4 use File::Basename qw(&basename &dirname);
8 sub link { # This is a cut-down version of installperl:link().
13 CORE::link($from, $to)
15 : ($from =~ m#^/afs/# || $to =~ m#^/afs/#)
16 ? die "AFS" # okay inside eval {}
17 : die "Couldn't link $from to $to: $!\n";
21 File::Copy::copy($from, $to)
23 : warn "Couldn't copy $from to $to: $!\n";
28 # List explicitly here the variables you want Configure to
29 # generate. Metaconfig only looks for shell variables, so you
30 # have to mention them as if they were shell variables, not
31 # %Config entries. Thus you write
33 # to ensure Configure will look for $Config{startperl}.
35 # This forces PL files to create target in same directory as PL file.
36 # This is so that make depend always knows where to find PL derivatives.
39 $file = basename($0, '.PL');
40 $file .= '.com' if $^O eq 'VMS';
42 open OUT,">$file" or die "Can't create $file: $!";
44 print "Extracting $file (with variable substitutions)\n";
46 # In this section, perl variables will be expanded during extraction.
47 # You can use $Config{...} to use Configure variables.
49 print OUT <<"!GROK!THIS!";
51 eval 'exec $Config{perlpath} -S \$0 \${1+"\$@"}'
52 if \$running_under_some_shell;
55 (\$startperl = <<'/../') =~ s/\\s*\\z//;
58 (\$perlpath = <<'/../') =~ s/\\s*\\z//;
63 # In the following, perl variables are not expanded during extraction.
65 print OUT <<'!NO!SUBS!';
67 $0 =~ s/^.*?(\w+)[\.\w]*$/$1/;
69 # (p)sed - a stream editor
70 # History: Aug 12 2000: Original version.
71 # Mar 25 2002: Rearrange generated Perl program.
79 psed - a stream editor
83 psed [-an] script [file ...]
84 psed [-an] [-e script] [-f script-file] [file ...]
86 s2p [-an] [-e script] [-f script-file]
90 A stream editor reads the input stream consisting of the specified files
91 (or standard input, if none are given), processes is line by line by
92 applying a script consisting of edit commands, and writes resulting lines
93 to standard output. The filename `C<->' may be used to read standard input.
95 The edit script is composed from arguments of B<-e> options and
96 script-files, in the given order. A single script argument may be specified
97 as the first parameter.
99 If this program is invoked with the name F<s2p>, it will act as a
100 sed-to-Perl translator. See L<"sed Script Translation">.
102 B<sed> returns an exit code of 0 on success or >0 if an error occurred.
110 A file specified as argument to the B<w> edit command is by default
111 opened before input processing starts. Using B<-a>, opening of such
112 files is delayed until the first line is actually written to the file.
114 =item B<-e> I<script>
116 The editing commands defined by I<script> are appended to the script.
117 Multiple commands must be separated by newlines.
119 =item B<-f> I<script-file>
121 Editing commands from the specified I<script-file> are read and appended
126 By default, a line is written to standard output after the editing script
127 has been applied to it. The B<-n> option suppresses automatic printing.
133 B<sed> command syntax is defined as
135 Z<> Z<> Z<> Z<>[I<address>[B<,>I<address>]][B<!>]I<function>[I<argument>]
137 with whitespace being permitted before or after addresses, and between
138 the function character and the argument. The I<address>es and the
139 address inverter (C<!>) are used to restrict the application of a
140 command to the selected line(s) of input.
142 Each command must be on a line of its own, except where noted in
145 The edit cycle performed on each input line consist of reading the line
146 (without its trailing newline character) into the I<pattern space>,
147 applying the applicable commands of the edit script, writing the final
148 contents of the pattern space and a newline to the standard output.
149 A I<hold space> is provided for saving the contents of the
150 pattern space for later use.
154 A sed address is either a line number or a pattern, which may be combined
155 arbitrarily to construct ranges. Lines are numbered across all input files.
157 Any address may be followed by an exclamation mark (`C<!>'), selecting
158 all lines not matching that address.
164 The line with the given number is selected.
168 A dollar sign (C<$>) is the line number of the last line of the input stream.
170 =item B</>I<regular expression>B</>
172 A pattern address is a basic regular expression (see
173 L<"Basic Regular Expressions">), between the delimiting character C</>.
174 Any other character except C<\> or newline may be used to delimit a
175 pattern address when the initial delimiter is prefixed with a
180 If no address is given, the command selects every line.
182 If one address is given, it selects the line (or lines) matching the
185 Two addresses select a range that begins whenever the first address
186 matches, and ends (including that line) when the second address matches.
187 If the first (second) address is a matching pattern, the second
188 address is not applied to the very same line to determine the end of
189 the range. Likewise, if the second address is a matching pattern, the
190 first address is not applied to the very same line to determine the
191 begin of another range. If both addresses are line numbers,
192 and the second line number is less than the first line number, then
193 only the first line is selected.
198 The maximum permitted number of addresses is indicated with each
199 function synopsis below.
201 The argument I<text> consists of one or more lines following the command.
202 Embedded newlines in I<text> must be preceded with a backslash. Other
203 backslashes in I<text> are deleted and the following character is taken
212 #--------------------------------------------------------------------------
213 $ComTab{'a'}=[ 1, 'txt', \&Emit, '{ push( @Q, <<'."'TheEnd' ) }\n" ]; #ok
215 =item [1addr]B<a\> I<text>
217 Write I<text> (which must start on the line following the command)
218 to standard output immediately before reading the next line
219 of input, either by executing the B<N> function or by beginning a new cycle.
223 #--------------------------------------------------------------------------
224 $ComTab{'b'}=[ 2, 'str', \&Branch, '{ goto XXX; }' ]; #ok
226 =item [2addr]B<b> [I<label>]
228 Branch to the B<:> function with the specified I<label>. If no label
229 is given, branch to the end of the script.
233 #--------------------------------------------------------------------------
234 $ComTab{'c'}=[ 2, 'txt', \&Change, <<'-X-' ]; #ok
235 { print <<'TheEnd'; } $doPrint = 0; goto EOS;
237 ### continue OK => next CYCLE;
239 =item [2addr]B<c\> I<text>
241 The line, or range of lines, selected by the address is deleted.
242 The I<text> (which must start on the line following the command)
243 is written to standard output. With an address range, this occurs at
244 the end of the range.
248 #--------------------------------------------------------------------------
249 $ComTab{'d'}=[ 2, '', \&Emit, <<'-X-' ]; #ok
254 ### continue OK => next CYCLE;
258 Deletes the pattern space and starts the next cycle.
262 #--------------------------------------------------------------------------
263 $ComTab{'D'}=[ 2, '', \&Emit, <<'-X-' ]; #ok
265 if(length($_)){ goto BOS } else { goto EOS }
268 ### continue OK => next CYCLE;
272 Deletes the pattern space through the first embedded newline or to the end.
273 If the pattern space becomes empty, a new cycle is started, otherwise
274 execution of the script is restarted.
278 #--------------------------------------------------------------------------
279 $ComTab{'g'}=[ 2, '', \&Emit, '{ $_ = $Hold };' ]; #ok
283 Replace the contents of the pattern space with the hold space.
287 #--------------------------------------------------------------------------
288 $ComTab{'G'}=[ 2, '', \&Emit, '{ $_ .= "\n"; $_ .= $Hold };' ]; #ok
292 Append a newline and the contents of the hold space to the pattern space.
296 #--------------------------------------------------------------------------
297 $ComTab{'h'}=[ 2, '', \&Emit, '{ $Hold = $_ }' ]; #ok
301 Replace the contents of the hold space with the pattern space.
305 #--------------------------------------------------------------------------
306 $ComTab{'H'}=[ 2, '', \&Emit, '{ $Hold .= "\n"; $Hold .= $_; }' ]; #ok
310 Append a newline and the contents of the pattern space to the hold space.
314 #--------------------------------------------------------------------------
315 $ComTab{'i'}=[ 1, 'txt', \&Emit, '{ print <<'."'TheEnd' }\n" ]; #ok
317 =item [1addr]B<i\> I<text>
319 Write the I<text> (which must start on the line following the command)
324 #--------------------------------------------------------------------------
325 $ComTab{'l'}=[ 2, '', \&Emit, '{ _l() }' ]; #okUTF8
329 Print the contents of the pattern space: non-printable characters are
330 shown in C-style escaped form; long lines are split and have a trailing
331 `C<\>' at the point of the split; the true end of a line is marked with
332 a `C<$>'. Escapes are: `\a', `\t', `\n', `\f', `\r', `\e' for
333 BEL, HT, LF, FF, CR, ESC, respectively, and `\' followed by a three-digit
334 octal number for all other non-printable characters.
338 #--------------------------------------------------------------------------
339 $ComTab{'n'}=[ 2, '', \&Emit, <<'-X-' ]; #ok
340 { print $_, "\n" if $doPrint;
343 last CYCLE unless getsARGV();
350 If automatic printing is enabled, write the pattern space to the standard
351 output. Replace the pattern space with the next line of input. If
352 there is no more input, processing is terminated.
356 #--------------------------------------------------------------------------
357 $ComTab{'N'}=[ 2, '', \&Emit, <<'-X-' ]; #ok
360 last CYCLE unless getsARGV( $h );
368 Append a newline and the next line of input to the pattern space. If
369 there is no more input, processing is terminated.
373 #--------------------------------------------------------------------------
374 $ComTab{'p'}=[ 2, '', \&Emit, '{ print $_, "\n"; }' ]; #ok
378 Print the pattern space to the standard output. (Use the B<-n> option
379 to suppress automatic printing at the end of a cycle if you want to
380 avoid double printing of lines.)
384 #--------------------------------------------------------------------------
385 $ComTab{'P'}=[ 2, '', \&Emit, <<'-X-' ]; #ok
386 { if( /^(.*)/ ){ print $1, "\n"; } }
391 Prints the pattern space through the first embedded newline or to the end.
395 #--------------------------------------------------------------------------
396 $ComTab{'q'}=[ 1, '', \&Emit, <<'-X-' ]; #ok
397 { print $_, "\n" if $doPrint;
404 Branch to the end of the script and quit without starting a new cycle.
408 #--------------------------------------------------------------------------
409 $ComTab{'r'}=[ 1, 'str', \&Emit, "{ _r( '-X-' ) }" ]; #ok
411 =item [1addr]B<r> I<file>
413 Copy the contents of the I<file> to standard output immediately before
414 the next attempt to read a line of input. Any error encountered while
415 reading I<file> is silently ignored.
419 #--------------------------------------------------------------------------
420 $ComTab{'s'}=[ 2, 'sub', \&Emit, '' ]; #ok
422 =item [2addr]B<s/>I<regular expression>B</>I<replacement>B</>I<flags>
424 Substitute the I<replacement> string for the first substring in
425 the pattern space that matches the I<regular expression>.
426 Any character other than backslash or newline can be used instead of a
427 slash to delimit the regular expression and the replacement.
428 To use the delimiter as a literal character within the regular expression
429 and the replacement, precede the character by a backslash (`C<\>').
431 Literal newlines may be embedded in the replacement string by
432 preceding a newline with a backslash.
434 Within the replacement, an ampersand (`C<&>') is replaced by the string
435 matching the regular expression. The strings `C<\1>' through `C<\9>' are
436 replaced by the corresponding subpattern (see L<"Basic Regular Expressions">).
437 To get a literal `C<&>' or `C<\>' in the replacement text, precede it
440 The following I<flags> modify the behaviour of the B<s> command:
446 The replacement is performed for all matching, non-overlapping substrings
447 of the pattern space.
451 Replace only the n-th matching substring of the pattern space.
455 If the substitution was made, print the new value of the pattern space.
459 If the substitution was made, write the new value of the pattern space
460 to the specified file.
466 #--------------------------------------------------------------------------
467 $ComTab{'t'}=[ 2, 'str', \&Branch, '{ goto XXX if _t() }' ]; #ok
469 =item [2addr]B<t> [I<label>]
471 Branch to the B<:> function with the specified I<label> if any B<s>
472 substitutions have been made since the most recent reading of an input line
473 or execution of a B<t> function. If no label is given, branch to the end of
479 #--------------------------------------------------------------------------
480 $ComTab{'w'}=[ 2, 'str', \&Write, "{ _w( '-X-' ) }" ]; #ok
482 =item [2addr]B<w> I<file>
484 The contents of the pattern space are written to the I<file>.
488 #--------------------------------------------------------------------------
489 $ComTab{'x'}=[ 2, '', \&Emit, '{ ($Hold, $_) = ($_, $Hold) }' ]; #ok
493 Swap the contents of the pattern space and the hold space.
497 #--------------------------------------------------------------------------
498 $ComTab{'y'}=[ 2, 'tra', \&Emit, '' ]; #ok
499 =item [2addr]B<y>B</>I<string1>B</>I<string2>B</>
501 In the pattern space, replace all characters occuring in I<string1> by the
502 character at the corresponding position in I<string2>. It is possible
503 to use any character (other than a backslash or newline) instead of a
504 slash to delimit the strings. Within I<string1> and I<string2>, a
505 backslash followed by any character other than a newline is that literal
506 character, and a backslash followed by an `n' is replaced by a newline
511 #--------------------------------------------------------------------------
512 $ComTab{'='}=[ 1, '', \&Emit, '{ print "$.\n" }' ]; #ok
516 Prints the current line number on the standard output.
520 #--------------------------------------------------------------------------
521 $ComTab{':'}=[ 0, 'str', \&Label, '' ]; #ok
523 =item [0addr]B<:> [I<label>]
525 The command specifies the position of the I<label>. It has no other effect.
529 #--------------------------------------------------------------------------
530 $ComTab{'{'}=[ 2, '', \&BeginBlock, '{' ]; #ok
531 $ComTab{'}'}=[ 0, '', \&EndBlock, ';}' ]; #ok
532 # ';' to avoid warning on empty {}-block
534 =item [2addr]B<{> [I<command>]
538 These two commands begin and end a command list. The first command may
539 be given on the same line as the opening B<{> command. The commands
540 within the list are jointly selected by the address(es) given on the
541 B<{> command (but may still have individual addresses).
545 #--------------------------------------------------------------------------
546 $ComTab{'#'}=[ 0, 'str', \&Comment, '' ]; #ok
548 =item [0addr]B<#> [I<comment>]
550 The entire line is ignored (treated as a comment). If, however, the first
551 two characters in the script are `C<#n>', automatic printing of output is
552 suppressed, as if the B<-n> option were given on the command line.
558 use vars qw{ $isEOF $Hold %wFiles @Q $CondReg $doPrint };
560 my $useDEBUG = exists( $ENV{PSEDDEBUG} );
561 my $useEXTBRE = $ENV{PSEDEXTBRE} || '';
562 $useEXTBRE =~ s/[^<>wWyB]//g; # gawk RE's handle these
564 my $doAutoPrint = 1; # automatic printing of pattern space (-n => 0)
565 my $doOpenWrite = 1; # open w command output files at start (-a => 0)
566 my $svOpenWrite = 0; # save $doOpenWrite
567 my $doGenerate = $0 eq 's2p';
569 # Collected and compiled script
571 my( @Commands, %Defined, @BlockStack, %Label, $labNum, $Code, $Func );
582 my( $msg, $loc ) = @_;
584 $loc .= ': ' if length( $loc );
585 warn( "$0: $loc$msg\n" );
590 return 'L_'.++$labNum;
593 # safeHere: create safe here delimiter and modify opcode and argument
596 my( $codref, $argref ) = @_;
598 while( $$argref =~ /^$eod$/m ){
601 $$codref =~ s/TheEnd/$eod/e;
602 $$argref .= "$eod\n";
605 # Emit: create address logic and emit command
608 my( $addr1, $addr2, $negated, $opcode, $arg, $fl ) = @_;
610 if( defined( $addr1 ) ){
611 if( defined( $addr2 ) ){
612 $addr1 .= $addr2 =~ /^\d+$/ ? "..$addr2" : "...$addr2";
614 $addr1 .= ' == $.' if $addr1 =~ /^\d+$/;
616 $cond = $negated ? "unless( $addr1 )\n" : "if( $addr1 )\n";
620 $Code .= "$cond$arg\n";
622 } elsif( $opcode =~ s/-X-/$arg/e ){
623 $Code .= "$cond$opcode\n";
625 } elsif( $opcode =~ /TheEnd/ ){
626 safeHere( \$opcode, \$arg );
627 $Code .= "$cond$opcode$arg";
630 $Code .= "$cond$opcode\n";
635 # Write (w command, w flag): store pathname
638 my( $addr1, $addr2, $negated, $opcode, $path, $fl ) = @_;
640 Emit( $addr1, $addr2, $negated, $opcode, $path, $fl );
644 # Label (: command): label definition
647 my( $addr1, $addr2, $negated, $opcode, $lab, $fl ) = @_;
650 if( length( $lab ) ){
652 if( ! exists( $Label{$lab} ) ){
653 $h = $Label{$lab}{name} = newLabel();
655 $h = $Label{$lab}{name};
656 if( exists( $Label{$lab}{defined} ) ){
657 my $dl = $Label{$lab}{defined};
658 Warn( "duplicate label $lab (first defined at $dl)", $fl );
662 $Label{$lab}{defined} = $fl;
668 # BeginBlock ({ command): push block start
670 sub BeginBlock($$$$$$){
671 my( $addr1, $addr2, $negated, $opcode, $arg, $fl ) = @_;
672 push( @BlockStack, [ $fl, $addr1, $addr2, $negated ] );
673 Emit( $addr1, $addr2, $negated, $opcode, $arg, $fl );
676 # EndBlock (} command): check proper nesting
678 sub EndBlock($$$$$$){
679 my( $addr1, $addr2, $negated, $opcode, $arg, $fl ) = @_;
681 my $jcom = pop( @BlockStack );
682 if( defined( $jcom ) ){
683 $rc = Emit( $addr1, $addr2, $negated, $opcode, $arg, $fl );
685 Warn( "unexpected `}'", $fl );
691 # Branch (t, b commands): check or create label, substitute default
694 my( $addr1, $addr2, $negated, $opcode, $lab, $fl ) = @_;
695 $lab =~ s/\s+//; # no spaces at end
697 if( length( $lab ) ){
698 if( ! exists( $Label{$lab} ) ){
699 $h = $Label{$lab}{name} = newLabel();
701 $h = $Label{$lab}{name};
703 push( @{$Label{$lab}{used}}, $fl );
707 $opcode =~ s/XXX/$h/e;
708 Emit( $addr1, $addr2, $negated, $opcode, '', $fl );
711 # Change (c command): is special due to range end watching
714 my( $addr1, $addr2, $negated, $opcode, $arg, $fl ) = @_;
715 my $kwd = $negated ? 'unless' : 'if';
716 if( defined( $addr2 ) ){
717 $addr1 .= $addr2 =~ /^\d+$/ ? "..$addr2" : "...$addr2";
719 $addr1 = '$icnt = ('.$addr1.')';
720 $opcode = 'if( $icnt =~ /E0$/ )' . $opcode;
723 $addr1 .= ' == $.' if $addr1 =~ /^\d+$/;
725 safeHere( \$opcode, \$arg );
726 $Code .= "$kwd( $addr1 ){\n $opcode$arg}\n";
731 # Comment (# command): A no-op. Who would've thought that!
734 my( $addr1, $addr2, $negated, $opcode, $arg, $fl ) = @_;
735 ### $Code .= "# $arg\n";
741 my( $del, $sref ) = @_;
743 print "stripRegex:$del:$$sref:\n" if $useDEBUG;
744 while( $$sref =~ s{^(.*?)(\\*)\Q$del\E(\s*)}{}s ){
746 $regex .= $1.$sl.$del;
747 if( length( $sl ) % 2 == 0 ){
755 # stripTrans: take a <del> terminated string from y command
756 # honoring and cleaning up of \-escaped <del>'s
759 my( $del, $sref ) = @_;
761 print "stripTrans:$del:$$sref:\n" if $useDEBUG;
762 while( $$sref =~ s{^(.*?)(\\*)\Q$del\E}{}s ){
765 if( length( $sl ) % 2 == 0 ){
776 # makey - construct Perl y/// from sed y///
779 my( $fr, $to, $fl ) = @_;
782 # Ensure that any '-' is up front.
783 # Diagnose duplicate contradicting mappings
785 for( my $i = 0; $i < length($fr); $i++ ){
786 my $fc = substr($fr,$i,1);
787 my $tc = substr($to,$i,1);
788 if( exists( $tr{$fc} ) && $tr{$fc} ne $tc ){
789 Warn( "ambiguous translation for character `$fc' in `y' command",
796 if( exists( $tr{'-'} ) ){
797 ( $fr, $to ) = ( '-', $tr{'-'} );
802 # might just as well sort it...
803 for my $fc ( sort keys( %tr ) ){
807 # make embedded delimiters and newlines safe
808 $fr =~ s/([{}])/\$1/g;
809 $to =~ s/([{}])/\$1/g;
812 return $error ? undef() : "{ y{$fr}{$to}; }";
816 # makes - construct Perl s/// from sed s///
819 my( $regex, $subst, $path, $global, $print, $nmatch, $fl ) = @_;
821 # make embedded newlines safe
822 $regex =~ s/\n/\\n/g;
823 $subst =~ s/\n/\\n/g;
828 if( length( $nmatch ) ){
831 while( --\$n && ( \$s = m ${regex}g ) ){}
832 \$s = ( substr( \$_, pos() ) =~ s ${regex}${subst}s ) if \$s;
837 { \$s = s ${regex}${subst}s${global};
842 $code .= ' print $_, "\n" if $s;'."\n";
844 if( defined( $path ) ){
846 $code .= " _w( '$path' ) if \$s;\n";
852 =head1 BASIC REGULAR EXPRESSIONS
854 A I<Basic Regular Expression> (BRE), as defined in POSIX 1003.2, consists
855 of I<atoms>, for matching parts of a string, and I<bounds>, specifying
856 repetitions of a preceding atom.
860 The possible atoms of a BRE are: B<.>, matching any single character;
861 B<^> and B<$>, matching the null string at the beginning or end
862 of a string, respectively; a I<bracket expressions>, enclosed
863 in B<[> and B<]> (see below); and any single character with no
864 other significance (matching that character). A B<\> before one
865 of: B<.>, B<^>, B<$>, B<[>, B<*>, B<\>, matching the character
866 after the backslash. A sequence of atoms enclosed in B<\(> and B<\)>
867 becomes an atom and establishes the target for a I<backreference>,
868 consisting of the substring that actually matches the enclosed atoms.
869 Finally, B<\> followed by one of the digits B<0> through B<9> is a
872 A B<^> that is not first, or a B<$> that is not last does not have
873 a special significance and need not be preceded by a backslash to
874 become literal. The same is true for a B<]>, that does not terminate
875 a bracket expression.
877 An unescaped backslash cannot be last in a BRE.
881 The BRE bounds are: B<*>, specifying 0 or more matches of the preceding
882 atom; B<\{>I<count>B<\}>, specifying that many repetitions;
883 B<\{>I<minimum>B<,\}>, giving a lower limit; and
884 B<\{>I<minimum>B<,>I<maximum>B<\}> finally defines a lower and upper
887 A bound appearing as the first item in a BRE is taken literally.
889 =head2 Bracket Expressions
891 A I<bracket expression> is a list of characters, character ranges
892 and character classes enclosed in B<[> and B<]> and matches any
893 single character from the represented set of characters.
895 A character range is written as two characters separated by B<-> and
896 represents all characters (according to the character collating sequence)
897 that are not less than the first and not greater than the second.
898 (Ranges are very collating-sequence-dependent, and portable programs
899 should avoid relying on them.)
901 A character class is one of the class names
908 enclosed in B<[:> and B<:]> and represents the set of characters
909 as defined in ctype(3).
911 If the first character after B<[> is B<^>, the sense of matching is
914 To include a literal `C<^>', place it anywhere else but first. To
915 include a literal 'C<]>' place it first or immediately after an
916 initial B<^>. To include a literal `C<->' make it the first (or
917 second after B<^>) or last character, or the second endpoint of
920 The special bracket expression constructs C<[[:E<lt>:]]> and C<[[:E<gt>:]]>
921 match the null string at the beginning and end of a word respectively.
922 (Note that neither is identical to Perl's `\b' atom.)
924 =head2 Additional Atoms
926 Since some sed implementations provide additional regular expression
927 atoms (not defined in POSIX 1003.2), B<psed> is capable of translating
928 the following backslash escapes:
932 =item B<\E<lt>> This is the same as C<[[:E<gt>:]]>.
934 =item B<\E<gt>> This is the same as C<[[:E<lt>:]]>.
936 =item B<\w> This is an abbreviation for C<[[:alnum:]_]>.
938 =item B<\W> This is an abbreviation for C<[^[:alnum:]_]>.
940 =item B<\y> Match the empty string at a word boundary.
942 =item B<\B> Match the empty string between any two either word or non-word characters.
946 To enable this feature, the environment variable PSEDEXTBRE must be set
947 to a string containing the requested characters, e.g.:
948 C<PSEDEXTBRE='E<lt>E<gt>wW'>.
953 # bre2p - convert BRE to Perl RE
956 my( $pref, $ic ) = @_;
957 $ic < length($$pref)-1 ? substr( $$pref, $ic+1, 1 ) : '';
961 my( $del, $pat, $fl ) = @_;
963 $led =~ tr/{([</})]>/;
964 $led = '' if $led eq $del;
966 $pat = substr( $pat, 1, length($pat) - 2 );
971 for( my $ic = 0; $ic < length( $pat ); $ic++ ){
972 my $c = substr( $pat, $ic, 1 );
974 ### backslash escapes
975 my $nc = peek($pat,$ic);
977 Warn( "`\\' cannot be last in pattern", $fl );
981 if( $nc eq $del ){ ## \<pattern del> => \<pattern del>
984 } elsif( $nc =~ /([[.*\\n])/ ){
985 ## check for \-escaped magics and \n:
986 ## \[ \. \* \\ \n stay as they are
989 } elsif( $nc eq '(' ){ ## \( => (
993 } elsif( $nc eq ')' ){ ## \) => )
997 Warn( "unmatched `\\)'", $fl );
1002 } elsif( $nc eq '{' ){ ## repetition factor \{<i>[,[<j>]]\}
1003 my $endpos = index( $pat, '\\}', $ic );
1005 Warn( "unmatched `\\{'", $fl );
1008 my $rep = substr( $pat, $ic+1, $endpos-($ic+1) );
1011 if( $res =~ /^\^?$/ ){
1012 $res .= "\\{$rep\}";
1013 } elsif( $rep =~ /^(\d+)(,?)(\d*)?$/ ){
1017 if( length( $max ) ){
1019 Warn( "maximum less than minimum in `\\{$rep\\}'",
1027 if( $min == 0 && $max eq '1' ){
1029 } elsif( $min == 1 && "$com$max" eq ',' ){
1031 } elsif( $min == 0 && "$com$max" eq ',' ){
1034 $res .= "{$min$com$max}";
1037 Warn( "invalid repeat clause `\\{$rep\\}'", $fl );
1041 } elsif( $nc =~ /^[1-9]$/ ){
1042 ## \1 .. \9 => \1 .. \9, but check for a following digit
1043 if( $nc > $backref ){
1044 Warn( "invalid backreference ($nc)", $fl );
1048 if( peek($pat,$ic) =~ /[0-9]/ ){
1052 } elsif( $useEXTBRE && ( $nc =~ /[$useEXTBRE]/ ) ){
1053 ## extensions - at most <>wWyB - not in POSIX
1054 if( $nc eq '<' ){ ## \< => \b(?=\w), be precise
1055 $res .= '\\b(?<=\\W)';
1056 } elsif( $nc eq '>' ){ ## \> => \b(?=\W), be precise
1057 $res .= '\\b(?=\\W)';
1058 } elsif( $nc eq 'y' ){ ## \y => \b
1060 } else { ## \B, \w, \W remain the same
1063 } elsif( $nc eq $led ){
1064 ## \<closing bracketing-delimiter> - keep '\'
1067 } else { ## \ <char> => <char> ("as if `\' were not present")
1071 } elsif( $c eq '.' ){ ## . => .
1074 } elsif( $c eq '*' ){ ## * => * but \* if there's nothing preceding it
1075 if( $res =~ /^\^?$/ ){
1077 } elsif( substr( $res, -1, 1 ) ne '*' ){
1081 } elsif( $c eq '[' ){
1082 ## parse []: [^...] [^]...] [-...]
1084 if( peek($pat,$ic) eq '^' ){
1088 my $nc = peek($pat,$ic);
1089 if( $nc eq ']' || $nc eq '-' ){
1093 # check that [ is not trailing
1094 if( $ic >= length( $pat ) - 1 ){
1095 Warn( "unmatched `['", $fl );
1098 # look for [:...:] and x-y
1099 my $rstr = substr( $pat, $ic+1 );
1100 if( $rstr =~ /^((?:\[:\(\w+|[><]\):\]|[^]-](?:-[^]])?)*)/ ){
1102 $ic += length( $cnt );
1103 $cnt =~ s/([\\\$])/\\$1/g; # `\', `$' are magic in Perl []
1104 # try some simplifications
1106 if( $red =~ s/0-9// ){
1108 if( $red =~ s/A-Z// && $red =~ s/a-z// && $red =~ s/_// ){
1114 # POSIX 1003.2 has this (optional) for begin/end word
1115 $add = '\\b(?=\\W)' if $add eq '[[:<:]]';
1116 $add = '\\b(?<=\\W)' if $add eq '[[:>:]]';
1120 ## may have a trailing `-' before `]'
1121 if( $ic < length($pat) - 1 &&
1122 substr( $pat, $ic+1 ) =~ /^(-?])/ ){
1123 $ic += length( $1 );
1125 # another simplification
1126 $add =~ s/^\[(\^?)(\\[dw])]$/ $1 eq '^' ? uc($2) : $2 /e;
1129 Warn( "unmatched `['", $fl );
1133 } elsif( $c eq $led ){ ## unescaped <closing bracketing-delimiter>
1136 } elsif( $c eq ']' ){ ## unmatched ] is not magic
1139 } elsif( $c =~ /[|+?{}()]/ ){ ## not magic in BRE, but in Perl: \-quote
1142 } elsif( $c eq '^' ){ ## not magic unless 1st, but in Perl: \-quote
1143 $res .= length( $res ) ? '\\^' : '^';
1145 } elsif( $c eq '$' ){ ## not magic unless last, but in Perl: \-quote
1146 $res .= $ic == length( $pat ) - 1 ? '$' : '\\$';
1154 Warn( "unmatched `\\('", $fl );
1158 # final cleanup: eliminate raw HTs
1160 return $del . $res . ( $led ? $led : $del );
1165 # sub2p - convert sed substitution to Perl substitution
1168 my( $del, $subst, $fl ) = @_;
1170 $led =~ tr/{([</})]>/;
1171 $led = '' if $led eq $del;
1173 $subst = substr( $subst, 1, length($subst) - 2 );
1176 for( my $ic = 0; $ic < length( $subst ); $ic++ ){
1177 my $c = substr( $subst, $ic, 1 );
1179 ### backslash escapes
1180 my $nc = peek($subst,$ic);
1182 Warn( "`\\' cannot be last in substitution", $fl );
1186 if( $nc =~ /[\\$del$led]/ ){ ## \ and delimiter
1188 } elsif( $nc =~ /[1-9]/ ){ ## \1 - \9 => ${1} - ${9}
1189 $res .= '${' . $nc . '}';
1190 } else { ## everything else (includes &): omit \
1193 } elsif( $c eq '&' ){ ## & => $&
1195 } elsif( $c =~ /[\$\@$led]/ ){ ## magic in Perl's substitution string
1202 # final cleanup: eliminate raw HTs
1204 return ( $led ? $del : $led ) . $res . ( $led ? $led : $del );
1210 my( $pdef, $pfil, $plin );
1211 for( my $icom = 0; $icom < @Commands; $icom++ ){
1212 my $cmd = $Commands[$icom];
1213 print "Parse:$cmd:\n" if $useDEBUG;
1215 next unless length( $cmd );
1217 if( exists( $Defined{$icom} ) ){
1218 $pdef = $Defined{$icom};
1219 if( $pdef =~ /^ #(\d+)/ ){
1220 $pfil = 'expression #';
1229 my $fl = "$pfil$plin";
1231 # insert command as comment in gnerated code
1233 $Code .= "# $cmd\n" if $doGenerate;
1237 my( $negated, $naddr, $addr1, $addr2 );
1239 if( $cmd =~ s/^(\d+)\s*// ){
1240 $addr1 = "$1"; $naddr++;
1241 } elsif( $cmd =~ s/^\$\s*// ){
1242 $addr1 = 'eofARGV()'; $naddr++;
1243 } elsif( $cmd =~ s{^(/)}{} || $cmd =~ s{^\\(.)}{} ){
1245 my $regex = stripRegex( $del, \$cmd );
1246 if( defined( $regex ) ){
1247 $addr1 = 'm '.bre2p( $del, $regex, $fl ).'s';
1250 Warn( "malformed regex, 1st address", $fl );
1255 if( defined( $addr1 ) && $cmd =~ s/,\s*// ){
1256 if( $cmd =~ s/^(\d+)\s*// ){
1257 $addr2 = "$1"; $naddr++;
1258 } elsif( $cmd =~ s/^\$\s*// ){
1259 $addr2 = 'eofARGV()'; $naddr++;
1260 } elsif( $cmd =~ s{^(/)}{} || $cmd =~ s{^\\(.)}{} ){
1262 my $regex = stripRegex( $del, \$cmd );
1263 if( defined( $regex ) ){
1264 $addr2 = 'm '. bre2p( $del, $regex, $fl ).'s';
1267 Warn( "malformed regex, 2nd address", $fl );
1272 Warn( "invalid address after `,'", $fl );
1278 # address modifier `!'
1280 $negated = $cmd =~ s/^!\s*//;
1281 if( defined( $addr1 ) ){
1282 print "Parse: addr1=$addr1" if $useDEBUG;
1283 if( defined( $addr2 ) ){
1284 print ", addr2=$addr2 " if $useDEBUG;
1285 # both numeric and addr1 > addr2 => eliminate addr2
1286 undef( $addr2 ) if $addr1 =~ /^\d+$/ &&
1287 $addr2 =~ /^\d+$/ && $addr1 > $addr2;
1290 print 'negated' if $useDEBUG && $negated;
1291 print " command:$cmd\n" if $useDEBUG;
1295 if( $cmd !~ s/^([:#={}abcdDgGhHilnNpPqrstwxy])\s*// ){
1296 my $h = substr( $cmd, 0, 1 );
1297 Warn( "unknown command `$h'", $fl );
1303 my $tabref = $ComTab{$key};
1305 if( $naddr > $tabref->[0] ){
1306 Warn( "excess address(es)", $fl );
1312 if( $tabref->[1] eq 'str' ){
1313 # take remainder - don't care if it is empty
1317 } elsif( $tabref->[1] eq 'txt' ){
1319 my $goon = $cmd =~ /(.*)\\$/;
1321 Warn( "extra characters after command ($cmd)", $fl );
1326 if( $icom > $#Commands ){
1327 Warn( "unexpected end of script", $fl );
1331 $cmd = $Commands[$icom];
1332 $Code .= "# $cmd\n" if $doGenerate;
1333 $goon = $cmd =~ s/\\$//;
1334 $cmd =~ s/\\(.)/$1/g;
1335 $arg .= "\n" if length( $arg );
1338 $arg .= "\n" if length( $arg );
1341 } elsif( $tabref->[1] eq 'sub' ){
1343 if( ! length( $cmd ) ){
1344 Warn( "`s' command requires argument", $fl );
1348 if( $cmd =~ s{^([^\\\n])}{} ){
1350 my $regex = stripRegex( $del, \$cmd );
1351 if( ! defined( $regex ) ){
1352 Warn( "malformed regular expression", $fl );
1356 $regex = bre2p( $del, $regex, $fl );
1358 # a trailing \ indicates embedded NL (in replacement string)
1359 while( $cmd =~ s/(?<!\\)\\$/\n/ ){
1361 if( $icom > $#Commands ){
1362 Warn( "unexpected end of script", $fl );
1366 $cmd .= $Commands[$icom];
1367 $Code .= "# $Commands[$icom]\n" if $doGenerate;
1370 my $subst = stripRegex( $del, \$cmd );
1371 if( ! defined( $regex ) ){
1372 Warn( "malformed substitution expression", $fl );
1376 $subst = sub2p( $del, $subst, $fl );
1378 # parse s/// modifier: g|p|0-9|w <file>
1379 my( $global, $nmatch, $print, $write ) =
1380 ( '', '', 0, undef );
1381 while( $cmd =~ s/^([gp0-9])// ){
1382 $1 eq 'g' ? ( $global = 'g' ) :
1383 $1 eq 'p' ? ( $print = $1 ) : ( $nmatch .= $1 );
1385 $write = $1 if $cmd =~ s/w\s*(.*)$//;
1386 ### $nmatch =~ s/^(\d)\1*$/$1/; ### may be dangerous?
1387 if( $global && length( $nmatch ) || length( $nmatch ) > 1 ){
1388 Warn( "conflicting flags `$global$nmatch'", $fl );
1393 $arg = makes( $regex, $subst,
1394 $write, $global, $print, $nmatch, $fl );
1395 if( ! defined( $arg ) ){
1401 Warn( "improper delimiter in s command", $fl );
1406 } elsif( $tabref->[1] eq 'tra' ){
1408 # a trailing \ indicates embedded newline
1409 while( $cmd =~ s/(?<!\\)\\$/\n/ ){
1411 if( $icom > $#Commands ){
1412 Warn( "unexpected end of script", $fl );
1416 $cmd .= $Commands[$icom];
1417 $Code .= "# $Commands[$icom]\n" if $doGenerate;
1419 if( ! length( $cmd ) ){
1420 Warn( "`y' command requires argument", $fl );
1424 my $d = substr( $cmd, 0, 1 ); $cmd = substr( $cmd, 1 );
1426 Warn( "`\\' not valid as delimiter in `y' command", $fl );
1430 my $fr = stripTrans( $d, \$cmd );
1431 if( ! defined( $fr ) || ! length( $cmd ) ){
1432 Warn( "malformed `y' command argument", $fl );
1436 my $to = stripTrans( $d, \$cmd );
1437 if( ! defined( $to ) ){
1438 Warn( "malformed `y' command argument", $fl );
1442 if( length($fr) != length($to) ){
1443 Warn( "string lengths in `y' command differ", $fl );
1447 if( ! defined( $arg = makey( $fr, $to, $fl ) ) ){
1454 # $cmd must be now empty - exception is {
1455 if( $cmd !~ /^\s*$/ ){
1457 # dirty hack to process command on '{' line
1458 $Commands[$icom--] = $cmd;
1460 Warn( "extra characters after command ($cmd)", $fl );
1468 if( &{$tabref->[2]}( $addr1, $addr2, $negated,
1469 $tabref->[3], $arg, $fl ) ){
1474 while( @BlockStack ){
1475 my $bl = pop( @BlockStack );
1476 Warn( "start of unterminated `{'", $bl );
1480 for my $lab ( keys( %Label ) ){
1481 if( ! exists( $Label{$lab}{defined} ) ){
1482 for my $used ( @{$Label{$lab}{used}} ){
1483 Warn( "undefined label `$lab'", $used );
1489 exit( 1 ) if $error;
1498 print STDERR "Usage: sed [-an] command [file...]\n";
1499 print STDERR " [-an] [-e command] [-f script-file] [file...]\n";
1503 # Here we go again...
1506 while( @ARGV && $ARGV[0] =~ /^-(.)(.*)$/ ){
1511 if( length( $arg ) ){
1512 push( @Commands, split( "\n", $arg ) );
1514 push( @Commands, shift( @ARGV ) );
1516 Warn( "option -e requires an argument" );
1521 $Defined{$#Commands} = " #$expr";
1526 if( length( $arg ) ){
1529 $path = shift( @ARGV );
1531 Warn( "option -f requires an argument" );
1535 my $fst = $#Commands + 1;
1536 open( SCRIPT, "<$path" ) || die( "$0: $path: could not open ($!)\n" );
1538 while( defined( $cmd = <SCRIPT> ) ){
1540 push( @Commands, $cmd );
1543 if( $#Commands >= $fst ){
1544 $Defined{$fst} = "$path";
1548 if( $opt eq '-' && $arg eq '' ){
1551 if( $opt eq 'h' || $opt eq '?' ){
1557 } elsif( $opt eq 'a' ){
1560 Warn( "illegal option `$opt'" );
1564 if( length( $arg ) ){
1565 unshift( @ARGV, "-$arg" );
1569 # A singleton command may be the 1st argument when there are no options.
1571 if( @Commands == 0 ){
1573 Warn( "no script command given" );
1577 push( @Commands, split( "\n", shift( @ARGV ) ) );
1578 $Defined{0} = ' #1';
1581 print STDERR "Files: @ARGV\n" if $useDEBUG;
1583 # generate leading code
1585 $Func = <<'[TheEnd]';
1587 # openARGV: open 1st input file
1590 unshift( @ARGV, '-' ) unless @ARGV;
1591 my $file = shift( @ARGV );
1592 open( ARG, "<$file" )
1593 || die( "$0: can't open $file for reading ($!)\n" );
1597 # getsARGV: Read another input line into argument (default: $_).
1598 # Move on to next input file, and reset EOF flag $isEOF.
1600 my $argref = @_ ? shift() : \$_;
1601 while( $isEOF || ! defined( $$argref = <ARG> ) ){
1603 return 0 unless @ARGV;
1604 my $file = shift( @ARGV );
1605 open( ARG, "<$file" )
1606 || die( "$0: can't open $file for reading ($!)\n" );
1612 # eofARGV: end-of-file test
1615 return @ARGV == 0 && ( $isEOF = eof( ARG ) );
1618 # makeHandle: Generates another file handle for some file (given by its path)
1619 # to be written due to a w command or an s command's w flag.
1623 if( ! exists( $wFiles{$path} ) || $wFiles{$path} eq '' ){
1624 $handle = $wFiles{$path} = gensym();
1626 if( ! open( $handle, ">$path" ) ){
1627 die( "$0: can't open $path for writing: ($!)\n" );
1631 $handle = $wFiles{$path};
1636 # printQ: Print queued output which is either a string or a reference
1641 # flush open w files so that reading this file gets it all
1642 if( exists( $wFiles{$$q} ) && $wFiles{$$q} ne '' ){
1643 open( $wFiles{$$q}, ">>$$q" );
1645 # copy file to stdout: slow, but safe
1646 if( open( RF, "<$$q" ) ){
1647 while( defined( my $line = <RF> ) ){
1661 # generate the sed loop
1663 $Code .= <<'[TheEnd]';
1669 # Run: the sed loop reading input and applying the script
1672 my( $h, $icnt, $s, $n );
1673 # hack (not unbreakable :-/) to avoid // matching an empty string
1674 my $z = "\000"; $z =~ /$z/;
1679 $doPrint = $doAutoPrint;
1681 while( getsARGV() ){
1683 $CondReg = 0; # cleared on t
1687 # parse - avoid opening files when doing s2p
1689 ( $svOpenWrite, $doOpenWrite ) = ( $doOpenWrite, $svOpenWrite )
1692 ( $svOpenWrite, $doOpenWrite ) = ( $doOpenWrite, $svOpenWrite )
1695 # append trailing code
1697 $Code .= <<'[TheEnd]';
1698 EOS: if( $doPrint ){
1701 $doPrint = $doAutoPrint;
1711 # append optional functions, prepend prototypes
1713 my $Proto = "# prototypes\n";
1715 $Proto .= "sub _l();\n";
1716 $Func .= <<'[TheEnd]';
1717 # _l: l command processing
1722 # transform non printing chars into escape notation
1724 if( $h =~ /[^[:print:]]/ ){
1731 $h =~ s/([^[:print:]])/sprintf("\\%03o", ord($1))/ge;
1733 # split into lines of length $mcpl
1734 while( length( $h ) > $mcpl ){
1735 my $l = substr( $h, 0, $mcpl-1 );
1736 $h = substr( $h, $mcpl );
1737 # remove incomplete \-escape from end of line
1738 if( $l =~ s/(?<!\\)(\\[0-7]{0,2})$// ){
1750 $Proto .= "sub _r(\$);\n";
1751 $Func .= <<'[TheEnd]';
1752 # _r: r command processing: Save a reference to the pathname.
1763 $Proto .= "sub _t();\n";
1764 $Func .= <<'[TheEnd]';
1765 # _t: t command - condition register test/reset
1777 $Proto .= "sub _w(\$);\n";
1778 $Func .= <<'[TheEnd]';
1779 # _w: w command and s command's w flag - write to file
1783 my $handle = $wFiles{$path};
1784 if( ! $doOpenWrite && ! defined( fileno( $handle ) ) ){
1785 open( $handle, ">$path" )
1786 || die( "$0: $path: cannot open ($!)\n" );
1788 print $handle $_, "\n";
1794 $Code = $Proto . $Code;
1796 # magic "#n" - same as -n option
1798 $doAutoPrint = 0 if substr( $Commands[0], 0, 2 ) eq '#n';
1800 # eval code - check for errors
1802 print "Code:\n$Code$Func" if $useDEBUG;
1805 print "Code:\n$Code$Func";
1806 die( "$0: internal error - generated incorrect Perl code: $@\n" );
1811 # write full Perl program
1814 # bang line, declarations, prototypes
1817 eval 'exec $perlpath -S \$0 \${1+"\$@"}'
1819 \$0 =~ s/^.*?(\\w+)\[\\.\\w+\]*\$/\$1/;
1823 use vars qw{ \$isEOF \$Hold \%wFiles \@Q \$CondReg
1824 \$doAutoPrint \$doOpenWrite \$doPrint };
1825 \$doAutoPrint = $doAutoPrint;
1826 \$doOpenWrite = $doOpenWrite;
1829 my $wf = "'" . join( "', '", keys( %wFiles ) ) . "'";
1834 exit( 1 ) unless makeHandle( \$p );
1846 # execute: make handles (and optionally open) all w files; run!
1847 for my $p ( keys( %wFiles ) ){
1848 exit( 1 ) unless makeHandle( $p );
1856 The environment variable C<PSEDEXTBRE> may be set to extend BREs.
1857 See L<"Additional Atoms">.
1863 =item ambiguous translation for character `%s' in `y' command
1865 The indicated character appears twice, with different translations.
1867 =item `[' cannot be last in pattern
1869 A `[' in a BRE indicates the beginning of a I<bracket expression>.
1871 =item `\' cannot be last in pattern
1873 A `\' in a BRE is used to make the subsequent character literal.
1875 =item `\' cannot be last in substitution
1877 A `\' in a subsitution string is used to make the subsequent character literal.
1879 =item conflicting flags `%s'
1881 In an B<s> command, either the `g' flag and an n-th occurrence flag, or
1882 multiple n-th occurrence flags are specified. Note that only the digits
1883 `1' through `9' are permitted.
1885 =item duplicate label %s (first defined at %s)
1887 =item excess address(es)
1889 The command has more than the permitted number of addresses.
1891 =item extra characters after command (%s)
1893 =item illegal option `%s'
1895 =item improper delimiter in s command
1897 The BRE and substitution may not be delimited with `\' or newline.
1899 =item invalid address after `,'
1901 =item invalid backreference (%s)
1903 The specified backreference number exceeds the number of backreferences
1906 =item invalid repeat clause `\{%s\}'
1908 The repeat clause does not contain a valid integer value, or pair of
1911 =item malformed regex, 1st address
1913 =item malformed regex, 2nd address
1915 =item malformed regular expression
1917 =item malformed substitution expression
1919 =item malformed `y' command argument
1921 The first or second string of a B<y> command is syntactically incorrect.
1923 =item maximum less than minimum in `\{%s\}'
1925 =item no script command given
1927 There must be at least one B<-e> or one B<-f> option specifying a
1928 script or script file.
1930 =item `\' not valid as delimiter in `y' command
1932 =item option -e requires an argument
1934 =item option -f requires an argument
1936 =item `s' command requires argument
1938 =item start of unterminated `{'
1940 =item string lengths in `y' command differ
1942 The translation table strings in a B<y> commanf must have equal lengths.
1944 =item undefined label `%s'
1946 =item unexpected `}'
1948 A B<}> command without a preceding B<{> command was encountered.
1950 =item unexpected end of script
1952 The end of the script was reached although a text line after a
1953 B<a>, B<c> or B<i> command indicated another line.
1955 =item unknown command `%s'
1957 =item unterminated `['
1959 A BRE contains an unterminated bracket expression.
1961 =item unterminated `\('
1963 A BRE contains an unterminated backreference.
1965 =item `\{' without closing `\}'
1967 A BRE contains an unterminated bounds specification.
1969 =item `\)' without preceding `\('
1971 =item `y' command requires argument
1977 The basic material for the preceding section was generated by running
1981 s/^.*Warn( *"\([^"]*\)".*$/\1/
1986 s/$[_[:alnum:]]\{1,\}/%s/g
1991 on the program's own text, and piping the output into C<sort -u>.
1994 =head1 SED SCRIPT TRANSLATION
1996 If this program is invoked with the name F<s2p> it will act as a
1997 sed-to-Perl translator. After option processing (all other
1998 arguments are ignored), a Perl program is printed on standard
1999 output, which will process the input stream (as read from all
2000 arguments) in the way defined by the sed script and the option setting
2001 used for the translation.
2005 perl(1), re_format(7)
2009 The B<l> command will show escape characters (ESC) as `C<\e>', but
2010 a vertical tab (VT) in octal.
2012 Trailing spaces are truncated from labels in B<:>, B<t> and B<b> commands.
2014 The meaning of an empty regular expression (`C<//>'), as defined by B<sed>,
2015 is "the last pattern used, at run time". This deviates from the Perl
2016 interpretation, which will re-use the "last last successfully executed
2017 regular expression". Since keeping track of pattern usage would create
2018 terribly cluttered code, and differences would only appear in obscure
2019 context (where other B<sed> implementations appear to deviate, too),
2020 the Perl semantics was adopted. Note that common usage of this feature,
2021 such as in C</abc/s//xyz/>, will work as expected.
2023 Collating elements (of bracket expressions in BREs) are not implemented.
2027 This B<sed> implementation conforms to the IEEE Std1003.2-1992 ("POSIX.2")
2028 definition of B<sed>, and is compatible with the I<OpenBSD>
2029 implementation, except where otherwise noted (see L<"BUGS">).
2033 This Perl implementation of I<sed> was written by Wolfgang Laun,
2034 I<Wolfgang.Laun@alcatel.at>.
2036 =head1 COPYRIGHT and LICENSE
2038 This program is free and open software. You may use, modify,
2039 distribute, and sell this program (and any modified variants) in any
2040 way you wish, provided you do not restrict others from doing the same.
2046 close OUT or die "Can't close $file: $!";
2047 chmod 0755, $file or die "Can't reset permissions for $file: $!\n";
2049 print "Linking s2p to psed.\n";
2050 if (defined $Config{d_link}) {
2053 unshift @INC, '../lib';
2055 File::Copy::syscopy('s2p', 'psed');
2057 exec("$Config{'eunicefix'} $file") if $Config{'eunicefix'} ne ':';