[p5sagit/p5-mst-13.2.git] / lib / Pod / Text.pm

package Pod::Text;

=head1 NAME

Pod::Text - convert POD data to formatted ASCII text

=head1 SYNOPSIS

	use Pod::Text;

	pod2text("perlfunc.pod");

Also:

	pod2text [B<-a>] [B<->I<width>] < input.pod

=head1 DESCRIPTION

Pod::Text is a module that can convert documentation in the POD format (such
as can be found throughout the Perl distribution) into formatted ASCII.
Termcap is optionally supported for boldface/underline, and can enabled via
C<$Pod::Text::termcap=1>. If termcap has not been enabled, then backspaces
will be used to simulate bold and underlined text.

A separate F<pod2text> program is included that is primarily a wrapper for
Pod::Text.

The single function C<pod2text()> can take the optional options B<-a>
for an alternative output format, then a B<->I<width> option with the
max terminal width, followed by one or two arguments. The first
should be the name of a file to read the pod from, or "E<lt>&STDIN" to read from
STDIN. A second argument, if provided, should be a filehandle glob where
output should be sent.

=head1 AUTHOR

Tom Christiansen E<lt>F<tchrist@mox.perl.com>E<gt>

=head1 TODO

Cleanup work. The input and output locations need to be more flexible,
termcap shouldn't be a global variable, and the terminal speed needs to
be properly calculated.

=cut

use Term::Cap;
require Exporter;
@ISA = Exporter;
@EXPORT = qw(pod2text);

use vars qw($VERSION);
$VERSION = "1.0203";

$termcap=0;

$opt_alt_format = 0;

#$use_format=1;

$UNDL = "\x1b[4m";
$INV = "\x1b[7m";
$BOLD = "\x1b[1m";
$NORM = "\x1b[0m";

sub pod2text {
shift if $opt_alt_format = ($_[0] eq '-a');

if($termcap and !$setuptermcap) {
	$setuptermcap=1;

    my($term) = Tgetent Term::Cap { TERM => undef, OSPEED => 9600 };
    $UNDL = $term->{'_us'};
    $INV = $term->{'_mr'};
    $BOLD = $term->{'_md'};
    $NORM = $term->{'_me'};
}

$SCREEN = ($_[0] =~ /^-(\d+)/ && (shift, $1))
       ||  $ENV{COLUMNS}
       || ($ENV{TERMCAP} =~ /co#(\d+)/)[0]
       || ($^O ne 'MSWin32' && (`stty -a 2>/dev/null` =~ /(\d+) columns/)[0])
       || 72;

@_ = ("<&STDIN") unless @_;
local($file,*OUTPUT) = @_;
*OUTPUT = *STDOUT if @_<2;

local $: = $:;
$: = " \n" if $opt_alt_format;	# Do not break ``-L/lib/'' into ``- L/lib/''.

$/ = "";

$FANCY = 0;

$cutting = 1;
$DEF_INDENT = 4;
$indent = $DEF_INDENT;
$needspace = 0;
$begun = "";

open(IN, $file) || die "Couldn't open $file: $!";

POD_DIRECTIVE: while (<IN>) {
    if ($cutting) {
	next unless /^=/;
	$cutting = 0;
    }
    if ($begun) {
        if (/^=end\s+$begun/) {
             $begun = "";
        }
        elsif ($begun eq "text") {
            print OUTPUT $_;
        }
        next;
    }
    1 while s{^(.*?)(\t+)(.*)$}{
	$1
	. (' ' x (length($2) * 8 - length($1) % 8))
	. $3
    }me;
    # Translate verbatim paragraph
    if (/^\s/) {
	output($_);
	next;
    }

    if (/^=for\s+(\S+)\s*(.*)/s) {
        if ($1 eq "text") {
            print OUTPUT $2,"";
        } else {
            # ignore unknown for
        }
        next;
    }
    elsif (/^=begin\s+(\S+)\s*(.*)/s) {
        $begun = $1;
        if ($1 eq "text") {
            print OUTPUT $2."";
        }
        next;
    }

sub prepare_for_output {

    s/\s*$/\n/;
    &init_noremap;

    # need to hide E<> first; they're processed in clear_noremap
    s/(E<[^<>]+>)/noremap($1)/ge;
    $maxnest = 10;
    while ($maxnest-- && /[A-Z]</) {
	unless ($FANCY) {
	    if ($opt_alt_format) {
		s/[BC]<(.*?)>/``$1''/sg;
		s/F<(.*?)>/"$1"/sg;
	    } else {
		s/C<(.*?)>/`$1'/sg;
	    }
	} else {
	    s/C<(.*?)>/noremap("E<lchevron>${1}E<rchevron>")/sge;
	}
        # s/[IF]<(.*?)>/italic($1)/ge;
        s/I<(.*?)>/*$1*/sg;
        # s/[CB]<(.*?)>/bold($1)/ge;
	s/X<.*?>//sg;
	# LREF: a manpage(3f)
	s:L<([a-zA-Z][^\s\/]+)(\([^\)]+\))?>:the $1$2 manpage:g;
	# LREF: an =item on another manpage
	s{
	    L<
		([^/]+)
		/
		(
		    [:\w]+
		    (\(\))?
		)
	    >
	} {the "$2" entry in the $1 manpage}gx;

	# LREF: an =item on this manpage
	s{
	   ((?:
	    L<
		/
		(
		    [:\w]+
		    (\(\))?
		)
	    >
	    (,?\s+(and\s+)?)?
	  )+)
	} { internal_lrefs($1) }gex;

	# LREF: a =head2 (head1?), maybe on a manpage, maybe right here
	# the "func" can disambiguate
	s{
	    L<
		(?:
		    ([a-zA-Z]\S+?) / 
		)?
		"?(.*?)"?
	    >
	}{
	    do {
		$1 	# if no $1, assume it means on this page.
		    ?  "the section on \"$2\" in the $1 manpage"
		    :  "the section on \"$2\""
	    }
	}sgex;

        s/[A-Z]<(.*?)>/$1/sg;
    }
    clear_noremap(1);
}

    &prepare_for_output;

    if (s/^=//) {
	# $needspace = 0;		# Assume this.
	# s/\n/ /g;
	($Cmd, $_) = split(' ', $_, 2);
	# clear_noremap(1);
	if ($Cmd eq 'cut') {
	    $cutting = 1;
	}
	elsif ($Cmd eq 'pod') {
	    $cutting = 0;
	}
	elsif ($Cmd eq 'head1') {
	    makespace();
	    if ($opt_alt_format) {
		print OUTPUT "\n";
		s/^(.+?)[ \t]*$/==== $1 ====/;
	    }
	    print OUTPUT;
	    # print OUTPUT uc($_);
	    $needspace = $opt_alt_format;
	}
	elsif ($Cmd eq 'head2') {
	    makespace();
	    # s/(\w+)/\u\L$1/g;
	    #print ' ' x $DEF_INDENT, $_;
	    # print "\xA7";
	    s/(\w)/\xA7 $1/ if $FANCY;
	    if ($opt_alt_format) {
		s/^(.+?)[ \t]*$/==   $1   ==/;
		print OUTPUT "\n", $_;
	    } else {
		print OUTPUT ' ' x ($DEF_INDENT/2), $_, "\n";
	    }
	    $needspace = $opt_alt_format;
	}
	elsif ($Cmd eq 'over') {
	    push(@indent,$indent);
	    $indent += ($_ + 0) || $DEF_INDENT;
	}
	elsif ($Cmd eq 'back') {
	    $indent = pop(@indent);
	    warn "Unmatched =back\n" unless defined $indent;
	}
	elsif ($Cmd eq 'item') {
	    makespace();
	    # s/\A(\s*)\*/$1\xb7/ if $FANCY;
	    # s/^(\s*\*\s+)/$1 /;
	    {
		if (length() + 3 < $indent) {
		    my $paratag = $_;
		    $_ = <IN>;
		    if (/^=/) {  # tricked!
			local($indent) = $indent[$#index - 1] || $DEF_INDENT;
			output($paratag);
			redo POD_DIRECTIVE;
		    }
		    &prepare_for_output;
		    IP_output($paratag, $_);
		} else {
		    local($indent) = $indent[$#index - 1] || $DEF_INDENT;
		    output($_, 0);
		}
	    }
	}
	else {
	    warn "Unrecognized directive: $Cmd\n";
	}
    }
    else {
	# clear_noremap(1);
	makespace();
	output($_, 1);
    }
}

close(IN);

}

#########################################################################

sub makespace {
    if ($needspace) {
	print OUTPUT "\n";
	$needspace = 0;
    }
}

sub bold {
    my $line = shift;
    return $line if $use_format;
    if($termcap) {
    	$line = "$BOLD$line$NORM";
    } else {
	    $line =~ s/(.)/$1\b$1/g;
	}
#    $line = "$BOLD$line$NORM" if $ansify;
    return $line;
}

sub italic {
    my $line = shift;
    return $line if $use_format;
    if($termcap) {
    	$line = "$UNDL$line$NORM";
    } else {
	    $line =~ s/(.)/$1\b_/g;
    }
#    $line = "$UNDL$line$NORM" if $ansify;
    return $line;
}

# Fill a paragraph including underlined and overstricken chars.
# It's not perfect for words longer than the margin, and it's probably
# slow, but it works.
sub fill {
    local $_ = shift;
    my $par = "";
    my $indent_space = " " x $indent;
    my $marg = $SCREEN-$indent;
    my $line = $indent_space;
    my $line_length;
    foreach (split) {
	my $word_length = length;
	$word_length -= 2 while /\010/g;  # Subtract backspaces

	if ($line_length + $word_length > $marg) {
	    $par .= $line . "\n";
	    $line= $indent_space . $_;
	    $line_length = $word_length;
	}
	else {
	    if ($line_length) {
		$line_length++;
		$line .= " ";
	    }
	    $line_length += $word_length;
	    $line .= $_;
	}
    }
    $par .= "$line\n" if $line;
    $par .= "\n";
    return $par;
}

sub IP_output {
    local($tag, $_) = @_;
    local($tag_indent) = $indent[$#index - 1] || $DEF_INDENT;
    $tag_cols = $SCREEN - $tag_indent;
    $cols = $SCREEN - $indent;
    $tag =~ s/\s*$//;
    s/\s+/ /g;
    s/^ //;
    $str = "format OUTPUT = \n"
	. (($opt_alt_format && $tag_indent > 1)
	   ? ":" . " " x ($tag_indent - 1)
	   : " " x ($tag_indent))
	. '@' . ('<' x ($indent - $tag_indent - 1))
	. "^" .  ("<" x ($cols - 1)) . "\n"
	. '$tag, $_'
	. "\n~~"
	. (" " x ($indent-2))
	. "^" .  ("<" x ($cols - 5)) . "\n"
	. '$_' . "\n\n.\n1";
    #warn $str; warn "tag is $tag, _ is $_";
    eval $str || die;
    write OUTPUT;
}

sub output {
    local($_, $reformat) = @_;
    if ($reformat) {
	$cols = $SCREEN - $indent;
	s/\s+/ /g;
	s/^ //;
	$str = "format OUTPUT = \n~~"
	    . (" " x ($indent-2))
	    . "^" .  ("<" x ($cols - 5)) . "\n"
	    . '$_' . "\n\n.\n1";
	eval $str || die;
	write OUTPUT;
    } else {
	s/^/' ' x $indent/gem;
	s/^\s+\n$/\n/gm;
	s/^  /: /s if defined($reformat) && $opt_alt_format;
	print OUTPUT;
    }
}

sub noremap {
    local($thing_to_hide) = shift;
    $thing_to_hide =~ tr/\000-\177/\200-\377/;
    return $thing_to_hide;
}

sub init_noremap {
    die "unmatched init" if $mapready++;
    #mask off high bit characters in input stream
    s/([\200-\377])/"E<".ord($1).">"/ge;
}

sub clear_noremap {
    my $ready_to_print = $_[0];
    die "unmatched clear" unless $mapready--;
    tr/\200-\377/\000-\177/;
    # now for the E<>s, which have been hidden until now
    # otherwise the interative \w<> processing would have
    # been hosed by the E<gt>
    s {
	    E<
	    (
	    	( \d+ )
	    	| ( [A-Za-z]+ )
	    )
	    >	
    } {
	 do {
	 	defined $2
	 	? chr($2)
	 	:
	     defined $HTML_Escapes{$3}
		? do { $HTML_Escapes{$3} }
		: do {
		    warn "Unknown escape: E<$1> in $_";
		    "E<$1>";
		}
	 }
    }egx if $ready_to_print;
}

sub internal_lrefs {
    local($_) = shift;
    s{L</([^>]+)>}{$1}g;
    my(@items) = split( /(?:,?\s+(?:and\s+)?)/ );
    my $retstr = "the ";
    my $i;
    for ($i = 0; $i <= $#items; $i++) {
	$retstr .= "C<$items[$i]>";
	$retstr .= ", " if @items > 2 && $i != $#items;
	$retstr .= " and " if $i+2 == @items;
    }

    $retstr .= " entr" . ( @items > 1  ? "ies" : "y" )
	    .  " elsewhere in this document ";

    return $retstr;

}

BEGIN {

%HTML_Escapes = (
    'amp'	=>	'&',	#   ampersand
    'lt'	=>	'<',	#   left chevron, less-than
    'gt'	=>	'>',	#   right chevron, greater-than
    'quot'	=>	'"',	#   double quote

    "Aacute"	=>	"\xC1",	#   capital A, acute accent
    "aacute"	=>	"\xE1",	#   small a, acute accent
    "Acirc"	=>	"\xC2",	#   capital A, circumflex accent
    "acirc"	=>	"\xE2",	#   small a, circumflex accent
    "AElig"	=>	"\xC6",	#   capital AE diphthong (ligature)
    "aelig"	=>	"\xE6",	#   small ae diphthong (ligature)
    "Agrave"	=>	"\xC0",	#   capital A, grave accent
    "agrave"	=>	"\xE0",	#   small a, grave accent
    "Aring"	=>	"\xC5",	#   capital A, ring
    "aring"	=>	"\xE5",	#   small a, ring
    "Atilde"	=>	"\xC3",	#   capital A, tilde
    "atilde"	=>	"\xE3",	#   small a, tilde
    "Auml"	=>	"\xC4",	#   capital A, dieresis or umlaut mark
    "auml"	=>	"\xE4",	#   small a, dieresis or umlaut mark
    "Ccedil"	=>	"\xC7",	#   capital C, cedilla
    "ccedil"	=>	"\xE7",	#   small c, cedilla
    "Eacute"	=>	"\xC9",	#   capital E, acute accent
    "eacute"	=>	"\xE9",	#   small e, acute accent
    "Ecirc"	=>	"\xCA",	#   capital E, circumflex accent
    "ecirc"	=>	"\xEA",	#   small e, circumflex accent
    "Egrave"	=>	"\xC8",	#   capital E, grave accent
    "egrave"	=>	"\xE8",	#   small e, grave accent
    "ETH"	=>	"\xD0",	#   capital Eth, Icelandic
    "eth"	=>	"\xF0",	#   small eth, Icelandic
    "Euml"	=>	"\xCB",	#   capital E, dieresis or umlaut mark
    "euml"	=>	"\xEB",	#   small e, dieresis or umlaut mark
    "Iacute"	=>	"\xCD",	#   capital I, acute accent
    "iacute"	=>	"\xED",	#   small i, acute accent
    "Icirc"	=>	"\xCE",	#   capital I, circumflex accent
    "icirc"	=>	"\xEE",	#   small i, circumflex accent
    "Igrave"	=>	"\xCD",	#   capital I, grave accent
    "igrave"	=>	"\xED",	#   small i, grave accent
    "Iuml"	=>	"\xCF",	#   capital I, dieresis or umlaut mark
    "iuml"	=>	"\xEF",	#   small i, dieresis or umlaut mark
    "Ntilde"	=>	"\xD1",		#   capital N, tilde
    "ntilde"	=>	"\xF1",		#   small n, tilde
    "Oacute"	=>	"\xD3",	#   capital O, acute accent
    "oacute"	=>	"\xF3",	#   small o, acute accent
    "Ocirc"	=>	"\xD4",	#   capital O, circumflex accent
    "ocirc"	=>	"\xF4",	#   small o, circumflex accent
    "Ograve"	=>	"\xD2",	#   capital O, grave accent
    "ograve"	=>	"\xF2",	#   small o, grave accent
    "Oslash"	=>	"\xD8",	#   capital O, slash
    "oslash"	=>	"\xF8",	#   small o, slash
    "Otilde"	=>	"\xD5",	#   capital O, tilde
    "otilde"	=>	"\xF5",	#   small o, tilde
    "Ouml"	=>	"\xD6",	#   capital O, dieresis or umlaut mark
    "ouml"	=>	"\xF6",	#   small o, dieresis or umlaut mark
    "szlig"	=>	"\xDF",		#   small sharp s, German (sz ligature)
    "THORN"	=>	"\xDE",	#   capital THORN, Icelandic
    "thorn"	=>	"\xFE",	#   small thorn, Icelandic
    "Uacute"	=>	"\xDA",	#   capital U, acute accent
    "uacute"	=>	"\xFA",	#   small u, acute accent
    "Ucirc"	=>	"\xDB",	#   capital U, circumflex accent
    "ucirc"	=>	"\xFB",	#   small u, circumflex accent
    "Ugrave"	=>	"\xD9",	#   capital U, grave accent
    "ugrave"	=>	"\xF9",	#   small u, grave accent
    "Uuml"	=>	"\xDC",	#   capital U, dieresis or umlaut mark
    "uuml"	=>	"\xFC",	#   small u, dieresis or umlaut mark
    "Yacute"	=>	"\xDD",	#   capital Y, acute accent
    "yacute"	=>	"\xFD",	#   small y, acute accent
    "yuml"	=>	"\xFF",	#   small y, dieresis or umlaut mark

    "lchevron"	=>	"\xAB",	#   left chevron (double less than)
    "rchevron"	=>	"\xBB",	#   right chevron (double greater than)
);
}

1;
Commit	Line	Data
69e00e79	1	package Pod::Text;
69e00e79	2
69e00e79	3	=head1 NAME
	4
	5	Pod::Text - convert POD data to formatted ASCII text
	6
	7	=head1 SYNOPSIS
	8
	9	use Pod::Text;
	10
	11	pod2text("perlfunc.pod");
	12
	13	Also:
	14
f2506fb2	15	pod2text [B<-a>] [B<->I<width>] < input.pod
69e00e79	16
	17	=head1 DESCRIPTION
	18
	19	Pod::Text is a module that can convert documentation in the POD format (such
	20	as can be found throughout the Perl distribution) into formatted ASCII.
	21	Termcap is optionally supported for boldface/underline, and can enabled via
	22	C<$Pod::Text::termcap=1>. If termcap has not been enabled, then backspaces
	23	will be used to simulate bold and underlined text.
	24
	25	A separate F<pod2text> program is included that is primarily a wrapper for
	26	Pod::Text.
	27
f2506fb2	28	The single function C<pod2text()> can take the optional options B<-a>
	29	for an alternative output format, then a B<->I<width> option with the
	30	max terminal width, followed by one or two arguments. The first
1fef88e7	31	should be the name of a file to read the pod from, or "E<lt>&STDIN" to read from
69e00e79	32	STDIN. A second argument, if provided, should be a filehandle glob where
	33	output should be sent.
	34
	35	=head1 AUTHOR
	36
1fef88e7	37	Tom Christiansen E<lt>F<tchrist@mox.perl.com>E<gt>
69e00e79	38
	39	=head1 TODO
	40
	41	Cleanup work. The input and output locations need to be more flexible,
	42	termcap shouldn't be a global variable, and the terminal speed needs to
	43	be properly calculated.
	44
	45	=cut
	46
	47	use Term::Cap;
	48	require Exporter;
	49	@ISA = Exporter;
	50	@EXPORT = qw(pod2text);
	51
f02a87df	52	use vars qw($VERSION);
f2506fb2	53	$VERSION = "1.0203";
f02a87df	54
69e00e79	55	$termcap=0;
69e00e79	56
f2506fb2	57	$opt_alt_format = 0;
f2506fb2	58
69e00e79	59	#$use_format=1;
	60
	61	$UNDL = "\x1b[4m";
	62	$INV = "\x1b[7m";
	63	$BOLD = "\x1b[1m";
	64	$NORM = "\x1b[0m";
	65
	66	sub pod2text {
f2506fb2	67	shift if $opt_alt_format = ($_[0] eq '-a');
69e00e79	68
	69	if($termcap and !$setuptermcap) {
	70	$setuptermcap=1;
	71
	72	my($term) = Tgetent Term::Cap { TERM => undef, OSPEED => 9600 };
	73	$UNDL = $term->{'_us'};
	74	$INV = $term->{'_mr'};
	75	$BOLD = $term->{'_md'};
	76	$NORM = $term->{'_me'};
	77	}
	78
	79	$SCREEN = ($_[0] =~ /^-(\d+)/ && (shift, $1))
69e00e79	80	\|\| $ENV{COLUMNS}
36477c24	81	\|\| ($ENV{TERMCAP} =~ /co#(\d+)/)[0]
2e0a0b36	82	\|\| ($^O ne 'MSWin32' && (`stty -a 2>/dev/null` =~ /(\d+) columns/)[0])
69e00e79	83	\|\| 72;
69e00e79	84
f2506fb2	85	@_ = ("<&STDIN") unless @_;
	86	local($file,*OUTPUT) = @_;
	87	OUTPUT = STDOUT if @_<2;
	88
	89	local $: = $:;
	90	$: = " \n" if $opt_alt_format; # Do not break ``-L/lib/'' into ``- L/lib/''.
	91
69e00e79	92	$/ = "";
	93
	94	$FANCY = 0;
	95
	96	$cutting = 1;
	97	$DEF_INDENT = 4;
	98	$indent = $DEF_INDENT;
	99	$needspace = 0;
8c634b6e	100	$begun = "";
69e00e79	101
2e917f57	102	open(IN, $file) \|\| die "Couldn't open $file: $!";
69e00e79	103
	104	POD_DIRECTIVE: while (<IN>) {
	105	if ($cutting) {
	106	next unless /^=/;
	107	$cutting = 0;
	108	}
8c634b6e	109	if ($begun) {
	110	if (/^=end\s+$begun/) {
	111	$begun = "";
	112	}
	113	elsif ($begun eq "text") {
78ff9ed7	114	print OUTPUT $_;
8c634b6e	115	}
	116	next;
	117	}
69e00e79	118	1 while s{^(.?)(\t+)(.)$}{
	119	$1
	120	. (' ' x (length($2) * 8 - length($1) % 8))
	121	. $3
	122	}me;
	123	# Translate verbatim paragraph
	124	if (/^\s/) {
69e00e79	125	output($_);
	126	next;
	127	}
	128
f02a87df	129	if (/^=for\s+(\S+)\s(.)/s) {
8c634b6e	130	if ($1 eq "text") {
78ff9ed7	131	print OUTPUT $2,"";
8c634b6e	132	} else {
	133	# ignore unknown for
	134	}
	135	next;
	136	}
f02a87df	137	elsif (/^=begin\s+(\S+)\s(.)/s) {
8c634b6e	138	$begun = $1;
8c634b6e	139	if ($1 eq "text") {
78ff9ed7	140	print OUTPUT $2."";
8c634b6e	141	}
	142	next;
	143	}
	144
69e00e79	145	sub prepare_for_output {
	146
	147	s/\s*$/\n/;
	148	&init_noremap;
	149
	150	# need to hide E<> first; they're processed in clear_noremap
	151	s/(E<[^<>]+>)/noremap($1)/ge;
	152	$maxnest = 10;
	153	while ($maxnest-- && /[A-Z]</) {
	154	unless ($FANCY) {
f2506fb2	155	if ($opt_alt_format) {
	156	s/[BC]<(.*?)>/``$1''/sg;
	157	s/F<(.*?)>/"$1"/sg;
	158	} else {
	159	s/C<(.*?)>/`$1'/sg;
	160	}
69e00e79	161	} else {
55497cff	162	s/C<(.*?)>/noremap("E<lchevron>${1}E<rchevron>")/sge;
69e00e79	163	}
69e00e79	164	# s/[IF]<(.*?)>/italic($1)/ge;
55497cff	165	s/I<(.?)>/$1*/sg;
69e00e79	166	# s/[CB]<(.*?)>/bold($1)/ge;
55497cff	167	s/X<.*?>//sg;
69e00e79	168	# LREF: a manpage(3f)
	169	s:L<([a-zA-Z][^\s\/]+)(\([^\)]+\))?>:the $1$2 manpage:g;
	170	# LREF: an =item on another manpage
	171	s{
	172	L<
	173	([^/]+)
	174	/
	175	(
	176	[:\w]+
	177	(\(\))?
	178	)
	179	>
	180	} {the "$2" entry in the $1 manpage}gx;
	181
	182	# LREF: an =item on this manpage
	183	s{
	184	((?:
	185	L<
	186	/
	187	(
	188	[:\w]+
	189	(\(\))?
	190	)
	191	>
	192	(,?\s+(and\s+)?)?
	193	)+)
	194	} { internal_lrefs($1) }gex;
	195
	196	# LREF: a =head2 (head1?), maybe on a manpage, maybe right here
	197	# the "func" can disambiguate
	198	s{
	199	L<
	200	(?:
	201	([a-zA-Z]\S+?) /
	202	)?
	203	"?(.*?)"?
	204	>
	205	}{
	206	do {
	207	$1 # if no $1, assume it means on this page.
	208	? "the section on \"$2\" in the $1 manpage"
	209	: "the section on \"$2\""
	210	}
55497cff	211	}sgex;
69e00e79	212
55497cff	213	s/[A-Z]<(.*?)>/$1/sg;
69e00e79	214	}
	215	clear_noremap(1);
	216	}
	217
	218	&prepare_for_output;
	219
	220	if (s/^=//) {
	221	# $needspace = 0; # Assume this.
	222	# s/\n/ /g;
	223	($Cmd, $_) = split(' ', $_, 2);
	224	# clear_noremap(1);
	225	if ($Cmd eq 'cut') {
	226	$cutting = 1;
	227	}
78ff9ed7	228	elsif ($Cmd eq 'pod') {
	229	$cutting = 0;
	230	}
69e00e79	231	elsif ($Cmd eq 'head1') {
69e00e79	232	makespace();
f2506fb2	233	if ($opt_alt_format) {
	234	print OUTPUT "\n";
	235	s/^(.+?)[ \t]*$/==== $1 ====/;
	236	}
69e00e79	237	print OUTPUT;
69e00e79	238	# print OUTPUT uc($_);
f2506fb2	239	$needspace = $opt_alt_format;
69e00e79	240	}
	241	elsif ($Cmd eq 'head2') {
	242	makespace();
	243	# s/(\w+)/\u\L$1/g;
	244	#print ' ' x $DEF_INDENT, $_;
	245	# print "\xA7";
	246	s/(\w)/\xA7 $1/ if $FANCY;
f2506fb2	247	if ($opt_alt_format) {
	248	s/^(.+?)[ \t]*$/== $1 ==/;
	249	print OUTPUT "\n", $_;
	250	} else {
	251	print OUTPUT ' ' x ($DEF_INDENT/2), $_, "\n";
	252	}
	253	$needspace = $opt_alt_format;
69e00e79	254	}
	255	elsif ($Cmd eq 'over') {
	256	push(@indent,$indent);
	257	$indent += ($_ + 0) \|\| $DEF_INDENT;
	258	}
	259	elsif ($Cmd eq 'back') {
	260	$indent = pop(@indent);
	261	warn "Unmatched =back\n" unless defined $indent;
69e00e79	262	}
	263	elsif ($Cmd eq 'item') {
	264	makespace();
	265	# s/\A(\s)\/$1\xb7/ if $FANCY;
	266	# s/^(\s\\s+)/$1 /;
	267	{
	268	if (length() + 3 < $indent) {
	269	my $paratag = $_;
2e917f57	270	$_ = <IN>;
69e00e79	271	if (/^=/) { # tricked!
	272	local($indent) = $indent[$#index - 1] \|\| $DEF_INDENT;
	273	output($paratag);
	274	redo POD_DIRECTIVE;
	275	}
	276	&prepare_for_output;
	277	IP_output($paratag, $_);
	278	} else {
	279	local($indent) = $indent[$#index - 1] \|\| $DEF_INDENT;
f2506fb2	280	output($_, 0);
69e00e79	281	}
	282	}
	283	}
	284	else {
	285	warn "Unrecognized directive: $Cmd\n";
	286	}
	287	}
	288	else {
	289	# clear_noremap(1);
	290	makespace();
	291	output($_, 1);
	292	}
	293	}
	294
	295	close(IN);
	296
	297	}
	298
	299	#########################################################################
	300
	301	sub makespace {
	302	if ($needspace) {
	303	print OUTPUT "\n";
	304	$needspace = 0;
	305	}
	306	}
	307
	308	sub bold {
	309	my $line = shift;
	310	return $line if $use_format;
	311	if($termcap) {
	312	$line = "$BOLD$line$NORM";
	313	} else {
	314	$line =~ s/(.)/$1\b$1/g;
	315	}
	316	# $line = "$BOLD$line$NORM" if $ansify;
	317	return $line;
	318	}
	319
	320	sub italic {
	321	my $line = shift;
	322	return $line if $use_format;
	323	if($termcap) {
	324	$line = "$UNDL$line$NORM";
	325	} else {
	326	$line =~ s/(.)/$1\b_/g;
	327	}
	328	# $line = "$UNDL$line$NORM" if $ansify;
	329	return $line;
	330	}
	331
	332	# Fill a paragraph including underlined and overstricken chars.
	333	# It's not perfect for words longer than the margin, and it's probably
	334	# slow, but it works.
	335	sub fill {
	336	local $_ = shift;
	337	my $par = "";
	338	my $indent_space = " " x $indent;
	339	my $marg = $SCREEN-$indent;
	340	my $line = $indent_space;
	341	my $line_length;
	342	foreach (split) {
	343	my $word_length = length;
	344	$word_length -= 2 while /\010/g; # Subtract backspaces
345
346	if ($line_length + $word_length > $marg) {
347	$par .= $line . "\n";
348	$line= $indent_space . $_;
349	$line_length = $word_length;
350	}
351	else {
352	if ($line_length) {
353	$line_length++;
354	$line .= " ";
355	}
356	$line_length += $word_length;
357	$line .= $_;
358	}
359	}
360	$par .= "$line\n" if $line;
361	$par .= "\n";
362	return $par;
363	}
364
365	sub IP_output {
366	local($tag, $_) = @_;
367	local($tag_indent) = $indent[$#index - 1] \|\| $DEF_INDENT;
368	$tag_cols = $SCREEN - $tag_indent;
369	$cols = $SCREEN - $indent;
370	$tag =~ s/\s*$//;
371	s/\s+/ /g;
372	s/^ //;
373	$str = "format OUTPUT = \n"
f2506fb2	374	. (($opt_alt_format && $tag_indent > 1)
	375	? ":" . " " x ($tag_indent - 1)
	376	: " " x ($tag_indent))
69e00e79	377	. '@' . ('<' x ($indent - $tag_indent - 1))
	378	. "^" . ("<" x ($cols - 1)) . "\n"
	379	. '$tag, $_'
	380	. "\n~~"
	381	. (" " x ($indent-2))
	382	. "^" . ("<" x ($cols - 5)) . "\n"
	383	. '$_' . "\n\n.\n1";
	384	#warn $str; warn "tag is $tag, _ is $_";
	385	eval $str \|\| die;
	386	write OUTPUT;
	387	}
	388
	389	sub output {
	390	local($_, $reformat) = @_;
	391	if ($reformat) {
	392	$cols = $SCREEN - $indent;
	393	s/\s+/ /g;
	394	s/^ //;
	395	$str = "format OUTPUT = \n~~"
	396	. (" " x ($indent-2))
	397	. "^" . ("<" x ($cols - 5)) . "\n"
	398	. '$_' . "\n\n.\n1";
	399	eval $str \|\| die;
	400	write OUTPUT;
	401	} else {
	402	s/^/' ' x $indent/gem;
	403	s/^\s+\n$/\n/gm;
f2506fb2	404	s/^ /: /s if defined($reformat) && $opt_alt_format;
69e00e79	405	print OUTPUT;
	406	}
	407	}
	408
	409	sub noremap {
	410	local($thing_to_hide) = shift;
	411	$thing_to_hide =~ tr/\000-\177/\200-\377/;
	412	return $thing_to_hide;
	413	}
	414
	415	sub init_noremap {
	416	die "unmatched init" if $mapready++;
26fb054b	417	#mask off high bit characters in input stream
26fb054b	418	s/([\200-\377])/"E<".ord($1).">"/ge;
69e00e79	419	}
	420
	421	sub clear_noremap {
	422	my $ready_to_print = $_[0];
	423	die "unmatched clear" unless $mapready--;
	424	tr/\200-\377/\000-\177/;
	425	# now for the E<>s, which have been hidden until now
	426	# otherwise the interative \w<> processing would have
	427	# been hosed by the E<gt>
	428	s {
26fb054b	429	E<
	430	(
	431	( \d+ )
	432	\| ( [A-Za-z]+ )
	433	)
69e00e79	434	>
	435	} {
	436	do {
26fb054b	437	defined $2
	438	? chr($2)
	439	:
	440	defined $HTML_Escapes{$3}
	441	? do { $HTML_Escapes{$3} }
69e00e79	442	: do {
f02a87df	443	warn "Unknown escape: E<$1> in $_";
69e00e79	444	"E<$1>";
	445	}
	446	}
	447	}egx if $ready_to_print;
	448	}
	449
	450	sub internal_lrefs {
	451	local($_) = shift;
	452	s{L</([^>]+)>}{$1}g;
	453	my(@items) = split( /(?:,?\s+(?:and\s+)?)/ );
	454	my $retstr = "the ";
	455	my $i;
	456	for ($i = 0; $i <= $#items; $i++) {
	457	$retstr .= "C<$items[$i]>";
	458	$retstr .= ", " if @items > 2 && $i != $#items;
	459	$retstr .= " and " if $i+2 == @items;
	460	}
	461
	462	$retstr .= " entr" . ( @items > 1 ? "ies" : "y" )
	463	. " elsewhere in this document ";
	464
	465	return $retstr;
	466
	467	}
	468
	469	BEGIN {
	470
	471	%HTML_Escapes = (
	472	'amp' => '&', # ampersand
	473	'lt' => '<', # left chevron, less-than
	474	'gt' => '>', # right chevron, greater-than
	475	'quot' => '"', # double quote
	476
	477	"Aacute" => "\xC1", # capital A, acute accent
	478	"aacute" => "\xE1", # small a, acute accent
	479	"Acirc" => "\xC2", # capital A, circumflex accent
	480	"acirc" => "\xE2", # small a, circumflex accent
	481	"AElig" => "\xC6", # capital AE diphthong (ligature)
	482	"aelig" => "\xE6", # small ae diphthong (ligature)
	483	"Agrave" => "\xC0", # capital A, grave accent
	484	"agrave" => "\xE0", # small a, grave accent
	485	"Aring" => "\xC5", # capital A, ring
	486	"aring" => "\xE5", # small a, ring
	487	"Atilde" => "\xC3", # capital A, tilde
	488	"atilde" => "\xE3", # small a, tilde
	489	"Auml" => "\xC4", # capital A, dieresis or umlaut mark
	490	"auml" => "\xE4", # small a, dieresis or umlaut mark
	491	"Ccedil" => "\xC7", # capital C, cedilla
	492	"ccedil" => "\xE7", # small c, cedilla
	493	"Eacute" => "\xC9", # capital E, acute accent
	494	"eacute" => "\xE9", # small e, acute accent
	495	"Ecirc" => "\xCA", # capital E, circumflex accent
	496	"ecirc" => "\xEA", # small e, circumflex accent
	497	"Egrave" => "\xC8", # capital E, grave accent
	498	"egrave" => "\xE8", # small e, grave accent
	499	"ETH" => "\xD0", # capital Eth, Icelandic
	500	"eth" => "\xF0", # small eth, Icelandic
	501	"Euml" => "\xCB", # capital E, dieresis or umlaut mark
	502	"euml" => "\xEB", # small e, dieresis or umlaut mark
	503	"Iacute" => "\xCD", # capital I, acute accent
	504	"iacute" => "\xED", # small i, acute accent
	505	"Icirc" => "\xCE", # capital I, circumflex accent
	506	"icirc" => "\xEE", # small i, circumflex accent
	507	"Igrave" => "\xCD", # capital I, grave accent
508	"igrave" => "\xED", # small i, grave accent
509	"Iuml" => "\xCF", # capital I, dieresis or umlaut mark
510	"iuml" => "\xEF", # small i, dieresis or umlaut mark
511	"Ntilde" => "\xD1", # capital N, tilde
512	"ntilde" => "\xF1", # small n, tilde
513	"Oacute" => "\xD3", # capital O, acute accent
514	"oacute" => "\xF3", # small o, acute accent
515	"Ocirc" => "\xD4", # capital O, circumflex accent
516	"ocirc" => "\xF4", # small o, circumflex accent
517	"Ograve" => "\xD2", # capital O, grave accent
518	"ograve" => "\xF2", # small o, grave accent
519	"Oslash" => "\xD8", # capital O, slash
520	"oslash" => "\xF8", # small o, slash
521	"Otilde" => "\xD5", # capital O, tilde
522	"otilde" => "\xF5", # small o, tilde
523	"Ouml" => "\xD6", # capital O, dieresis or umlaut mark
524	"ouml" => "\xF6", # small o, dieresis or umlaut mark
525	"szlig" => "\xDF", # small sharp s, German (sz ligature)
526	"THORN" => "\xDE", # capital THORN, Icelandic
527	"thorn" => "\xFE", # small thorn, Icelandic
528	"Uacute" => "\xDA", # capital U, acute accent
529	"uacute" => "\xFA", # small u, acute accent
530	"Ucirc" => "\xDB", # capital U, circumflex accent
531	"ucirc" => "\xFB", # small u, circumflex accent
532	"Ugrave" => "\xD9", # capital U, grave accent
533	"ugrave" => "\xF9", # small u, grave accent
534	"Uuml" => "\xDC", # capital U, dieresis or umlaut mark
535	"uuml" => "\xFC", # small u, dieresis or umlaut mark
536	"Yacute" => "\xDD", # capital Y, acute accent
537	"yacute" => "\xFD", # small y, acute accent
538	"yuml" => "\xFF", # small y, dieresis or umlaut mark
539
540	"lchevron" => "\xAB", # left chevron (double less than)
541	"rchevron" => "\xBB", # right chevron (double greater than)
542	);
543	}
544
545	1;