(
197                            qq 'The "a" flag may only appear a maximum of twice'
198                            );
199                        }
200                        elsif ($a_count == 2) {
201                            $_ = 'aa';
202                        }
203                    }
204		    if ($on) {
205			if ($seen_charset) {
206			    require Carp;
207                            if ($seen_charset ne $_) {
208                                Carp::carp(
209                                qq 'The "$seen_charset" and "$_" flags '
210                                .qq 'are exclusive'
211                                );
212                            }
213                            else {
214                                Carp::carp(
215                                qq 'The "$seen_charset" flag may not appear '
216                                .qq 'twice'
217                                );
218                            }
219			}
220			$^H{reflags_charset} = $reflags{$_};
221			$seen_charset = $_;
222		    }
223		    else {
224			delete $^H{reflags_charset}
225                                     if defined $^H{reflags_charset}
226                                        && $^H{reflags_charset} == $reflags{$_};
227		    }
228		} elsif (exists $reflags{$_}) {
229                    if ($_ eq 'x') {
230                        $x_count++;
231                        if ($x_count > 2) {
232			    require Carp;
233                            Carp::carp(
234                            qq 'The "x" flag may only appear a maximum of twice'
235                            );
236                        }
237                        elsif ($x_count == 2) {
238                            $_ = 'xx';  # First time through got the /x
239                        }
240                    }
241
242                    $on
243		      ? $reflags |= $reflags{$_}
244		      : ($reflags &= ~$reflags{$_});
245		} else {
246		    require Carp;
247		    Carp::carp(
248		     qq'Unknown regular expression flag "$_"'
249		    );
250		    next ARG;
251		}
252	    }
253	    ($^H{reflags} = $reflags or defined $^H{reflags_charset})
254	                    ? $^H |= $flags_hint
255	                    : ($^H &= ~$flags_hint);
256	} else {
257	    require Carp;
258	    Carp::carp("Unknown \"re\" subpragma '$s' (known ones are: ",
259                       join(', ', map {qq('$_')} 'debug', 'debugcolor', sort keys %bitmask),
260                       ")");
261	}
262    }
263
264    if ($turning_all_off) {
265        _load_unload(0);
266        $^H{reflags} = 0;
267        $^H{reflags_charset} = 0;
268        $^H &= ~$flags_hint;
269    }
270
271    $bits;
272}
273
274sub import {
275    shift;
276    $^H |= bits(1, @_);
277}
278
279sub unimport {
280    shift;
281    $^H &= ~ bits(0, @_);
282}
283
2841;
285
286__END__
287
288=head1 NAME
289
290re - Perl pragma to alter regular expression behaviour
291
292=head1 SYNOPSIS
293
294    use re 'taint';
295    ($x) = ($^X =~ /^(.*)$/s);     # $x is tainted here
296
297    $pat = '(?{ $foo = 1 })';
298    use re 'eval';
299    /foo${pat}bar/;		   # won't fail (when not under -T
300                                   # switch)
301
302    {
303	no re 'taint';		   # the default
304	($x) = ($^X =~ /^(.*)$/s); # $x is not tainted here
305
306	no re 'eval';		   # the default
307	/foo${pat}bar/;		   # disallowed (with or without -T
308                                   # switch)
309    }
310
311    use re 'strict';               # Raise warnings for more conditions
312
313    use re '/ix';
314    "FOO" =~ / foo /; # /ix implied
315    no re '/x';
316    "FOO" =~ /foo/; # just /i implied
317
318    use re 'debug';		   # output debugging info during
319    /^(.*)$/s;			   # compile and run time
320
321
322    use re 'debugcolor';	   # same as 'debug', but with colored
323                                   # output
324    ...
325
326    use re qw(Debug All);          # Same as "use re 'debug'", but you
327                                   # can use "Debug" with things other
328                                   # than 'All'
329    use re qw(Debug More);         # 'All' plus output more details
330    no re qw(Debug ALL);           # Turn on (almost) all re debugging
331                                   # in this scope
332
333    use re qw(is_regexp regexp_pattern); # import utility functions
334    my ($pat,$mods)=regexp_pattern(qr/foo/i);
335    if (is_regexp($obj)) {
336        print "Got regexp: ",
337            scalar regexp_pattern($obj); # just as perl would stringify
338    }                                    # it but no hassle with blessed
339                                         # re's.
340
341(We use $^X in these examples because it's tainted by default.)
342
343=head1 DESCRIPTION
344
345=head2 'taint' mode
346
347When C<use re 'taint'> is in effect, and a tainted string is the target
348of a regexp, the regexp memories (or values returned by the m// operator
349in list context) are tainted.  This feature is useful when regexp operations
350on tainted data aren't meant to extract safe substrings, but to perform
351other transformations.
352
353=head2 'eval' mode
354
355When C<use re 'eval'> is in effect, a regexp is allowed to contain
356C<(?{ ... })> zero-width assertions and C<(??{ ... })> postponed
357subexpressions that are derived from variable interpolation, rather than
358appearing literally within the regexp.  That is normally disallowed, since
359it is a
360potential security risk.  Note that this pragma is ignored when the regular
361expression is obtained from tainted data, i.e.  evaluation is always
362disallowed with tainted regular expressions.  See L<perlre/(?{ code })>
363and L<perlre/(??{ code })>.
364
365For the purpose of this pragma, interpolation of precompiled regular
366expressions (i.e., the result of C<qr//>) is I<not> considered variable
367interpolation.  Thus:
368
369    /foo${pat}bar/
370
371I<is> allowed if $pat is a precompiled regular expression, even
372if $pat contains C<(?{ ... })> assertions or C<(??{ ... })> subexpressions.
373
374=head2 'strict' mode
375
376Note that this is an experimental feature which may be changed or removed in a
377future Perl release.
378
379When C<use re 'strict'> is in effect, stricter checks are applied than
380otherwise when compiling regular expressions patterns.  These may cause more
381warnings to be raised than otherwise, and more things to be fatal instead of
382just warnings.  The purpose of this is to find and report at compile time some
383things, which may be legal, but have a reasonable possibility of not being the
384programmer's actual intent.  This automatically turns on the C<"regexp">
385warnings category (if not already on) within its scope.
386
387As an example of something that is caught under C<"strict'>, but not
388otherwise, is the pattern
389
390 qr/\xABC/
391
392The C<"\x"> construct without curly braces should be followed by exactly two
393hex digits; this one is followed by three.  This currently evaluates as
394equivalent to
395
396 qr/\x{AB}C/
397
398that is, the character whose code point value is C<0xAB>, followed by the
399letter C<C>.  But since C<C> is a a hex digit, there is a reasonable chance
400that the intent was
401
402 qr/\x{ABC}/
403
404that is the single character at C<0xABC>.  Under C<'strict'> it is an error to
405not follow C<\x> with exactly two hex digits.  When not under C<'strict'> a
406warning is generated if there is only one hex digit, and no warning is raised
407if there are more than two.
408
409It is expected that what exactly C<'strict'> does will evolve over time as we
410gain experience with it.  This means that programs that compile under it in
411today's Perl may not compile, or may have more or fewer warnings, in future
412Perls.  There is no backwards compatibility promises with regards to it.  Also
413there are already proposals for an alternate syntax for enabling it.  For
414these reasons, using it will raise a C<experimental::re_strict> class warning,
415unless that category is turned off.
416
417Note that if a pattern compiled within C<'strict'> is recompiled, say by
418interpolating into another pattern, outside of C<'strict'>, it is not checked
419again for strictness.  This is because if it works under strict it must work
420under non-strict.
421
422=head2 '/flags' mode
423
424When C<use re '/I<flags>'> is specified, the given I<flags> are automatically
425added to every regular expression till the end of the lexical scope.
426I<flags> can be any combination of
427C<'a'>,
428C<'aa'>,
429C<'d'>,
430C<'i'>,
431C<'l'>,
432C<'m'>,
433C<'n'>,
434C<'p'>,
435C<'s'>,
436C<'u'>,
437C<'x'>,
438and/or
439C<'xx'>.
440
441C<no re '/I<flags>'> will turn off the effect of C<use re '/I<flags>'> for the
442given flags.
443
444For example, if you want all your regular expressions to have /msxx on by
445default, simply put
446
447    use re '/msxx';
448
449at the top of your code.
450
451The character set C</adul> flags cancel each other out. So, in this example,
452
453    use re "/u";
454    "ss" =~ /\xdf/;
455    use re "/d";
456    "ss" =~ /\xdf/;
457
458the second C<use re> does an implicit C<no re '/u'>.
459
460Similarly,
461
462    use re "/xx";   # Doubled-x
463    ...
464    use re "/x";    # Single x from here on
465    ...
466
467Turning on one of the character set flags with C<use re> takes precedence over the
468C<locale> pragma and the 'unicode_strings' C<feature>, for regular
469expressions. Turning off one of these flags when it is active reverts to
470the behaviour specified by whatever other pragmata are in scope. For
471example:
472
473    use feature "unicode_strings";
474    no re "/u"; # does nothing
475    use re "/l";
476    no re "/l"; # reverts to unicode_strings behaviour
477
478=head2 'debug' mode
479
480When C<use re 'debug'> is in effect, perl emits debugging messages when
481compiling and using regular expressions.  The output is the same as that
482obtained by running a C<-DDEBUGGING>-enabled perl interpreter with the
483B<-Dr> switch. It may be quite voluminous depending on the complexity
484of the match.  Using C<debugcolor> instead of C<debug> enables a
485form of output that can be used to get a colorful display on terminals
486that understand termcap color sequences.  Set C<$ENV{PERL_RE_TC}> to a
487comma-separated list of C<termcap> properties to use for highlighting
488strings on/off, pre-point part on/off.
489See L<perldebug/"Debugging Regular Expressions"> for additional info.
490
491As of 5.9.5 the directive C<use re 'debug'> and its equivalents are
492lexically scoped, as the other directives are.  However they have both
493compile-time and run-time effects.
494
495See L<perlmodlib/Pragmatic Modules>.
496
497=head2 'Debug' mode
498
499Similarly C<use re 'Debug'> produces debugging output, the difference
500being that it allows the fine tuning of what debugging output will be
501emitted. Options are divided into three groups, those related to
502compilation, those related to execution and those related to special
503purposes. The options are as follows:
504
505=over 4
506
507=item Compile related options
508
509=over 4
510
511=item COMPILE
512
513Turns on all compile related debug options.
514
515=item PARSE
516
517Turns on debug output related to the process of parsing the pattern.
518
519=item OPTIMISE
520
521Enables output related to the optimisation phase of compilation.
522
523=item TRIEC
524
525Detailed info about trie compilation.
526
527=item DUMP
528
529Dump the final program out after it is compiled and optimised.
530
531=item FLAGS
532
533Dump the flags associated with the program
534
535=item TEST
536
537Print output intended for testing the internals of the compile process
538
539=back
540
541=item Execute related options
542
543=over 4
544
545=item EXECUTE
546
547Turns on all execute related debug options.
548
549=item MATCH
550
551Turns on debugging of the main matching loop.
552
553=item TRIEE
554
555Extra debugging of how tries execute.
556
557=item INTUIT
558
559Enable debugging of start-point optimisations.
560
561=back
562
563=item Extra debugging options
564
565=over 4
566
567=item EXTRA
568
569Turns on all "extra" debugging options.
570
571=item BUFFERS
572
573Enable debugging the capture group storage during match. Warning,
574this can potentially produce extremely large output.
575
576=item TRIEM
577
578Enable enhanced TRIE debugging. Enhances both TRIEE
579and TRIEC.
580
581=item STATE
582
583Enable debugging of states in the engine.
584
585=item STACK
586
587Enable debugging of the recursion stack in the engine. Enabling
588or disabling this option automatically does the same for debugging
589states as well. This output from this can be quite large.
590
591=item GPOS
592
593Enable debugging of the \G modifier.
594
595=item OPTIMISEM
596
597Enable enhanced optimisation debugging and start-point optimisations.
598Probably not useful except when debugging the regexp engine itself.
599
600=item OFFSETS
601
602Dump offset information. This can be used to see how regops correlate
603to the pattern. Output format is
604
605   NODENUM:POSITION[LENGTH]
606
607Where 1 is the position of the first char in the string. Note that position
608can be 0, or larger than the actual length of the pattern, likewise length
609can be zero.
610
611=item OFFSETSDBG
612
613Enable debugging of offsets information. This emits copious
614amounts of trace information and doesn't mesh well with other
615debug options.
616
617Almost definitely only useful to people hacking
618on the offsets part of the debug engine.
619
620
621=back
622
623=item Other useful flags
624
625These are useful shortcuts to save on the typing.
626
627=over 4
628
629=item ALL
630
631Enable all options at once except OFFSETS, OFFSETSDBG and BUFFERS.
632(To get every single option without exception, use both ALL and EXTRA, or
633starting in 5.30 on a C<-DDEBUGGING>-enabled perl interpreter, use
634the B<-Drv> command-line switches.)
635
636=item All
637
638Enable DUMP and all execute options. Equivalent to:
639
640  use re 'debug';
641
642=item MORE
643
644=item More
645
646Enable the options enabled by "All", plus STATE, TRIEC, and TRIEM.
647
648=back
649
650=back
651
652As of 5.9.5 the directive C<use re 'debug'> and its equivalents are
653lexically scoped, as are the other directives.  However they have both
654compile-time and run-time effects.
655
656=head2 Exportable Functions
657
658As of perl 5.9.5 're' debug contains a number of utility functions that
659may be optionally exported into the caller's namespace. They are listed
660below.
661
662=over 4
663
664=item is_regexp($ref)
665
666Returns true if the argument is a compiled regular expression as returned
667by C<qr//>, false if it is not.
668
669This function will not be confused by overloading or blessing. In
670internals terms, this extracts the regexp pointer out of the
671PERL_MAGIC_qr structure so it cannot be fooled.
672
673=item regexp_pattern($ref)
674
675If the argument is a compiled regular expression as returned by C<qr//>,
676then this function returns the pattern.
677
678In list context it returns a two element list, the first element
679containing the pattern and the second containing the modifiers used when
680the pattern was compiled.
681
682  my ($pat, $mods) = regexp_pattern($ref);
683
684In scalar context it returns the same as perl would when stringifying a raw
685C<qr//> with the same pattern inside.  If the argument is not a compiled
686reference then this routine returns false but defined in scalar context,
687and the empty list in list context. Thus the following
688
689    if (regexp_pattern($ref) eq '(?^i:foo)')
690
691will be warning free regardless of what $ref actually is.
692
693Like C<is_regexp> this function will not be confused by overloading
694or blessing of the object.
695
696=item regmust($ref)
697
698If the argument is a compiled regular expression as returned by C<qr//>,
699then this function returns what the optimiser considers to be the longest
700anchored fixed string and longest floating fixed string in the pattern.
701
702A I<fixed string> is defined as being a substring that must appear for the
703pattern to match. An I<anchored fixed string> is a fixed string that must
704appear at a particular offset from the beginning of the match. A I<floating
705fixed string> is defined as a fixed string that can appear at any point in
706a range of positions relative to the start of the match. For example,
707
708    my $qr = qr/here .* there/x;
709    my ($anchored, $floating) = regmust($qr);
710    print "anchored:'$anchored'\nfloating:'$floating'\n";
711
712results in
713
714    anchored:'here'
715    floating:'there'
716
717Because the C<here> is before the C<.*> in the pattern, its position
718can be determined exactly. That's not true, however, for the C<there>;
719it could appear at any point after where the anchored string appeared.
720Perl uses both for its optimisations, preferring the longer, or, if they are
721equal, the floating.
722
723B<NOTE:> This may not necessarily be the definitive longest anchored and
724floating string. This will be what the optimiser of the Perl that you
725are using thinks is the longest. If you believe that the result is wrong
726please report it via the L<perlbug> utility.
727
728=item regname($name,$all)
729
730Returns the contents of a named buffer of the last successful match. If
731$all is true, then returns an array ref containing one entry per buffer,
732otherwise returns the first defined buffer.
733
734=item regnames($all)
735
736Returns a list of all of the named buffers defined in the last successful
737match. If $all is true, then it returns all names defined, if not it returns
738only names which were involved in the match.
739
740=item regnames_count()
741
742Returns the number of distinct names defined in the pattern used
743for the last successful match.
744
745B<Note:> this result is always the actual number of distinct
746named buffers defined, it may not actually match that which is
747returned by C<regnames()> and related routines when those routines
748have not been called with the $all parameter set.
749
750=back
751
752=head1 SEE ALSO
753
754L<perlmodlib/Pragmatic Modules>.
755
756=cut
757