1 # EXTRACT VARIOUSLY DELIMITED TEXT SEQUENCES FROM STRINGS.
2 # FOR FULL DOCUMENTATION SEE Balanced.pod
7 package Text::Balanced;
11 use vars qw { $VERSION @ISA %EXPORT_TAGS };
14 @ISA = qw ( Exporter );
16 %EXPORT_TAGS = ( ALL => [ qw(
31 Exporter::export_ok_tags('ALL');
35 sub _match_bracketed($$$$$$);
36 sub _match_variable($$);
37 sub _match_codeblock($$$$$$$);
38 sub _match_quotelike($$$$);
40 # HANDLE RETURN VALUES IN VARIOUS CONTEXTS
43 my ($message, $pos) = @_;
44 $@ = bless { error=>$message, pos=>$pos }, "Text::Balanced::ErrorMsg";
49 my ($wantarray, $textref, $message, $pos) = @_;
50 _failmsg $message, $pos if $message;
51 return ("",$$textref,"") if $wantarray;
58 my ($wantarray,$textref) = splice @_, 0, 2;
59 my ($extrapos, $extralen) = @_>18 ? splice(@_, -2, 2) : (0,0);
60 my ($startlen) = $_[5];
61 my $remainderpos = $_[2];
65 while (my ($from, $len) = splice @_, 0, 2)
67 push @res, substr($$textref,$from,$len);
69 if ($extralen) { # CORRECT FILLET
70 my $extra = substr($res[0], $extrapos-$startlen, $extralen, "\n");
71 $res[1] = "$extra$res[1]";
72 eval { substr($$textref,$remainderpos,0) = $extra;
73 substr($$textref,$extrapos,$extralen,"\n")} ;
74 #REARRANGE HERE DOC AND FILLET IF POSSIBLE
75 pos($$textref) = $remainderpos-$extralen+1; # RESET \G
78 pos($$textref) = $remainderpos; # RESET \G
84 my $match = substr($$textref,$_[0],$_[1]);
85 substr($match,$extrapos-$_[0]-$startlen,$extralen,"") if $extralen;
87 ? substr($$textref, $extrapos, $extralen)."\n" : "";
88 eval {substr($$textref,$_[4],$_[1]+$_[5])=$extra} ; #CHOP OUT PREFIX & MATCH, IF POSSIBLE
89 pos($$textref) = $_[4]; # RESET \G
94 # BUILD A PATTERN MATCHING A SIMPLE DELIMITED STRING
96 sub gen_delimited_pat($;$) # ($delimiters;$escapes)
98 my ($dels, $escs) = @_;
99 return "" unless $dels =~ /\S/;
100 $escs = '\\' unless $escs;
101 $escs .= substr($escs,-1) x (length($dels)-length($escs));
104 for ($i=0; $i<length $dels; $i++)
106 my $del = quotemeta substr($dels,$i,1);
107 my $esc = quotemeta substr($escs,$i,1);
110 push @pat, "$del(?:[^$del]*(?:(?:$del$del)[^$del]*)*)$del";
114 push @pat, "$del(?:[^$esc$del]*(?:$esc.[^$esc$del]*)*)$del";
117 my $pat = join '|', @pat;
121 *delimited_pat = \&gen_delimited_pat;
124 # THE EXTRACTION FUNCTIONS
126 sub extract_delimited (;$$$$)
128 my $textref = defined $_[0] ? \$_[0] : \$_;
129 my $wantarray = wantarray;
130 my $del = defined $_[1] ? $_[1] : qq{\'\"\`};
131 my $pre = defined $_[2] ? $_[2] : '\s*';
132 my $esc = defined $_[3] ? $_[3] : qq{\\};
133 my $pat = gen_delimited_pat($del, $esc);
134 my $startpos = pos $$textref || 0;
135 return _fail($wantarray, $textref, "Not a delimited pattern", 0)
136 unless $$textref =~ m/\G($pre)($pat)/gc;
137 my $prelen = length($1);
138 my $matchpos = $startpos+$prelen;
139 my $endpos = pos $$textref;
140 return _succeed $wantarray, $textref,
141 $matchpos, $endpos-$matchpos, # MATCH
142 $endpos, length($$textref)-$endpos, # REMAINDER
143 $startpos, $prelen; # PREFIX
146 sub extract_bracketed (;$$$)
148 my $textref = defined $_[0] ? \$_[0] : \$_;
149 my $ldel = defined $_[1] ? $_[1] : '{([<';
150 my $pre = defined $_[2] ? $_[2] : '\s*';
151 my $wantarray = wantarray;
154 $ldel =~ s/'//g and $qdel .= q{'};
155 $ldel =~ s/"//g and $qdel .= q{"};
156 $ldel =~ s/`//g and $qdel .= q{`};
157 $ldel =~ s/q//g and $quotelike = 1;
158 $ldel =~ tr/[](){}<>\0-\377/[[(({{<</ds;
160 unless ($rdel =~ tr/[({</])}>/)
162 return _fail $wantarray, $textref,
163 "Did not find a suitable bracket in delimiter: \"$_[1]\"",
167 $ldel = join('|', map { quotemeta $_ } split('', $ldel));
168 $rdel = join('|', map { quotemeta $_ } split('', $rdel));
171 my $startpos = pos $$textref || 0;
172 my @match = _match_bracketed($textref,$pre, $ldel, $qdel, $quotelike, $rdel);
174 return _fail ($wantarray, $textref) unless @match;
176 return _succeed ( $wantarray, $textref,
177 $match[2], $match[5]+2, # MATCH
178 @match[8,9], # REMAINDER
179 @match[0,1], # PREFIX
183 sub _match_bracketed($$$$$$) # $textref, $pre, $ldel, $qdel, $quotelike, $rdel
185 my ($textref, $pre, $ldel, $qdel, $quotelike, $rdel) = @_;
186 my ($startpos, $ldelpos, $endpos) = (pos $$textref = pos $$textref||0);
187 unless ($$textref =~ m/\G$pre/gc)
189 _failmsg "Did not find prefix: /$pre/", $startpos;
193 $ldelpos = pos $$textref;
195 unless ($$textref =~ m/\G($ldel)/gc)
197 _failmsg "Did not find opening bracket after prefix: \"$pre\"",
199 pos $$textref = $startpos;
203 my @nesting = ( $1 );
204 my $textlen = length $$textref;
205 while (pos $$textref < $textlen)
207 next if $$textref =~ m/\G\\./gcs;
209 if ($$textref =~ m/\G($ldel)/gc)
213 elsif ($$textref =~ m/\G($rdel)/gc)
215 my ($found, $brackettype) = ($1, $1);
218 _failmsg "Unmatched closing bracket: \"$found\"",
220 pos $$textref = $startpos;
223 my $expected = pop(@nesting);
224 $expected =~ tr/({[</)}]>/;
225 if ($expected ne $brackettype)
227 _failmsg qq{Mismatched closing bracket: expected "$expected" but found "$found"},
229 pos $$textref = $startpos;
232 last if $#nesting < 0;
234 elsif ($qdel && $$textref =~ m/\G([$qdel])/gc)
236 $$textref =~ m/\G[^\\$1]*(?:\\.[^\\$1]*)*(\Q$1\E)/gsc and next;
237 _failmsg "Unmatched embedded quote ($1)",
239 pos $$textref = $startpos;
242 elsif ($quotelike && _match_quotelike($textref,"",1,0))
247 else { $$textref =~ m/\G(?:[a-zA-Z0-9]+|.)/gcs }
251 _failmsg "Unmatched opening bracket(s): "
252 . join("..",@nesting)."..",
254 pos $$textref = $startpos;
258 $endpos = pos $$textref;
261 $startpos, $ldelpos-$startpos, # PREFIX
262 $ldelpos, 1, # OPENING BRACKET
263 $ldelpos+1, $endpos-$ldelpos-2, # CONTENTS
264 $endpos-1, 1, # CLOSING BRACKET
265 $endpos, length($$textref)-$endpos, # REMAINDER
271 my $brack = reverse $_[0];
272 $brack =~ tr/[({</])}>/;
276 my $XMLNAME = q{[a-zA-Z_:][a-zA-Z0-9_:.-]*};
278 sub extract_tagged (;$$$$$) # ($text, $opentag, $closetag, $pre, \%options)
280 my $textref = defined $_[0] ? \$_[0] : \$_;
283 my $pre = defined $_[3] ? $_[3] : '\s*';
284 my %options = defined $_[4] ? %{$_[4]} : ();
285 my $omode = defined $options{fail} ? $options{fail} : '';
286 my $bad = ref($options{reject}) eq 'ARRAY' ? join('|', @{$options{reject}})
287 : defined($options{reject}) ? $options{reject}
290 my $ignore = ref($options{ignore}) eq 'ARRAY' ? join('|', @{$options{ignore}})
291 : defined($options{ignore}) ? $options{ignore}
295 if (!defined $ldel) { $ldel = '<\w+(?:' . gen_delimited_pat(q{'"}) . '|[^>])*>'; }
298 my @match = _match_tagged($textref, $pre, $ldel, $rdel, $omode, $bad, $ignore);
300 return _fail(wantarray, $textref) unless @match;
301 return _succeed wantarray, $textref,
302 $match[2], $match[3]+$match[5]+$match[7], # MATCH
303 @match[8..9,0..1,2..7]; # REM, PRE, BITS
306 sub _match_tagged # ($$$$$$$)
308 my ($textref, $pre, $ldel, $rdel, $omode, $bad, $ignore) = @_;
311 my ($startpos, $opentagpos, $textpos, $parapos, $closetagpos, $endpos) = ( pos($$textref) = pos($$textref)||0 );
313 unless ($$textref =~ m/\G($pre)/gc)
315 _failmsg "Did not find prefix: /$pre/", pos $$textref;
319 $opentagpos = pos($$textref);
321 unless ($$textref =~ m/\G$ldel/gc)
323 _failmsg "Did not find opening tag: /$ldel/", pos $$textref;
327 $textpos = pos($$textref);
332 unless ($rdelspec =~ s/\A([[(<{]+)($XMLNAME).*/ quotemeta "$1\/$2". revbracket($1) /oes)
334 _failmsg "Unable to construct closing tag to match: $rdel",
341 $rdelspec = eval "qq{$rdel}";
344 while (pos($$textref) < length($$textref))
346 next if $$textref =~ m/\G\\./gc;
348 if ($$textref =~ m/\G(\n[ \t]*\n)/gc )
350 $parapos = pos($$textref) - length($1)
351 unless defined $parapos;
353 elsif ($$textref =~ m/\G($rdelspec)/gc )
355 $closetagpos = pos($$textref)-length($1);
358 elsif ($ignore && $$textref =~ m/\G(?:$ignore)/gc)
362 elsif ($bad && $$textref =~ m/\G($bad)/gcs)
364 pos($$textref) -= length($1); # CUT OFF WHATEVER CAUSED THE SHORTNESS
365 goto short if ($omode eq 'PARA' || $omode eq 'MAX');
366 _failmsg "Found invalid nested tag: $1", pos $$textref;
369 elsif ($$textref =~ m/\G($ldel)/gc)
372 pos($$textref) -= length($tag); # REWIND TO NESTED TAG
373 unless (_match_tagged(@_)) # MATCH NESTED TAG
375 goto short if $omode eq 'PARA' || $omode eq 'MAX';
376 _failmsg "Found unbalanced nested tag: $tag",
381 else { $$textref =~ m/./gcs }
385 $closetagpos = pos($$textref);
386 goto matched if $omode eq 'MAX';
387 goto failed unless $omode eq 'PARA';
389 if (defined $parapos) { pos($$textref) = $parapos }
390 else { $parapos = pos($$textref) }
393 $startpos, $opentagpos-$startpos, # PREFIX
394 $opentagpos, $textpos-$opentagpos, # OPENING TAG
395 $textpos, $parapos-$textpos, # TEXT
396 $parapos, 0, # NO CLOSING TAG
397 $parapos, length($$textref)-$parapos, # REMAINDER
401 $endpos = pos($$textref);
403 $startpos, $opentagpos-$startpos, # PREFIX
404 $opentagpos, $textpos-$opentagpos, # OPENING TAG
405 $textpos, $closetagpos-$textpos, # TEXT
406 $closetagpos, $endpos-$closetagpos, # CLOSING TAG
407 $endpos, length($$textref)-$endpos, # REMAINDER
411 _failmsg "Did not find closing tag", pos $$textref unless $@;
412 pos($$textref) = $startpos;
416 sub extract_variable (;$$)
418 my $textref = defined $_[0] ? \$_[0] : \$_;
419 return ("","","") unless defined $$textref;
420 my $pre = defined $_[1] ? $_[1] : '\s*';
422 my @match = _match_variable($textref,$pre);
424 return _fail wantarray, $textref unless @match;
426 return _succeed wantarray, $textref,
427 @match[2..3,4..5,0..1]; # MATCH, REMAINDER, PREFIX
430 sub _match_variable($$)
435 my ($textref, $pre) = @_;
436 my $startpos = pos($$textref) = pos($$textref)||0;
437 unless ($$textref =~ m/\G($pre)/gc)
439 _failmsg "Did not find prefix: /$pre/", pos $$textref;
442 my $varpos = pos($$textref);
443 unless ($$textref =~ m{\G\$\s*(\d+|[][&`'+*./|,";%=~:?!\@<>()-]|\^[a-z]?)}gci)
445 unless ($$textref =~ m/\G((\$#?|[*\@\%]|\\&)+)/gc)
447 _failmsg "Did not find leading dereferencer", pos $$textref;
448 pos $$textref = $startpos;
453 unless ($$textref =~ m/\G\s*(?:::|')?(?:[_a-z]\w*(?:::|'))*[_a-z]\w*/gci
454 or _match_codeblock($textref, "", '\{', '\}', '\{', '\}', 0)
455 or $deref eq '$#' or $deref eq '$$' )
457 _failmsg "Bad identifier after dereferencer", pos $$textref;
458 pos $$textref = $startpos;
465 next if _match_codeblock($textref,
466 qr/\s*->\s*(?:[_a-zA-Z]\w+\s*)?/,
467 qr/[({[]/, qr/[)}\]]/,
468 qr/[({[]/, qr/[)}\]]/, 0);
469 next if _match_codeblock($textref,
470 qr/\s*/, qr/[{[]/, qr/[}\]]/,
471 qr/[{[]/, qr/[}\]]/, 0);
472 next if _match_variable($textref,'\s*->\s*');
473 next if $$textref =~ m/\G\s*->\s*\w+(?![{([])/gc;
477 my $endpos = pos($$textref);
478 return ($startpos, $varpos-$startpos,
479 $varpos, $endpos-$varpos,
480 $endpos, length($$textref)-$endpos
484 sub extract_codeblock (;$$$$$)
486 my $textref = defined $_[0] ? \$_[0] : \$_;
487 my $wantarray = wantarray;
488 my $ldel_inner = defined $_[1] ? $_[1] : '{';
489 my $pre = defined $_[2] ? $_[2] : '\s*';
490 my $ldel_outer = defined $_[3] ? $_[3] : $ldel_inner;
492 my $rdel_inner = $ldel_inner;
493 my $rdel_outer = $ldel_outer;
495 for ($ldel_inner, $ldel_outer) { tr/[]()<>{}\0-\377/[[((<<{{/ds }
496 for ($rdel_inner, $rdel_outer) { tr/[]()<>{}\0-\377/]]))>>}}/ds }
497 for ($ldel_inner, $ldel_outer, $rdel_inner, $rdel_outer)
499 $_ = '('.join('|',map { quotemeta $_ } split('',$_)).')'
503 my @match = _match_codeblock($textref, $pre,
504 $ldel_outer, $rdel_outer,
505 $ldel_inner, $rdel_inner,
507 return _fail($wantarray, $textref) unless @match;
508 return _succeed($wantarray, $textref,
509 @match[2..3,4..5,0..1] # MATCH, REMAINDER, PREFIX
514 sub _match_codeblock($$$$$$$)
516 my ($textref, $pre, $ldel_outer, $rdel_outer, $ldel_inner, $rdel_inner, $rd) = @_;
517 my $startpos = pos($$textref) = pos($$textref) || 0;
518 unless ($$textref =~ m/\G($pre)/gc)
520 _failmsg qq{Did not match prefix /$pre/ at"} .
521 substr($$textref,pos($$textref),20) .
526 my $codepos = pos($$textref);
527 unless ($$textref =~ m/\G($ldel_outer)/gc) # OUTERMOST DELIMITER
529 _failmsg qq{Did not find expected opening bracket at "} .
530 substr($$textref,pos($$textref),20) .
533 pos $$textref = $startpos;
537 $closing =~ tr/([<{/)]>}/;
540 while (pos($$textref) < length($$textref))
543 if ($rd && $$textref =~ m#\G(\Q(?)\E|\Q(s?)\E|\Q(s)\E)#gc)
549 if ($$textref =~ m/\G\s*#.*/gc)
554 if ($$textref =~ m/\G\s*($rdel_outer)/gc)
556 unless ($matched = ($closing && $1 eq $closing) )
558 next if $1 eq '>'; # MIGHT BE A "LESS THAN"
559 _failmsg q{Mismatched closing bracket at "} .
560 substr($$textref,pos($$textref),20) .
561 qq{...". Expected '$closing'},
567 if (_match_variable($textref,'\s*') ||
568 _match_quotelike($textref,'\s*',$patvalid,$patvalid) )
575 # NEED TO COVER MANY MORE CASES HERE!!!
576 if ($$textref =~ m#\G\s*( [-+*x/%^&|.]=?
579 | (\*\*|&&|\|\||<<|>>)=?
580 | split|grep|map|return
587 if ( _match_codeblock($textref, '\s*', $ldel_inner, $rdel_inner, $ldel_inner, $rdel_inner, $rd) )
593 if ($$textref =~ m/\G\s*$ldel_outer/gc)
595 _failmsg q{Improperly nested codeblock at "} .
596 substr($$textref,pos($$textref),20) .
603 $$textref =~ m/\G\s*(\w+|[-=>]>|.|\Z)/gc;
605 continue { $@ = undef }
609 _failmsg 'No match found for opening bracket', pos $$textref
614 my $endpos = pos($$textref);
615 return ( $startpos, $codepos-$startpos,
616 $codepos, $endpos-$codepos,
617 $endpos, length($$textref)-$endpos,
623 'none' => '[cgimsox]*',
625 's' => '[cegimsox]*',
635 sub extract_quotelike (;$$)
637 my $textref = $_[0] ? \$_[0] : \$_;
638 my $wantarray = wantarray;
639 my $pre = defined $_[1] ? $_[1] : '\s*';
641 my @match = _match_quotelike($textref,$pre,1,0);
642 return _fail($wantarray, $textref) unless @match;
643 return _succeed($wantarray, $textref,
644 $match[2], $match[18]-$match[2], # MATCH
645 @match[18,19], # REMAINDER
646 @match[0,1], # PREFIX
647 @match[2..17], # THE BITS
648 @match[20,21], # ANY FILLET?
652 sub _match_quotelike($$$$) # ($textref, $prepat, $allow_raw_match)
654 my ($textref, $pre, $rawmatch, $qmark) = @_;
656 my ($textlen,$startpos,
658 $preld1pos,$ld1pos,$str1pos,$rd1pos,
659 $preld2pos,$ld2pos,$str2pos,$rd2pos,
660 $modpos) = ( length($$textref), pos($$textref) = pos($$textref) || 0 );
662 unless ($$textref =~ m/\G($pre)/gc)
664 _failmsg qq{Did not find prefix /$pre/ at "} .
665 substr($$textref, pos($$textref), 20) .
670 $oppos = pos($$textref);
672 my $initial = substr($$textref,$oppos,1);
674 if ($initial && $initial =~ m|^[\"\'\`]|
675 || $rawmatch && $initial =~ m|^/|
676 || $qmark && $initial =~ m|^\?|)
678 unless ($$textref =~ m/ \Q$initial\E [^\\$initial]* (\\.[^\\$initial]*)* \Q$initial\E /gcsx)
680 _failmsg qq{Did not find closing delimiter to match '$initial' at "} .
681 substr($$textref, $oppos, 20) .
684 pos $$textref = $startpos;
687 $modpos= pos($$textref);
690 if ($initial eq '/' || $initial eq '?')
692 $$textref =~ m/\G$mods{none}/gc
695 my $endpos = pos($$textref);
697 $startpos, $oppos-$startpos, # PREFIX
698 $oppos, 0, # NO OPERATOR
699 $oppos, 1, # LEFT DEL
700 $oppos+1, $rd1pos-$oppos-1, # STR/PAT
701 $rd1pos, 1, # RIGHT DEL
702 $modpos, 0, # NO 2ND LDEL
703 $modpos, 0, # NO 2ND STR
704 $modpos, 0, # NO 2ND RDEL
705 $modpos, $endpos-$modpos, # MODIFIERS
706 $endpos, $textlen-$endpos, # REMAINDER
710 unless ($$textref =~ m{\G((?:m|s|qq|qx|qw|q|qr|tr|y)\b(?=\s*\S)|<<)}gc)
712 _failmsg q{No quotelike operator found after prefix at "} .
713 substr($$textref, pos($$textref), 20) .
716 pos $$textref = $startpos;
721 $preld1pos = pos($$textref);
723 $ld1pos = pos($$textref);
725 if ($$textref =~ m{\G([A-Za-z_]\w*)}gc) {
728 elsif ($$textref =~ m{ \G ' ([^'\\]* (?:\\.[^'\\]*)*) '
729 | \G " ([^"\\]* (?:\\.[^"\\]*)*) "
730 | \G ` ([^`\\]* (?:\\.[^`\\]*)*) `
737 my $extrapos = pos($$textref);
738 $$textref =~ m{.*\n}gc;
739 $str1pos = pos($$textref);
740 unless ($$textref =~ m{.*?\n(?=$label\n)}gc) {
741 _failmsg qq{Missing here doc terminator ('$label') after "} .
742 substr($$textref, $startpos, 20) .
745 pos $$textref = $startpos;
748 $rd1pos = pos($$textref);
749 $$textref =~ m{$label\n}gc;
750 $ld2pos = pos($$textref);
752 $startpos, $oppos-$startpos, # PREFIX
753 $oppos, length($op), # OPERATOR
754 $ld1pos, $extrapos-$ld1pos, # LEFT DEL
755 $str1pos, $rd1pos-$str1pos, # STR/PAT
756 $rd1pos, $ld2pos-$rd1pos, # RIGHT DEL
757 $ld2pos, 0, # NO 2ND LDEL
758 $ld2pos, 0, # NO 2ND STR
759 $ld2pos, 0, # NO 2ND RDEL
760 $ld2pos, 0, # NO MODIFIERS
761 $ld2pos, $textlen-$ld2pos, # REMAINDER
762 $extrapos, $str1pos-$extrapos, # FILLETED BIT
766 $$textref =~ m/\G\s*/gc;
767 $ld1pos = pos($$textref);
768 $str1pos = $ld1pos+1;
770 unless ($$textref =~ m/\G(\S)/gc) # SHOULD USE LOOKAHEAD
772 _failmsg "No block delimiter found after quotelike $op",
774 pos $$textref = $startpos;
777 pos($$textref) = $ld1pos; # HAVE TO DO THIS BECAUSE LOOKAHEAD BROKEN
778 my ($ldel1, $rdel1) = ("\Q$1","\Q$1");
779 if ($ldel1 =~ /[[(<{]/)
781 $rdel1 =~ tr/[({</])}>/;
782 _match_bracketed($textref,"",$ldel1,"","",$rdel1)
783 || do { pos $$textref = $startpos; return };
787 $$textref =~ /$ldel1[^\\$ldel1]*(\\.[^\\$ldel1]*)*$ldel1/gcs
788 || do { pos $$textref = $startpos; return };
790 $ld2pos = $rd1pos = pos($$textref)-1;
792 my $second_arg = $op =~ /s|tr|y/ ? 1 : 0;
796 if ($ldel1 =~ /[[(<{]/)
798 unless ($$textref =~ /\G\s*(\S)/gc) # SHOULD USE LOOKAHEAD
800 _failmsg "Missing second block for quotelike $op",
802 pos $$textref = $startpos;
805 $ldel2 = $rdel2 = "\Q$1";
806 $rdel2 =~ tr/[({</])}>/;
810 $ldel2 = $rdel2 = $ldel1;
812 $str2pos = $ld2pos+1;
814 if ($ldel2 =~ /[[(<{]/)
816 pos($$textref)--; # OVERCOME BROKEN LOOKAHEAD
817 _match_bracketed($textref,"",$ldel2,"","",$rdel2)
818 || do { pos $$textref = $startpos; return };
822 $$textref =~ /[^\\$ldel2]*(\\.[^\\$ldel2]*)*$ldel2/gcs
823 || do { pos $$textref = $startpos; return };
825 $rd2pos = pos($$textref)-1;
829 $ld2pos = $str2pos = $rd2pos = $rd1pos;
832 $modpos = pos $$textref;
834 $$textref =~ m/\G($mods{$op})/gc;
835 my $endpos = pos $$textref;
838 $startpos, $oppos-$startpos, # PREFIX
839 $oppos, length($op), # OPERATOR
840 $ld1pos, 1, # LEFT DEL
841 $str1pos, $rd1pos-$str1pos, # STR/PAT
842 $rd1pos, 1, # RIGHT DEL
843 $ld2pos, $second_arg, # 2ND LDEL (MAYBE)
844 $str2pos, $rd2pos-$str2pos, # 2ND STR (MAYBE)
845 $rd2pos, $second_arg, # 2ND RDEL (MAYBE)
846 $modpos, $endpos-$modpos, # MODIFIERS
847 $endpos, $textlen-$endpos, # REMAINDER
853 sub { extract_variable($_[0], '') },
854 sub { extract_quotelike($_[0],'') },
855 sub { extract_codeblock($_[0],'{}','') },
858 sub extract_multiple (;$$$$) # ($text, $functions_ref, $max_fields, $ignoreunknown)
860 my $textref = defined($_[0]) ? \$_[0] : \$_;
862 my ($lastpos, $firstpos);
867 my @func = defined $_[1] ? @{$_[1]} : @{$def_func};
868 my $max = defined $_[2] && $_[2]>0 ? $_[2] : 1_000_000_000;
876 carp "extract_multiple reset maximal count to 1 in scalar context"
877 if $^W && defined($_[2]) && $max > 1;
886 foreach $func ( @func )
888 if (ref($func) eq 'HASH')
890 push @class, (keys %$func)[0];
891 $func = (values %$func)[0];
899 FIELD: while (pos($$textref) < length($$textref))
903 foreach my $i ( 0..$#func )
908 $lastpos = pos $$textref;
909 if (ref($func) eq 'CODE')
910 { ($field,undef,$pref) = @bits = $func->($$textref) }
911 elsif (ref($func) eq 'Text::Balanced::Extractor')
912 { @bits = $field = $func->extract($$textref) }
913 elsif( $$textref =~ m/\G$func/gc )
914 { @bits = $field = defined($1) ? $1 : $& }
916 if (defined($field) && length($field))
919 $unkpos = pos $$textref
920 if length($pref) && !defined($unkpos);
923 push @fields, substr($$textref, $unkpos, $lastpos-$unkpos).$pref;
924 $firstpos = $unkpos unless defined $firstpos;
926 last FIELD if @fields == $max;
930 ? bless (\$field, $class)
932 $firstpos = $lastpos unless defined $firstpos;
933 $lastpos = pos $$textref;
934 last FIELD if @fields == $max;
938 if ($$textref =~ /\G(.)/gcs)
940 $unkpos = pos($$textref)-1
941 unless $igunk || defined $unkpos;
947 push @fields, substr($$textref, $unkpos);
948 $firstpos = $unkpos unless defined $firstpos;
949 $lastpos = length $$textref;
954 pos $$textref = $lastpos;
955 return @fields if wantarray;
958 eval { substr($$textref,$firstpos,$lastpos-$firstpos)="";
959 pos $$textref = $firstpos };
964 sub gen_extract_tagged # ($opentag, $closetag, $pre, \%options)
968 my $pre = defined $_[2] ? $_[2] : '\s*';
969 my %options = defined $_[3] ? %{$_[3]} : ();
970 my $omode = defined $options{fail} ? $options{fail} : '';
971 my $bad = ref($options{reject}) eq 'ARRAY' ? join('|', @{$options{reject}})
972 : defined($options{reject}) ? $options{reject}
975 my $ignore = ref($options{ignore}) eq 'ARRAY' ? join('|', @{$options{ignore}})
976 : defined($options{ignore}) ? $options{ignore}
980 if (!defined $ldel) { $ldel = '<\w+(?:' . gen_delimited_pat(q{'"}) . '|[^>])*>'; }
983 for ($ldel, $pre, $bad, $ignore) { $_ = qr/$_/ if $_ }
988 my $textref = defined $_[0] ? \$_[0] : \$_;
989 my @match = Text::Balanced::_match_tagged($textref, $pre, $ldel, $rdel, $omode, $bad, $ignore);
991 return _fail(wantarray, $textref) unless @match;
992 return _succeed wantarray, $textref,
993 $match[2], $match[3]+$match[5]+$match[7], # MATCH
994 @match[8..9,0..1,2..7]; # REM, PRE, BITS
997 bless $closure, 'Text::Balanced::Extractor';
1000 package Text::Balanced::Extractor;
1002 sub extract($$) # ($self, $text)
1007 package Text::Balanced::ErrorMsg;
1009 use overload '""' => sub { "$_[0]->{error}, detected at offset $_[0]->{pos}" };
1017 Text::Balanced - Extract delimited text sequences from strings.
1022 use Text::Balanced qw (
1035 # Extract the initial substring of $text that is delimited by
1036 # two (unescaped) instances of the first character in $delim.
1038 ($extracted, $remainder) = extract_delimited($text,$delim);
1041 # Extract the initial substring of $text that is bracketed
1042 # with a delimiter(s) specified by $delim (where the string
1043 # in $delim contains one or more of '(){}[]<>').
1045 ($extracted, $remainder) = extract_bracketed($text,$delim);
1048 # Extract the initial substring of $text that is bounded by
1051 ($extracted, $remainder) = extract_tagged($text);
1054 # Extract the initial substring of $text that is bounded by
1055 # a C<BEGIN>...C<END> pair. Don't allow nested C<BEGIN> tags
1057 ($extracted, $remainder) =
1058 extract_tagged($text,"BEGIN","END",undef,{bad=>["BEGIN"]});
1061 # Extract the initial substring of $text that represents a
1062 # Perl "quote or quote-like operation"
1064 ($extracted, $remainder) = extract_quotelike($text);
1067 # Extract the initial substring of $text that represents a block
1068 # of Perl code, bracketed by any of character(s) specified by $delim
1069 # (where the string $delim contains one or more of '(){}[]<>').
1071 ($extracted, $remainder) = extract_codeblock($text,$delim);
1074 # Extract the initial substrings of $text that would be extracted by
1075 # one or more sequential applications of the specified functions
1076 # or regular expressions
1078 @extracted = extract_multiple($text,
1079 [ \&extract_bracketed,
1080 \&extract_quotelike,
1081 \&some_other_extractor_sub,
1086 # Create a string representing an optimized pattern (a la Friedl)
1087 # that matches a substring delimited by any of the specified characters
1088 # (in this case: any type of quote or a slash)
1090 $patstring = gen_delimited_pat(q{'"`/});
1093 # Generate a reference to an anonymous sub that is just like extract_tagged
1094 # but pre-compiled and optimized for a specific pair of tags, and consequently
1095 # much faster (i.e. 3 times faster). It uses qr// for better performance on
1096 # repeated calls, so it only works under Perl 5.005 or later.
1098 $extract_head = gen_extract_tagged('<HEAD>','</HEAD>');
1100 ($extracted, $remainder) = $extract_head->($text);
1105 The various C<extract_...> subroutines may be used to extract a
1106 delimited string (possibly after skipping a specified prefix string).
1107 The search for the string always begins at the current C<pos>
1108 location of the string's variable (or at index zero, if no C<pos>
1109 position is defined).
1111 =head2 General behaviour in list contexts
1113 In a list context, all the subroutines return a list, the first three
1114 elements of which are always:
1120 The extracted string, including the specified delimiters.
1121 If the extraction fails an empty string is returned.
1125 The remainder of the input string (i.e. the characters after the
1126 extracted string). On failure, the entire string is returned.
1130 The skipped prefix (i.e. the characters before the extracted string).
1131 On failure, the empty string is returned.
1135 Note that in a list context, the contents of the original input text (the first
1136 argument) are not modified in any way.
1138 However, if the input text was passed in a variable, that variable's
1139 C<pos> value is updated to point at the first character after the
1140 extracted text. That means that in a list context the various
1141 subroutines can be used much like regular expressions. For example:
1143 while ( $next = (extract_quotelike($text))[0] )
1145 # process next quote-like (in $next)
1149 =head2 General behaviour in scalar and void contexts
1151 In a scalar context, the extracted string is returned, having first been
1152 removed from the input text. Thus, the following code also processes
1153 each quote-like operation, but actually removes them from $text:
1155 while ( $next = extract_quotelike($text) )
1157 # process next quote-like (in $next)
1160 Note that if the input text is a read-only string (i.e. a literal),
1161 no attempt is made to remove the extracted text.
1163 In a void context the behaviour of the extraction subroutines is
1164 exactly the same as in a scalar context, except (of course) that the
1165 extracted substring is not returned.
1167 =head2 A note about prefixes
1169 Prefix patterns are matched without any trailing modifiers (C</gimsox> etc.)
1170 This can bite you if you're expecting a prefix specification like
1171 '.*?(?=<H1>)' to skip everything up to the first <H1> tag. Such a prefix
1172 pattern will only succeed if the <H1> tag is on the current line, since
1173 . normally doesn't match newlines.
1175 To overcome this limitation, you need to turn on /s matching within
1176 the prefix pattern, using the C<(?s)> directive: '(?s).*?(?=<H1>)'
1179 =head2 C<extract_delimited>
1181 The C<extract_delimited> function formalizes the common idiom
1182 of extracting a single-character-delimited substring from the start of
1183 a string. For example, to extract a single-quote delimited string, the
1184 following code is typically used:
1186 ($remainder = $text) =~ s/\A('(\\.|[^'])*')//s;
1189 but with C<extract_delimited> it can be simplified to:
1191 ($extracted,$remainder) = extract_delimited($text, "'");
1193 C<extract_delimited> takes up to four scalars (the input text, the
1194 delimiters, a prefix pattern to be skipped, and any escape characters)
1195 and extracts the initial substring of the text that
1196 is appropriately delimited. If the delimiter string has multiple
1197 characters, the first one encountered in the text is taken to delimit
1199 The third argument specifies a prefix pattern that is to be skipped
1200 (but must be present!) before the substring is extracted.
1201 The final argument specifies the escape character to be used for each
1204 All arguments are optional. If the escape characters are not specified,
1205 every delimiter is escaped with a backslash (C<\>).
1206 If the prefix is not specified, the
1207 pattern C<'\s*'> - optional whitespace - is used. If the delimiter set
1208 is also not specified, the set C</["'`]/> is used. If the text to be processed
1209 is not specified either, C<$_> is used.
1211 In list context, C<extract_delimited> returns an array of three
1212 elements, the extracted substring (I<including the surrounding
1213 delimiters>), the remainder of the text, and the skipped prefix (if
1214 any). If a suitable delimited substring is not found, the first
1215 element of the array is the empty string, the second is the complete
1216 original text, and the prefix returned in the third element is an
1219 In a scalar context, just the extracted substring is returned. In
1220 a void context, the extracted substring (and any prefix) are simply
1221 removed from the beginning of the first argument.
1225 # Remove a single-quoted substring from the very beginning of $text:
1227 $substring = extract_delimited($text, "'", '');
1229 # Remove a single-quoted Pascalish substring (i.e. one in which
1230 # doubling the quote character escapes it) from the very
1231 # beginning of $text:
1233 $substring = extract_delimited($text, "'", '', "'");
1235 # Extract a single- or double- quoted substring from the
1236 # beginning of $text, optionally after some whitespace
1237 # (note the list context to protect $text from modification):
1239 ($substring) = extract_delimited $text, q{"'};
1242 # Delete the substring delimited by the first '/' in $text:
1244 $text = join '', (extract_delimited($text,'/','[^/]*')[2,1];
1246 Note that this last example is I<not> the same as deleting the first
1247 quote-like pattern. For instance, if C<$text> contained the string:
1249 "if ('./cmd' =~ m/$UNIXCMD/s) { $cmd = $1; }"
1251 then after the deletion it would contain:
1253 "if ('.$UNIXCMD/s) { $cmd = $1; }"
1257 "if ('./cmd' =~ ms) { $cmd = $1; }"
1260 See L<"extract_quotelike"> for a (partial) solution to this problem.
1263 =head2 C<extract_bracketed>
1265 Like C<"extract_delimited">, the C<extract_bracketed> function takes
1266 up to three optional scalar arguments: a string to extract from, a delimiter
1267 specifier, and a prefix pattern. As before, a missing prefix defaults to
1268 optional whitespace and a missing text defaults to C<$_>. However, a missing
1269 delimiter specifier defaults to C<'{}()[]E<lt>E<gt>'> (see below).
1271 C<extract_bracketed> extracts a balanced-bracket-delimited
1272 substring (using any one (or more) of the user-specified delimiter
1273 brackets: '(..)', '{..}', '[..]', or '<..>'). Optionally it will also
1274 respect quoted unbalanced brackets (see below).
1276 A "delimiter bracket" is a bracket in list of delimiters passed as
1277 C<extract_bracketed>'s second argument. Delimiter brackets are
1278 specified by giving either the left or right (or both!) versions
1279 of the required bracket(s). Note that the order in which
1280 two or more delimiter brackets are specified is not significant.
1282 A "balanced-bracket-delimited substring" is a substring bounded by
1283 matched brackets, such that any other (left or right) delimiter
1284 bracket I<within> the substring is also matched by an opposite
1285 (right or left) delimiter bracket I<at the same level of nesting>. Any
1286 type of bracket not in the delimiter list is treated as an ordinary
1289 In other words, each type of bracket specified as a delimiter must be
1290 balanced and correctly nested within the substring, and any other kind of
1291 ("non-delimiter") bracket in the substring is ignored.
1293 For example, given the string:
1295 $text = "{ an '[irregularly :-(] {} parenthesized >:-)' string }";
1297 then a call to C<extract_bracketed> in a list context:
1299 @result = extract_bracketed( $text, '{}' );
1303 ( "{ an '[irregularly :-(] {} parenthesized >:-)' string }" , "" , "" )
1305 since both sets of C<'{..}'> brackets are properly nested and evenly balanced.
1306 (In a scalar context just the first element of the array would be returned. In
1307 a void context, C<$text> would be replaced by an empty string.)
1309 Likewise the call in:
1311 @result = extract_bracketed( $text, '{[' );
1313 would return the same result, since all sets of both types of specified
1314 delimiter brackets are correctly nested and balanced.
1316 However, the call in:
1318 @result = extract_bracketed( $text, '{([<' );
1320 would fail, returning:
1322 ( undef , "{ an '[irregularly :-(] {} parenthesized >:-)' string }" );
1324 because the embedded pairs of C<'(..)'>s and C<'[..]'>s are "cross-nested" and
1325 the embedded C<'E<gt>'> is unbalanced. (In a scalar context, this call would
1326 return an empty string. In a void context, C<$text> would be unchanged.)
1328 Note that the embedded single-quotes in the string don't help in this
1329 case, since they have not been specified as acceptable delimiters and are
1330 therefore treated as non-delimiter characters (and ignored).
1332 However, if a particular species of quote character is included in the
1333 delimiter specification, then that type of quote will be correctly handled.
1334 for example, if C<$text> is:
1336 $text = '<A HREF=">>>>">link</A>';
1340 @result = extract_bracketed( $text, '<">' );
1344 ( '<A HREF=">>>>">', 'link</A>', "" )
1346 as expected. Without the specification of C<"> as an embedded quoter:
1348 @result = extract_bracketed( $text, '<>' );
1350 the result would be:
1352 ( '<A HREF=">', '>>>">link</A>', "" )
1354 In addition to the quote delimiters C<'>, C<">, and C<`>, full Perl quote-like
1355 quoting (i.e. q{string}, qq{string}, etc) can be specified by including the
1356 letter 'q' as a delimiter. Hence:
1358 @result = extract_bracketed( $text, '<q>' );
1360 would correctly match something like this:
1362 $text = '<leftop: conj /and/ conj>';
1364 See also: C<"extract_quotelike"> and C<"extract_codeblock">.
1367 =head2 C<extract_tagged>
1369 C<extract_tagged> extracts and segments text between (balanced)
1372 The subroutine takes up to five optional arguments:
1378 A string to be processed (C<$_> if the string is omitted or C<undef>)
1382 A string specifying a pattern to be matched as the opening tag.
1383 If the pattern string is omitted (or C<undef>) then a pattern
1384 that matches any standard HTML/XML tag is used.
1388 A string specifying a pattern to be matched at the closing tag.
1389 If the pattern string is omitted (or C<undef>) then the closing
1390 tag is constructed by inserting a C</> after any leading bracket
1391 characters in the actual opening tag that was matched (I<not> the pattern
1392 that matched the tag). For example, if the opening tag pattern
1393 is specified as C<'{{\w+}}'> and actually matched the opening tag
1394 C<"{{DATA}}">, then the constructed closing tag would be C<"{{/DATA}}">.
1398 A string specifying a pattern to be matched as a prefix (which is to be
1399 skipped). If omitted, optional whitespace is skipped.
1403 A hash reference containing various parsing options (see below)
1407 The various options that can be specified are:
1411 =item C<reject =E<gt> $listref>
1413 The list reference contains one or more strings specifying patterns
1414 that must I<not> appear within the tagged text.
1416 For example, to extract
1417 an HTML link (which should not contain nested links) use:
1419 extract_tagged($text, '<A>', '</A>', undef, {reject => ['<A>']} );
1421 =item C<ignore =E<gt> $listref>
1423 The list reference contains one or more strings specifying patterns
1424 that are I<not> be be treated as nested tags within the tagged text
1425 (even if they would match the start tag pattern).
1427 For example, to extract an arbitrary XML tag, but ignore "empty" elements:
1429 extract_tagged($text, undef, undef, undef, {ignore => ['<[^>]*/>']} );
1431 (also see L<"gen_delimited_pat"> below).
1434 =item C<fail =E<gt> $str>
1436 The C<fail> option indicates the action to be taken if a matching end
1437 tag is not encountered (i.e. before the end of the string or some
1438 C<reject> pattern matches). By default, a failure to match a closing
1439 tag causes C<extract_tagged> to immediately fail.
1441 However, if the string value associated with <reject> is "MAX", then
1442 C<extract_tagged> returns the complete text up to the point of failure.
1443 If the string is "PARA", C<extract_tagged> returns only the first paragraph
1444 after the tag (up to the first line that is either empty or contains
1445 only whitespace characters).
1446 If the string is "", the default behaviour (i.e. failure) is reinstated.
1448 For example, suppose the start tag "/para" introduces a paragraph, which then
1449 continues until the next "/endpara" tag or until another "/para" tag is
1452 $text = "/para line 1\n\nline 3\n/para line 4";
1454 extract_tagged($text, '/para', '/endpara', undef,
1455 {reject => '/para', fail => MAX );
1457 # EXTRACTED: "/para line 1\n\nline 3\n"
1459 Suppose instead, that if no matching "/endpara" tag is found, the "/para"
1460 tag refers only to the immediately following paragraph:
1462 $text = "/para line 1\n\nline 3\n/para line 4";
1464 extract_tagged($text, '/para', '/endpara', undef,
1465 {reject => '/para', fail => MAX );
1467 # EXTRACTED: "/para line 1\n"
1469 Note that the specified C<fail> behaviour applies to nested tags as well.
1473 On success in a list context, an array of 6 elements is returned. The elements are:
1479 the extracted tagged substring (including the outermost tags),
1483 the remainder of the input text,
1487 the prefix substring (if any),
1495 the text between the opening and closing tags
1499 the closing tag (or "" if no closing tag was found)
1503 On failure, all of these values (except the remaining text) are C<undef>.
1505 In a scalar context, C<extract_tagged> returns just the complete
1506 substring that matched a tagged text (including the start and end
1507 tags). C<undef> is returned on failure. In addition, the original input
1508 text has the returned substring (and any prefix) removed from it.
1510 In a void context, the input text just has the matched substring (and
1511 any specified prefix) removed.
1514 =head2 C<gen_extract_tagged>
1516 (Note: This subroutine is only available under Perl5.005)
1518 C<gen_extract_tagged> generates a new anonymous subroutine which
1519 extracts text between (balanced) specified tags. In other words,
1520 it generates a function identical in function to C<extract_tagged>.
1522 The difference between C<extract_tagged> and the anonymous
1523 subroutines generated by
1524 C<gen_extract_tagged>, is that those generated subroutines:
1530 do not have to reparse tag specification or parsing options every time
1531 they are called (whereas C<extract_tagged> has to effectively rebuild
1532 its tag parser on every call);
1536 make use of the new qr// construct to pre-compile the regexes they use
1537 (whereas C<extract_tagged> uses standard string variable interpolation
1538 to create tag-matching patterns).
1542 The subroutine takes up to four optional arguments (the same set as
1543 C<extract_tagged> except for the string to be processed). It returns
1544 a reference to a subroutine which in turn takes a single argument (the text to
1547 In other words, the implementation of C<extract_tagged> is exactly
1553 $extractor = gen_extract_tagged(@_);
1554 return $extractor->($text);
1557 (although C<extract_tagged> is not currently implemented that way, in order
1558 to preserve pre-5.005 compatibility).
1560 Using C<gen_extract_tagged> to create extraction functions for specific tags
1561 is a good idea if those functions are going to be called more than once, since
1562 their performance is typically twice as good as the more general-purpose
1566 =head2 C<extract_quotelike>
1568 C<extract_quotelike> attempts to recognize, extract, and segment any
1569 one of the various Perl quotes and quotelike operators (see
1570 L<perlop(3)>) Nested backslashed delimiters, embedded balanced bracket
1571 delimiters (for the quotelike operators), and trailing modifiers are
1572 all caught. For example, in:
1574 extract_quotelike 'q # an octothorpe: \# (not the end of the q!) #'
1576 extract_quotelike ' "You said, \"Use sed\"." '
1578 extract_quotelike ' s{([A-Z]{1,8}\.[A-Z]{3})} /\L$1\E/; '
1580 extract_quotelike ' tr/\\\/\\\\/\\\//ds; '
1582 the full Perl quotelike operations are all extracted correctly.
1584 Note too that, when using the /x modifier on a regex, any comment
1585 containing the current pattern delimiter will cause the regex to be
1586 immediately terminated. In other words:
1589 (?i) # CASE INSENSITIVE
1590 [a-z_] # LEADING ALPHABETIC/UNDERSCORE
1591 [a-z0-9]* # FOLLOWED BY ANY NUMBER OF ALPHANUMERICS
1594 will be extracted as if it were:
1597 (?i) # CASE INSENSITIVE
1598 [a-z_] # LEADING ALPHABETIC/'
1600 This behaviour is identical to that of the actual compiler.
1602 C<extract_quotelike> takes two arguments: the text to be processed and
1603 a prefix to be matched at the very beginning of the text. If no prefix
1604 is specified, optional whitespace is the default. If no text is given,
1607 In a list context, an array of 11 elements is returned. The elements are:
1613 the extracted quotelike substring (including trailing modifiers),
1617 the remainder of the input text,
1621 the prefix substring (if any),
1625 the name of the quotelike operator (if any),
1629 the left delimiter of the first block of the operation,
1633 the text of the first block of the operation
1634 (that is, the contents of
1635 a quote, the regex of a match or substitution or the target list of a
1640 the right delimiter of the first block of the operation,
1644 the left delimiter of the second block of the operation
1645 (that is, if it is an C<s>, C<tr>, or C<y>),
1649 the text of the second block of the operation
1650 (that is, the replacement of a substitution or the translation list
1655 the right delimiter of the second block of the operation (if any),
1659 the trailing modifiers on the operation (if any).
1663 For each of the fields marked "(if any)" the default value on success is
1665 On failure, all of these values (except the remaining text) are C<undef>.
1668 In a scalar context, C<extract_quotelike> returns just the complete substring
1669 that matched a quotelike operation (or C<undef> on failure). In a scalar or
1670 void context, the input text has the same substring (and any specified
1675 # Remove the first quotelike literal that appears in text
1677 $quotelike = extract_quotelike($text,'.*?');
1679 # Replace one or more leading whitespace-separated quotelike
1680 # literals in $_ with "<QLL>"
1682 do { $_ = join '<QLL>', (extract_quotelike)[2,1] } until $@;
1685 # Isolate the search pattern in a quotelike operation from $text
1687 ($op,$pat) = (extract_quotelike $text)[3,5];
1690 print "search pattern: $pat\n";
1694 print "$op is not a pattern matching operation\n";
1698 =head2 C<extract_quotelike> and "here documents"
1700 C<extract_quotelike> can successfully extract "here documents" from an input
1701 string, but with an important caveat in list contexts.
1703 Unlike other types of quote-like literals, a here document is rarely
1704 a contiguous substring. For example, a typical piece of code using
1705 here document might look like this:
1708 This is the message.
1712 Given this as an input string in a scalar context, C<extract_quotelike>
1713 would correctly return the string "<<'EOMSG'\nThis is the message.\nEOMSG",
1714 leaving the string " || die;\nexit;" in the original variable. In other words,
1715 the two separate pieces of the here document are successfully extracted and
1718 In a list context, C<extract_quotelike> would return the list
1724 "<<'EOMSG'\nThis is the message.\nEOMSG\n" (i.e. the full extracted here document,
1725 including fore and aft delimiters),
1729 " || die;\nexit;" (i.e. the remainder of the input text, concatenated),
1733 "" (i.e. the prefix substring -- trivial in this case),
1737 "<<" (i.e. the "name" of the quotelike operator)
1741 "'EOMSG'" (i.e. the left delimiter of the here document, including any quotes),
1745 "This is the message.\n" (i.e. the text of the here document),
1749 "EOMSG" (i.e. the right delimiter of the here document),
1753 "" (a here document has no second left delimiter, second text, second right
1754 delimiter, or trailing modifiers).
1758 However, the matching position of the input variable would be set to
1759 "exit;" (i.e. I<after> the closing delimiter of the here document),
1760 which would cause the earlier " || die;\nexit;" to be skipped in any
1761 sequence of code fragment extractions.
1763 To avoid this problem, when it encounters a here document while
1764 extracting from a modifiable string, C<extract_quotelike> silently
1765 rearranges the string to an equivalent piece of Perl:
1768 This is the message.
1773 in which the here document I<is> contiguous. It still leaves the
1774 matching position after the here document, but now the rest of the line
1775 on which the here document starts is not skipped.
1777 To prevent <extract_quotelike> from mucking about with the input in this way
1778 (this is the only case where a list-context C<extract_quotelike> does so),
1779 you can pass the input variable as an interpolated literal:
1781 $quotelike = extract_quotelike("$var");
1784 =head2 C<extract_codeblock>
1786 C<extract_codeblock> attempts to recognize and extract a balanced
1787 bracket delimited substring that may contain unbalanced brackets
1788 inside Perl quotes or quotelike operations. That is, C<extract_codeblock>
1789 is like a combination of C<"extract_bracketed"> and
1790 C<"extract_quotelike">.
1792 C<extract_codeblock> takes the same initial three parameters as C<extract_bracketed>:
1793 a text to process, a set of delimiter brackets to look for, and a prefix to
1794 match first. It also takes an optional fourth parameter, which allows the
1795 outermost delimiter brackets to be specified separately (see below).
1797 Omitting the first argument (input text) means process C<$_> instead.
1798 Omitting the second argument (delimiter brackets) indicates that only C<'{'> is to be used.
1799 Omitting the third argument (prefix argument) implies optional whitespace at the start.
1800 Omitting the fourth argument (outermost delimiter brackets) indicates that the
1801 value of the second argument is to be used for the outermost delimiters.
1803 Once the prefix an the outermost opening delimiter bracket have been
1804 recognized, code blocks are extracted by stepping through the input text and
1805 trying the following alternatives in sequence:
1811 Try and match a closing delimiter bracket. If the bracket was the same
1812 species as the last opening bracket, return the substring to that
1813 point. If the bracket was mismatched, return an error.
1817 Try to match a quote or quotelike operator. If found, call
1818 C<extract_quotelike> to eat it. If C<extract_quotelike> fails, return
1819 the error it returned. Otherwise go back to step 1.
1823 Try to match an opening delimiter bracket. If found, call
1824 C<extract_codeblock> recursively to eat the embedded block. If the
1825 recursive call fails, return an error. Otherwise, go back to step 1.
1829 Unconditionally match a bareword or any other single character, and
1830 then go back to step 1.
1837 # Find a while loop in the text
1839 if ($text =~ s/.*?while\s*\{/{/)
1841 $loop = "while " . extract_codeblock($text);
1844 # Remove the first round-bracketed list (which may include
1845 # round- or curly-bracketed code blocks or quotelike operators)
1847 extract_codeblock $text, "(){}", '[^(]*';
1850 The ability to specify a different outermost delimiter bracket is useful
1851 in some circumstances. For example, in the Parse::RecDescent module,
1852 parser actions which are to be performed only on a successful parse
1853 are specified using a C<E<lt>defer:...E<gt>> directive. For example:
1855 sentence: subject verb object
1856 <defer: {$::theVerb = $item{verb}} >
1858 Parse::RecDescent uses C<extract_codeblock($text, '{}E<lt>E<gt>')> to extract the code
1859 within the C<E<lt>defer:...E<gt>> directive, but there's a problem.
1861 A deferred action like this:
1863 <defer: {if ($count>10) {$count--}} >
1865 will be incorrectly parsed as:
1867 <defer: {if ($count>
1869 because the "less than" operator is interpreted as a closing delimiter.
1871 But, by extracting the directive using
1872 S<C<extract_codeblock($text, '{}', undef, 'E<lt>E<gt>')>>
1873 the '>' character is only treated as a delimited at the outermost
1874 level of the code block, so the directive is parsed correctly.
1876 =head2 C<extract_multiple>
1878 The C<extract_multiple> subroutine takes a string to be processed and a
1879 list of extractors (subroutines or regular expressions) to apply to that string.
1881 In an array context C<extract_multiple> returns an array of substrings
1882 of the original string, as extracted by the specified extractors.
1883 In a scalar context, C<extract_multiple> returns the first
1884 substring successfully extracted from the original string. In both
1885 scalar and void contexts the original string has the first successfully
1886 extracted substring removed from it. In all contexts
1887 C<extract_multiple> starts at the current C<pos> of the string, and
1888 sets that C<pos> appropriately after it matches.
1890 Hence, the aim of a call to C<extract_multiple> in a list context
1891 is to split the processed string into as many non-overlapping fields as
1892 possible, by repeatedly applying each of the specified extractors
1893 to the remainder of the string. Thus C<extract_multiple> is
1894 a generalized form of Perl's C<split> subroutine.
1896 The subroutine takes up to four optional arguments:
1902 A string to be processed (C<$_> if the string is omitted or C<undef>)
1906 A reference to a list of subroutine references and/or qr// objects and/or
1907 literal strings and/or hash references, specifying the extractors
1908 to be used to split the string. If this argument is omitted (or
1912 sub { extract_variable($_[0], '') },
1913 sub { extract_quotelike($_[0],'') },
1914 sub { extract_codeblock($_[0],'{}','') },
1922 A number specifying the maximum number of fields to return. If this
1923 argument is omitted (or C<undef>), split continues as long as possible.
1925 If the third argument is I<N>, then extraction continues until I<N> fields
1926 have been successfully extracted, or until the string has been completely
1929 Note that in scalar and void contexts the value of this argument is
1930 automatically reset to 1 (under C<-w>, a warning is issued if the argument
1935 A value indicating whether unmatched substrings (see below) within the
1936 text should be skipped or returned as fields. If the value is true,
1937 such substrings are skipped. Otherwise, they are returned.
1941 The extraction process works by applying each extractor in
1942 sequence to the text string.
1944 If the extractor is a subroutine it is called in a list context and is
1945 expected to return a list of a single element, namely the extracted
1946 text. It may optionally also return two further arguments: a string
1947 representing the text left after extraction (like $' for a pattern
1948 match), and a string representing any prefix skipped before the
1949 extraction (like $` in a pattern match). Note that this is designed
1950 to facilitate the use of other Text::Balanced subroutines with
1951 C<extract_multiple>. Note too that the value returned by an extractor
1952 subroutine need not bear any relationship to the corresponding substring
1953 of the original text (see examples below).
1955 If the extractor is a precompiled regular expression or a string,
1956 it is matched against the text in a scalar context with a leading
1957 '\G' and the gc modifiers enabled. The extracted value is either
1958 $1 if that variable is defined after the match, or else the
1959 complete match (i.e. $&).
1961 If the extractor is a hash reference, it must contain exactly one element.
1962 The value of that element is one of the
1963 above extractor types (subroutine reference, regular expression, or string).
1964 The key of that element is the name of a class into which the successful
1965 return value of the extractor will be blessed.
1967 If an extractor returns a defined value, that value is immediately
1968 treated as the next extracted field and pushed onto the list of fields.
1969 If the extractor was specified in a hash reference, the field is also
1970 blessed into the appropriate class,
1972 If the extractor fails to match (in the case of a regex extractor), or returns an empty list or an undefined value (in the case of a subroutine extractor), it is
1973 assumed to have failed to extract.
1974 If none of the extractor subroutines succeeds, then one
1975 character is extracted from the start of the text and the extraction
1976 subroutines reapplied. Characters which are thus removed are accumulated and
1977 eventually become the next field (unless the fourth argument is true, in which
1978 case they are discarded).
1980 For example, the following extracts substrings that are valid Perl variables:
1982 @fields = extract_multiple($text,
1983 [ sub { extract_variable($_[0]) } ],
1986 This example separates a text into fields which are quote delimited,
1987 curly bracketed, and anything else. The delimited and bracketed
1988 parts are also blessed to identify them (the "anything else" is unblessed):
1990 @fields = extract_multiple($text,
1992 { Delim => sub { extract_delimited($_[0],q{'"}) } },
1993 { Brack => sub { extract_bracketed($_[0],'{}') } },
1996 This call extracts the next single substring that is a valid Perl quotelike
1997 operator (and removes it from $text):
1999 $quotelike = extract_multiple($text,
2001 sub { extract_quotelike($_[0]) },
2004 Finally, here is yet another way to do comma-separated value parsing:
2006 @fields = extract_multiple($csv_text,
2008 sub { extract_delimited($_[0],q{'"}) },
2013 The list in the second argument means:
2014 I<"Try and extract a ' or " delimited string, otherwise extract anything up to a comma...">.
2015 The undef third argument means:
2016 I<"...as many times as possible...">,
2017 and the true value in the fourth argument means
2018 I<"...discarding anything else that appears (i.e. the commas)">.
2020 If you wanted the commas preserved as separate fields (i.e. like split
2021 does if your split pattern has capturing parentheses), you would
2022 just make the last parameter undefined (or remove it).
2025 =head2 C<gen_delimited_pat>
2027 The C<gen_delimited_pat> subroutine takes a single (string) argument and
2028 > builds a Friedl-style optimized regex that matches a string delimited
2029 by any one of the characters in the single argument. For example:
2031 gen_delimited_pat(q{'"})
2035 (?:\"(?:\\\"|(?!\").)*\"|\'(?:\\\'|(?!\').)*\')
2037 Note that the specified delimiters are automatically quotemeta'd.
2039 A typical use of C<gen_delimited_pat> would be to build special purpose tags
2040 for C<extract_tagged>. For example, to properly ignore "empty" XML elements
2041 (which might contain quoted strings):
2043 my $empty_tag = '<(' . gen_delimited_pat(q{'"}) . '|.)+/>';
2045 extract_tagged($text, undef, undef, undef, {ignore => [$empty_tag]} );
2048 C<gen_delimited_pat> may also be called with an optional second argument,
2049 which specifies the "escape" character(s) to be used for each delimiter.
2050 For example to match a Pascal-style string (where ' is the delimiter
2051 and '' is a literal ' within the string):
2053 gen_delimited_pat(q{'},q{'});
2055 Different escape characters can be specified for different delimiters.
2056 For example, to specify that '/' is the escape for single quotes
2057 and '%' is the escape for double quotes:
2059 gen_delimited_pat(q{'"},q{/%});
2061 If more delimiters than escape chars are specified, the last escape char
2062 is used for the remaining delimiters.
2063 If no escape char is specified for a given specified delimiter, '\' is used.
2066 C<gen_delimited_pat> was previously called
2067 C<delimited_pat>. That name may still be used, but is now deprecated.
2072 In a list context, all the functions return C<(undef,$original_text)>
2073 on failure. In a scalar context, failure is indicated by returning C<undef>
2074 (in this case the input text is not modified in any way).
2076 In addition, on failure in I<any> context, the C<$@> variable is set.
2077 Accessing C<$@-E<gt>{error}> returns one of the error diagnostics listed
2079 Accessing C<$@-E<gt>{pos}> returns the offset into the original string at
2080 which the error was detected (although not necessarily where it occurred!)
2081 Printing C<$@> directly produces the error message, with the offset appended.
2082 On success, the C<$@> variable is guaranteed to be C<undef>.
2084 The available diagnostics are:
2088 =item C<Did not find a suitable bracket: "%s">
2090 The delimiter provided to C<extract_bracketed> was not one of
2091 C<'()[]E<lt>E<gt>{}'>.
2093 =item C<Did not find prefix: /%s/>
2095 A non-optional prefix was specified but wasn't found at the start of the text.
2097 =item C<Did not find opening bracket after prefix: "%s">
2099 C<extract_bracketed> or C<extract_codeblock> was expecting a
2100 particular kind of bracket at the start of the text, and didn't find it.
2102 =item C<No quotelike operator found after prefix: "%s">
2104 C<extract_quotelike> didn't find one of the quotelike operators C<q>,
2105 C<qq>, C<qw>, C<qx>, C<s>, C<tr> or C<y> at the start of the substring
2108 =item C<Unmatched closing bracket: "%c">
2110 C<extract_bracketed>, C<extract_quotelike> or C<extract_codeblock> encountered
2111 a closing bracket where none was expected.
2113 =item C<Unmatched opening bracket(s): "%s">
2115 C<extract_bracketed>, C<extract_quotelike> or C<extract_codeblock> ran
2116 out of characters in the text before closing one or more levels of nested
2119 =item C<Unmatched embedded quote (%s)>
2121 C<extract_bracketed> attempted to match an embedded quoted substring, but
2122 failed to find a closing quote to match it.
2124 =item C<Did not find closing delimiter to match '%s'>
2126 C<extract_quotelike> was unable to find a closing delimiter to match the
2127 one that opened the quote-like operation.
2129 =item C<Mismatched closing bracket: expected "%c" but found "%s">
2131 C<extract_bracketed>, C<extract_quotelike> or C<extract_codeblock> found
2132 a valid bracket delimiter, but it was the wrong species. This usually
2133 indicates a nesting error, but may indicate incorrect quoting or escaping.
2135 =item C<No block delimiter found after quotelike "%s">
2137 C<extract_quotelike> or C<extract_codeblock> found one of the
2138 quotelike operators C<q>, C<qq>, C<qw>, C<qx>, C<s>, C<tr> or C<y>
2139 without a suitable block after it.
2141 =item C<Did not find leading dereferencer>
2143 C<extract_variable> was expecting one of '$', '@', or '%' at the start of
2144 a variable, but didn't find any of them.
2146 =item C<Bad identifier after dereferencer>
2148 C<extract_variable> found a '$', '@', or '%' indicating a variable, but that
2149 character was not followed by a legal Perl identifier.
2151 =item C<Did not find expected opening bracket at %s>
2153 C<extract_codeblock> failed to find any of the outermost opening brackets
2154 that were specified.
2156 =item C<Improperly nested codeblock at %s>
2158 A nested code block was found that started with a delimiter that was specified
2159 as being only to be used as an outermost bracket.
2161 =item C<Missing second block for quotelike "%s">
2163 C<extract_codeblock> or C<extract_quotelike> found one of the
2164 quotelike operators C<s>, C<tr> or C<y> followed by only one block.
2166 =item C<No match found for opening bracket>
2168 C<extract_codeblock> failed to find a closing bracket to match the outermost
2171 =item C<Did not find opening tag: /%s/>
2173 C<extract_tagged> did not find a suitable opening tag (after any specified
2174 prefix was removed).
2176 =item C<Unable to construct closing tag to match: /%s/>
2178 C<extract_tagged> matched the specified opening tag and tried to
2179 modify the matched text to produce a matching closing tag (because
2180 none was specified). It failed to generate the closing tag, almost
2181 certainly because the opening tag did not start with a
2182 bracket of some kind.
2184 =item C<Found invalid nested tag: %s>
2186 C<extract_tagged> found a nested tag that appeared in the "reject" list
2187 (and the failure mode was not "MAX" or "PARA").
2189 =item C<Found unbalanced nested tag: %s>
2191 C<extract_tagged> found a nested opening tag that was not matched by a
2192 corresponding nested closing tag (and the failure mode was not "MAX" or "PARA").
2194 =item C<Did not find closing tag>
2196 C<extract_tagged> reached the end of the text without finding a closing tag
2197 to match the original opening tag (and the failure mode was not
2208 Damian Conway (damian@conway.org)
2211 =head1 BUGS AND IRRITATIONS
2213 There are undoubtedly serious bugs lurking somewhere in this code, if
2214 only because parts of it give the impression of understanding a great deal
2215 more about Perl than they really do.
2217 Bug reports and other feedback are most welcome.
2222 Copyright (c) 1997-2001, Damian Conway. All Rights Reserved.
2223 This module is free software. It may be used, redistributed
2224 and/or modified under the same terms as Perl itself.