[perl5.git] / installhtml

#!/usr/bin/perl -w

use Config;		# for config options in the makefile
use Getopt::Long;	# for command-line parsing
use Cwd;
use Pod::Html;

umask 022;

=head1 NAME

installhtml - converts a collection of POD pages to HTML format.

=head1 SYNOPSIS

    installhtml  [--help] [--podpath=<name>:...:<name>] [--podroot=<name>]
         [--htmldir=<name>] [--htmlroot=<name>]  [--norecurse] [--recurse]
         [--splithead=<name>,...,<name>]   [--splititem=<name>,...,<name>]
	 [--libpods=<name>,...,<name>]  [--verbose]

=head1 DESCRIPTION

I<installhtml> converts a collection of POD pages to a corresponding
collection of HTML pages.  This is primarily used to convert the pod
pages found in the perl distribution.

=head1 OPTIONS

=over 4

=item B<--help> help

Displays the usage.

=item B<--podpath> POD search path

The list of diretories to search for .pod and .pm files to be converted.
Default is `podroot/.'.

=item B<--podroot> POD search path base directory

The base directory to search for all .pod and .pm files to be converted.
Default is current directory.

=item B<--htmldir> HTML destination directory

The base directory which all HTML files will be written to.  This should
be a path relative to the filesystem, not the resulting URL.

=item B<--htmlroot> URL base directory

The base directory which all resulting HTML files will be visible at in
a URL.  The default is `/'.

=item B<--recurse> recurse on subdirectories

Whether or not to convert all .pm and .pod files found in subdirectories
too.  Default is to not recurse.

=item B<--splithead> POD files to split on =head directive

Colon-seperated list of pod files to split by the =head directive.  These
files should have names specified relative to podroot.

=item B<--splititem> POD files to split on =item directive

Colon-seperated list of all pod files to split by the =item directive.
I<installhtml> does not do the actual split, rather it invokes I<splitpod>
to do the dirty work.  As with --splithead, these files should have names
specified relative to podroot.

=item B<--libpods> library PODs for LE<lt>E<gt> links

Colon-seperated list of "library" pod files.  This is the same list that
will be passed to pod2html when any pod is converted.

=item B<--verbose> verbose output

Self-explanatory.

=back

=head1 EXAMPLE

The following command-line is an example of the one we use to convert
perl documentation:

    ./installhtml --podpath=lib:ext:pod:vms   \
		  --podroot=/usr/src/perl     \
                  --htmldir=/perl/nmanual     \
		  --htmlroot=/perl/nmanual    \
		  --splithead=pod/perlipc.pod \
		  --splititem=pod/perlfunc    \
		  --libpods=perlfunc:perlguts:perlvar:perlrun:perlop \
                  --recurse \
		  --verbose

=head1 AUTHOR

Chris Hall E<lt>hallc@cs.colorado.eduE<gt>

=head1 TODO

=cut

$usage =<<END_OF_USAGE;
Usage: $0 --help --podpath=<name>:...:<name> --podroot=<name>
         --htmldir=<name> --htmlroot=<name> --norecurse --recurse
         --splithead=<name>,...,<name> --splititem=<name>,...,<name>
	 --libpods=<name>,...,<name> --verbose

    --help      - this message
    --podpath   - colon-separated list of directories containing .pod and
                  .pm files to be converted (. by default).
    --podroot   - filesystem base directory from which all relative paths in
                  podpath stem (default is .).
    --htmldir   - directory to store resulting html files in relative
                  to the filesystem (\$podroot/html by default). 
    --htmlroot  - http-server base directory from which all relative paths
                  in podpath stem (default is /).
    --libpods   - comma-separated list of files to search for =item pod
                  directives in as targets of C<> and implicit links (empty
                  by default).
    --norecurse - don't recurse on those subdirectories listed in podpath.
                  (default behavior).
    --recurse   - recurse on those subdirectories listed in podpath
    --splithead - comma-separated list of .pod or .pm files to split.  will
                  split each file into several smaller files at every occurance
                  of a pod =head[1-6] directive.
    --splititem - comma-separated list of .pod or .pm files to split using
                  splitpod.
    --splitpod  - where the program splitpod can be found (\$podroot/pod by
                  default).
    --verbose   - self-explanatory.

END_OF_USAGE

@libpods = ();
@podpath = ( "." );	# colon-separated list of directories containing .pod
			# and .pm files to be converted.
$podroot = ".";		# assume the pods we want are here
$htmldir = "";		# nothing for now...
$htmlroot = "/";	# default value
$recurse = 0;		# default behavior
@splithead = ();	# don't split any files by default
@splititem = ();	# don't split any files by default
$splitpod = "";		# nothing for now.

$verbose = 0; 	    	# whether or not to print debugging info

$pod2html = "pod/pod2html";


# parse the command-line
$result = GetOptions( qw(
	help
	podpath=s
	podroot=s
	htmldir=s
	htmlroot=s
	libpods=s
	recurse!
	splithead=s
	splititem=s
	splitpod=s
	verbose
));
usage("invalid parameters") unless $result;
parse_command_line();


# set these variables to appropriate values if the user didn't specify
#  values for them.
$htmldir = "$htmlroot/html" unless $htmldir;
$splitpod = "$podroot/pod" unless $splitpod;


# make sure that the destination directory exists
(mkdir($htmldir, 0755) ||
	die "$0: cannot make directory $htmldir: $!\n") if ! -d $htmldir;


# the following array will eventually contain files that are to be
# ignored in the conversion process.  these are files that have been
# process by splititem or splithead and should not be converted as a
# result.
@ignore = ();


# split pods.  its important to do this before convert ANY pods because
#  it may effect some of the links
@splitdirs = ();    # files in these directories won't get an index
split_on_head($podroot, $htmldir, \@splitdirs, \@ignore, @splithead);
split_on_item($podroot, \@splitdirs, \@ignore, @splititem);


# convert the pod pages found in @poddirs
#warn "converting files\n" if $verbose;
#warn "\@ignore\t= @ignore\n" if $verbose;
foreach $dir (@podpath) {
    installdir($dir, $recurse, $podroot, \@splitdirs, \@ignore);
}


# now go through and create master indices for each pod we split
foreach $dir (@splititem) {
    print "creating index $htmldir/$dir.html\n" if $verbose;
    create_index("$htmldir/$dir.html", "$htmldir/$dir");
}

foreach $dir (@splithead) {
    $dir .= ".pod" unless $dir =~ /(\.pod|\.pm)$/;
    # let pod2html create the file
    runpod2html($dir, 1);

    # now go through and truncate after the index
    $dir =~ /^(.*?)(\.pod|\.pm)?$/sm;
    $file = "$htmldir/$1";
    print "creating index $file.html\n" if $verbose;

    # read in everything until what would have been the first =head
    # directive, patching the index as we go.
    open(H, "<$file.html") ||
	die "$0: error opening $file.html for input: $!\n";
    $/ = "";
    @data = ();
    while (<H>) {
	last if /NAME=/;
	s,HREF="#(.*)">,HREF="$file/$1.html">,g;
	push @data, $_;
    } 
    close(H);

    # now rewrite the file 
    open(H, ">$file.html") ||
	die "$0: error opening $file.html for output: $!\n";
    print H "@data\n";
    close(H);
}

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


sub usage {
    warn "$0: @_\n" if @_;
    die $usage;
}


sub parse_command_line {
    usage() if defined $opt_help;
    $opt_help = ""; 	    	    # make -w shut up

    # list of directories
    @podpath   = split(":", $opt_podpath) if defined $opt_podpath;

    # lists of files
    @splithead = split(",", $opt_splithead) if defined $opt_splithead;
    @splititem = split(",", $opt_splititem) if defined $opt_splititem;
    @libpods   = split(",", $opt_libpods) if defined $opt_libpods;

    $htmldir  = $opt_htmldir	    if defined $opt_htmldir;
    $htmlroot = $opt_htmlroot	    if defined $opt_htmlroot;
    $podroot  = $opt_podroot	    if defined $opt_podroot;
    $splitpod = $opt_splitpod	    if defined $opt_splitpod;

    $recurse  = $opt_recurse	    if defined $opt_recurse;
    $verbose  = $opt_verbose	    if defined $opt_verbose;
}


sub create_index {
    my($html, $dir) = @_;
    my(@files, @filedata, @index, $file);

    # get the list of .html files in this directory
    opendir(DIR, $dir) ||
	die "$0: error opening directory $dir for reading: $!\n";
    @files = sort(grep(/\.html$/, readdir(DIR)));
    closedir(DIR);

    open(HTML, ">$html") ||
	die "$0: error opening $html for output: $!\n";

    # for each .html file in the directory, extract the index
    #	embedded in the file and throw it into the big index.
    print HTML "<DL COMPACT>\n";
    foreach $file (@files) {
	$/ = "";

	open(IN, "<$dir/$file") ||
	    die "$0: error opening $dir/$file for input: $!\n";
	@filedata = <IN>;
	close(IN);

	# pull out the NAME section
	($name) = grep(/NAME=/, @filedata);
	$name =~ m,/H1>\s(\S+)\s[\s-]*(.*?)\s*$,sm;
	print HTML qq(<A HREF="$dir/$file">);
	print HTML "<DT>$1</A><DD>$2\n" if defined $1;
#	print HTML qq(<A HREF="$dir/$file">$1</A><BR>\n") if defined $1;

	next;

	@index = grep(/<!-- INDEX BEGIN -->.*<!-- INDEX END -->/s,
		    @filedata);
	for (@index) {
	    s/<!-- INDEX BEGIN -->(\s*<!--)(.*)(-->\s*)<!-- INDEX END -->/$2/s;
	    s,#,$dir/$file#,g;
	    # print HTML "$_\n";
	    print HTML "$_\n<P><HR><P>\n";
	}
    }
    print HTML "</DL>\n";

    close(HTML);
}


sub split_on_head {
    my($podroot, $htmldir, $splitdirs, $ignore, @splithead) = @_;
    my($pod, $dirname, $filename);

    # split the files specified in @splithead on =head[1-6] pod directives
    print "splitting files by head.\n" if $verbose && $#splithead >= 0;
    foreach $pod (@splithead) {
	# figure out the directory name and filename
	$pod      =~ s,^([^/]*)$,/$1,;
	$pod      =~ m,(.*?)/(.*?)(\.pod)?$,;
	$dirname  = $1;
	$filename = "$2.pod";

	# since we are splitting this file it shouldn't be converted.
	push(@$ignore, "$podroot/$dirname/$filename");

	# split the pod
	splitpod("$podroot/$dirname/$filename", "$podroot/$dirname", $htmldir,
	    $splitdirs);
    }
}


sub split_on_item {
    my($podroot, $splitdirs, $ignore, @splititem) = @_;
    my($pwd, $dirname, $filename);

    print "splitting files by item.\n" if $verbose && $#splititem >= 0;
    $pwd = getcwd();
    foreach $pod (@splititem) {
	# figure out the directory to split into
	$pod      =~ s,^([^/]*)$,/$1,;
	$pod      =~ m,(.*?)/(.*?)(\.pod)?$,;
	$dirname  = "$1/$2";
	$filename = "$2.pod";

	# since we are splitting this file it shouldn't be converted.
	push(@$ignore, "$podroot/$dirname.pod");

	# split the pod
	push(@$splitdirs, "$podroot/$dirname");
	if (! -d "$podroot/$dirname") {
	    mkdir("$podroot/$dirname", 0755) ||
		    die "$0: error creating directory $podroot/$dirname: $!\n";
	}
	chdir("$podroot/$dirname") ||
	    die "$0: error changing to directory $podroot/$dirname: $!\n";
	system("./splitpod", "../$filename");
    }
    chdir($pwd);
}


#
# splitpod - splits a .pod file into several smaller .pod files
#  where a new file is started each time a =head[1-6] pod directive
#  is encountered in the input file.
#
sub splitpod {
    my($pod, $poddir, $htmldir, $splitdirs) = @_;
    my(@poddata, @filedata, @heads);
    my($file, $i, $j, $prevsec, $section, $nextsec);

    print "splitting $pod\n" if $verbose;

    # read the file in paragraphs
    $/ = "";
    open(SPLITIN, "<$pod") ||
	die "$0: error opening $pod for input: $!\n";
    @filedata = <SPLITIN>;
    close(SPLITIN) ||
	die "$0: error closing $pod: $!\n";

    # restore the file internally by =head[1-6] sections
    @poddata = ();
    for ($i = 0, $j = -1; $i <= $#filedata; $i++) {
	$j++ if ($filedata[$i] =~ /^\s*=head[1-6]/);
	if ($j >= 0) { 
	    $poddata[$j]  = "" unless defined $poddata[$j];
	    $poddata[$j] .= "\n$filedata[$i]" if $j >= 0;
	}
    }

    # create list of =head[1-6] sections so that we can rewrite
    #  L<> links as necessary.
    %heads = ();
    foreach $i (0..$#poddata) {
	$heads{htmlize($1)} = 1 if $poddata[$i] =~ /=head[1-6]\s+(.*)/;
    }

    # create a directory of a similar name and store all the
    #  files in there
    $pod =~ s,.*/(.*),$1,;	# get the last part of the name
    $dir = $pod;
    $dir =~ s/\.pod//g;
    push(@$splitdirs, "$poddir/$dir");
    mkdir("$poddir/$dir", 0755) ||
	die "$0: could not create directory $poddir/$dir: $!\n"
	unless -d "$poddir/$dir";

    $poddata[0] =~ /^\s*=head[1-6]\s+(.*)/;
    $section    = "";
    $nextsec    = $1;

    # for each section of the file create a separate pod file
    for ($i = 0; $i <= $#poddata; $i++) {
	# determine the "prev" and "next" links
	$prevsec = $section;
	$section = $nextsec;
	if ($i < $#poddata) {
	    $poddata[$i+1] =~ /^\s*=head[1-6]\s+(.*)/;
	    $nextsec       = $1;
	} else {
	    $nextsec = "";
	}

	# determine an appropriate filename (this must correspond with
	#  what pod2html will try and guess)
	# $poddata[$i] =~ /^\s*=head[1-6]\s+(.*)/;
	$file = "$dir/" . htmlize($section) . ".pod";

	# create the new .pod file
	print "\tcreating $poddir/$file\n" if $verbose;
	open(SPLITOUT, ">$poddir/$file") ||
	    die "$0: error opening $poddir/$file for output: $!\n";
	$poddata[$i] =~ s,L<([^<>]*)>,
			    defined $heads{htmlize($1)} ? "L<$dir/$1>" : "L<$1>"
			 ,ge;
	print SPLITOUT $poddata[$i]."\n\n";
	print SPLITOUT "=over 4\n\n";
	print SPLITOUT "=item *\n\nBack to L<$dir/\"$prevsec\">\n\n" if $prevsec;
	print SPLITOUT "=item *\n\nForward to L<$dir/\"$nextsec\">\n\n" if $nextsec;
	print SPLITOUT "=item *\n\nUp to L<$dir>\n\n";
	print SPLITOUT "=back\n\n";
	close(SPLITOUT) ||
	    die "$0: error closing $poddir/$file: $!\n";
    }
}


#
# installdir - takes care of converting the .pod and .pm files in the
#  current directory to .html files and then installing those.
#
sub installdir {
    my($dir, $recurse, $podroot, $splitdirs, $ignore) = @_;
    my(@dirlist, @podlist, @pmlist, $doindex);

    @dirlist = ();	# directories to recurse on
    @podlist = ();	# .pod files to install
    @pmlist  = ();	# .pm files to install

    # should files in this directory get an index?
    $doindex = (grep($_ eq "$podroot/$dir", @$splitdirs) ? 0 : 1);

    opendir(DIR, "$podroot/$dir")
	|| die "$0: error opening directory $podroot/$dir: $!\n";

    # find the directories to recurse on
    @dirlist = map { "$dir/$_" }
	grep(-d "$podroot/$dir/$_" && !/^\.{1,2}/, readdir(DIR)) if $recurse;
    rewinddir(DIR);

    # find all the .pod files within the directory
    @podlist = map { /^(.*)\.pod$/; "$dir/$1" }
	grep(! -d "$podroot/$dir/$_" && /\.pod$/, readdir(DIR));
    rewinddir(DIR);

    # find all the .pm files within the directory
    @pmlist = map { /^(.*)\.pm$/; "$dir/$1" }
	grep(! -d "$podroot/$dir/$_" && /\.pm$/, readdir(DIR));

    closedir(DIR);

    # recurse on all subdirectories we kept track of
    foreach $dir (@dirlist) {
	installdir($dir, $recurse, $podroot, $splitdirs, $ignore);
    }

    # install all the pods we found
    foreach $pod (@podlist) {
	# check if we should ignore it.
	next if grep($_ eq "$podroot/$pod.pod", @$ignore);

	# check if a .pm files exists too
	if (grep($_ eq "$pod.pm", @pmlist)) {
	    print  "$0: Warning both `$podroot/$pod.pod' and "
		. "`$podroot/$pod.pm' exist, using pod\n";
	    push(@ignore, "$pod.pm");
	}
	runpod2html("$pod.pod", $doindex);
    }

    # install all the .pm files we found
    foreach $pm (@pmlist) {
	# check if we should ignore it.
	next if grep($_ eq "$pm.pm", @ignore);

	runpod2html("$pm.pm", $doindex);
    }
}


#
# runpod2html - invokes pod2html to convert a .pod or .pm file to a .html
#  file.
#
sub runpod2html {
    my($pod, $doindex) = @_;
    my($html, $i, $dir, @dirs);

    $html = $pod;
    $html =~ s/\.(pod|pm)$/.html/g;

    # make sure the destination directories exist
    @dirs = split("/", $html);
    $dir  = "$htmldir/";
    for ($i = 0; $i < $#dirs; $i++) {
	if (! -d "$dir$dirs[$i]") {
	    mkdir("$dir$dirs[$i]", 0755) ||
		die "$0: error creating directory $dir$dirs[$i]: $!\n";
	}
	$dir .= "$dirs[$i]/";
    }

    # invoke pod2html
    print "$podroot/$pod => $htmldir/$html\n" if $verbose;
#system("./pod2html",
        Pod::Html'pod2html(
        #Pod::Html'pod2html($pod2html,
	"--htmlroot=$htmlroot",
	"--podpath=".join(":", @podpath),
	"--podroot=$podroot", "--netscape",
	($doindex ? "--index" : "--noindex"),
	"--" . ($recurse ? "" : "no") . "recurse",
	($#libpods >= 0) ? "--libpods=" . join(":", @libpods) : "",
	"--infile=$podroot/$pod", "--outfile=$htmldir/$html");
    die "$0: error running $pod2html: $!\n" if $?;
}

sub htmlize { htmlify(0, @_) }
Commit	Line	Data
54310121	1	#!/usr/bin/perl -w
	2
	3	use Config; # for config options in the makefile
	4	use Getopt::Long; # for command-line parsing
	5	use Cwd;
	6	use Pod::Html;
	7
	8	umask 022;
	9
	10	=head1 NAME
	11
	12	installhtml - converts a collection of POD pages to HTML format.
	13
	14	=head1 SYNOPSIS
	15
	16	installhtml [--help] [--podpath=<name>:...:<name>] [--podroot=<name>]
	17	[--htmldir=<name>] [--htmlroot=<name>] [--norecurse] [--recurse]
	18	[--splithead=<name>,...,<name>] [--splititem=<name>,...,<name>]
	19	[--libpods=<name>,...,<name>] [--verbose]
	20
	21	=head1 DESCRIPTION
	22
	23	I<installhtml> converts a collection of POD pages to a corresponding
	24	collection of HTML pages. This is primarily used to convert the pod
	25	pages found in the perl distribution.
	26
	27	=head1 OPTIONS
	28
	29	=over 4
	30
	31	=item B<--help> help
	32
	33	Displays the usage.
	34
	35	=item B<--podpath> POD search path
	36
	37	The list of diretories to search for .pod and .pm files to be converted.
	38	Default is `podroot/.'.
	39
	40	=item B<--podroot> POD search path base directory
	41
	42	The base directory to search for all .pod and .pm files to be converted.
	43	Default is current directory.
	44
	45	=item B<--htmldir> HTML destination directory
	46
	47	The base directory which all HTML files will be written to. This should
	48	be a path relative to the filesystem, not the resulting URL.
	49
	50	=item B<--htmlroot> URL base directory
	51
	52	The base directory which all resulting HTML files will be visible at in
	53	a URL. The default is `/'.
	54
	55	=item B<--recurse> recurse on subdirectories
	56
	57	Whether or not to convert all .pm and .pod files found in subdirectories
	58	too. Default is to not recurse.
	59
	60	=item B<--splithead> POD files to split on =head directive
	61
	62	Colon-seperated list of pod files to split by the =head directive. These
	63	files should have names specified relative to podroot.
	64
65	=item B<--splititem> POD files to split on =item directive
66
67	Colon-seperated list of all pod files to split by the =item directive.
68	I<installhtml> does not do the actual split, rather it invokes I<splitpod>
69	to do the dirty work. As with --splithead, these files should have names
70	specified relative to podroot.
71
72	=item B<--libpods> library PODs for LE<lt>E<gt> links
73
74	Colon-seperated list of "library" pod files. This is the same list that
75	will be passed to pod2html when any pod is converted.
76
77	=item B<--verbose> verbose output
78
79	Self-explanatory.
80
81	=back
82
83	=head1 EXAMPLE
84
85	The following command-line is an example of the one we use to convert
86	perl documentation:
87
88	./installhtml --podpath=lib:ext:pod:vms \
89	--podroot=/usr/src/perl \
90	--htmldir=/perl/nmanual \
91	--htmlroot=/perl/nmanual \
92	--splithead=pod/perlipc.pod \
93	--splititem=pod/perlfunc \
94	--libpods=perlfunc:perlguts:perlvar:perlrun:perlop \
95	--recurse \
96	--verbose
97
98	=head1 AUTHOR
99
100	Chris Hall E<lt>hallc@cs.colorado.eduE<gt>
101
102	=head1 TODO
103
104	=cut
105
106	$usage =<<END_OF_USAGE;
107	Usage: $0 --help --podpath=<name>:...:<name> --podroot=<name>
108	--htmldir=<name> --htmlroot=<name> --norecurse --recurse
109	--splithead=<name>,...,<name> --splititem=<name>,...,<name>
110	--libpods=<name>,...,<name> --verbose
111
112	--help - this message
113	--podpath - colon-separated list of directories containing .pod and
114	.pm files to be converted (. by default).
115	--podroot - filesystem base directory from which all relative paths in
116	podpath stem (default is .).
117	--htmldir - directory to store resulting html files in relative
118	to the filesystem (\$podroot/html by default).
119	--htmlroot - http-server base directory from which all relative paths
120	in podpath stem (default is /).
121	--libpods - comma-separated list of files to search for =item pod
122	directives in as targets of C<> and implicit links (empty
123	by default).
124	--norecurse - don't recurse on those subdirectories listed in podpath.
125	(default behavior).
126	--recurse - recurse on those subdirectories listed in podpath
127	--splithead - comma-separated list of .pod or .pm files to split. will
128	split each file into several smaller files at every occurance
129	of a pod =head[1-6] directive.
130	--splititem - comma-separated list of .pod or .pm files to split using
131	splitpod.
132	--splitpod - where the program splitpod can be found (\$podroot/pod by
133	default).
134	--verbose - self-explanatory.
135
136	END_OF_USAGE
137
138	@libpods = ();
139	@podpath = ( "." ); # colon-separated list of directories containing .pod
140	# and .pm files to be converted.
141	$podroot = "."; # assume the pods we want are here
142	$htmldir = ""; # nothing for now...
143	$htmlroot = "/"; # default value
144	$recurse = 0; # default behavior
145	@splithead = (); # don't split any files by default
146	@splititem = (); # don't split any files by default
147	$splitpod = ""; # nothing for now.
148
149	$verbose = 0; # whether or not to print debugging info
150
151	$pod2html = "pod/pod2html";
152
153
154	# parse the command-line
155	$result = GetOptions( qw(
156	help
157	podpath=s
158	podroot=s
159	htmldir=s
160	htmlroot=s
161	libpods=s
162	recurse!
163	splithead=s
164	splititem=s
165	splitpod=s
166	verbose
167	));
168	usage("invalid parameters") unless $result;
169	parse_command_line();
170
171
172	# set these variables to appropriate values if the user didn't specify
173	# values for them.
174	$htmldir = "$htmlroot/html" unless $htmldir;
175	$splitpod = "$podroot/pod" unless $splitpod;
176
177
178	# make sure that the destination directory exists
179	(mkdir($htmldir, 0755) \|\|
180	die "$0: cannot make directory $htmldir: $!\n") if ! -d $htmldir;
181
182
183	# the following array will eventually contain files that are to be
184	# ignored in the conversion process. these are files that have been
185	# process by splititem or splithead and should not be converted as a
186	# result.
187	@ignore = ();
188
189
190	# split pods. its important to do this before convert ANY pods because
191	# it may effect some of the links
192	@splitdirs = (); # files in these directories won't get an index
193	split_on_head($podroot, $htmldir, \@splitdirs, \@ignore, @splithead);
194	split_on_item($podroot, \@splitdirs, \@ignore, @splititem);
195
196
197	# convert the pod pages found in @poddirs
198	#warn "converting files\n" if $verbose;
199	#warn "\@ignore\t= @ignore\n" if $verbose;
200	foreach $dir (@podpath) {
201	installdir($dir, $recurse, $podroot, \@splitdirs, \@ignore);
202	}
203
204
205	# now go through and create master indices for each pod we split
206	foreach $dir (@splititem) {
207	print "creating index $htmldir/$dir.html\n" if $verbose;
208	create_index("$htmldir/$dir.html", "$htmldir/$dir");
209	}
210
211	foreach $dir (@splithead) {
212	$dir .= ".pod" unless $dir =~ /(\.pod\|\.pm)$/;
213	# let pod2html create the file
214	runpod2html($dir, 1);
215
216	# now go through and truncate after the index
217	$dir =~ /^(.*?)(\.pod\|\.pm)?$/sm;
218	$file = "$htmldir/$1";
219	print "creating index $file.html\n" if $verbose;
220
221	# read in everything until what would have been the first =head
222	# directive, patching the index as we go.
223	open(H, "<$file.html") \|\|
224	die "$0: error opening $file.html for input: $!\n";
225	$/ = "";
226	@data = ();
227	while (<H>) {
228	last if /NAME=/;
229	s,HREF="#(.*)">,HREF="$file/$1.html">,g;
230	push @data, $_;
231	}
232	close(H);
233
234	# now rewrite the file
235	open(H, ">$file.html") \|\|
236	die "$0: error opening $file.html for output: $!\n";
237	print H "@data\n";
238	close(H);
239	}
240
241	##############################################################################
242
243
244	sub usage {
245	warn "$0: @_\n" if @_;
246	die $usage;
247	}
248
249
250	sub parse_command_line {
251	usage() if defined $opt_help;
252	$opt_help = ""; # make -w shut up
253
254	# list of directories
255	@podpath = split(":", $opt_podpath) if defined $opt_podpath;
256
257	# lists of files
258	@splithead = split(",", $opt_splithead) if defined $opt_splithead;
259	@splititem = split(",", $opt_splititem) if defined $opt_splititem;
260	@libpods = split(",", $opt_libpods) if defined $opt_libpods;
261
262	$htmldir = $opt_htmldir if defined $opt_htmldir;
263	$htmlroot = $opt_htmlroot if defined $opt_htmlroot;
264	$podroot = $opt_podroot if defined $opt_podroot;
265	$splitpod = $opt_splitpod if defined $opt_splitpod;
266
267	$recurse = $opt_recurse if defined $opt_recurse;
268	$verbose = $opt_verbose if defined $opt_verbose;
269	}
270
271
272	sub create_index {
273	my($html, $dir) = @_;
274	my(@files, @filedata, @index, $file);
275
276	# get the list of .html files in this directory
277	opendir(DIR, $dir) \|\|
278	die "$0: error opening directory $dir for reading: $!\n";
279	@files = sort(grep(/\.html$/, readdir(DIR)));
280	closedir(DIR);
281
282	open(HTML, ">$html") \|\|
283	die "$0: error opening $html for output: $!\n";
284
285	# for each .html file in the directory, extract the index
286	# embedded in the file and throw it into the big index.
287	print HTML "<DL COMPACT>\n";
288	foreach $file (@files) {
289	$/ = "";
290
291	open(IN, "<$dir/$file") \|\|
292	die "$0: error opening $dir/$file for input: $!\n";
293	@filedata = <IN>;
294	close(IN);
295
296	# pull out the NAME section
297	($name) = grep(/NAME=/, @filedata);
298	$name =~ m,/H1>\s(\S+)\s[\s-](.?)\s*$,sm;
299	print HTML qq(<A HREF="$dir/$file">);
300	print HTML "<DT>$1</A><DD>$2\n" if defined $1;
301	# print HTML qq(<A HREF="$dir/$file">$1</A><BR>\n") if defined $1;
302
303	next;
304
305	@index = grep(/<!-- INDEX BEGIN -->.*<!-- INDEX END -->/s,
306	@filedata);
307	for (@index) {
308	s/<!-- INDEX BEGIN -->(\s<!--)(.)(-->\s*)<!-- INDEX END -->/$2/s;
309	s,#,$dir/$file#,g;
310	# print HTML "$_\n";
311	print HTML "$_\n<P><HR><P>\n";
312	}
313	}
314	print HTML "</DL>\n";
315
316	close(HTML);
317	}
318
319
320	sub split_on_head {
321	my($podroot, $htmldir, $splitdirs, $ignore, @splithead) = @_;
322	my($pod, $dirname, $filename);
323
324	# split the files specified in @splithead on =head[1-6] pod directives
325	print "splitting files by head.\n" if $verbose && $#splithead >= 0;
326	foreach $pod (@splithead) {
327	# figure out the directory name and filename
328	$pod =~ s,^([^/]*)$,/$1,;
329	$pod =~ m,(.?)/(.?)(\.pod)?$,;
330	$dirname = $1;
331	$filename = "$2.pod";
332
333	# since we are splitting this file it shouldn't be converted.
334	push(@$ignore, "$podroot/$dirname/$filename");
335
336	# split the pod
337	splitpod("$podroot/$dirname/$filename", "$podroot/$dirname", $htmldir,
338	$splitdirs);
339	}
340	}
341
342
343	sub split_on_item {
344	my($podroot, $splitdirs, $ignore, @splititem) = @_;
345	my($pwd, $dirname, $filename);
346
347	print "splitting files by item.\n" if $verbose && $#splititem >= 0;
348	$pwd = getcwd();
349	foreach $pod (@splititem) {
350	# figure out the directory to split into
351	$pod =~ s,^([^/]*)$,/$1,;
352	$pod =~ m,(.?)/(.?)(\.pod)?$,;
353	$dirname = "$1/$2";
354	$filename = "$2.pod";
355
356	# since we are splitting this file it shouldn't be converted.
357	push(@$ignore, "$podroot/$dirname.pod");
358
359	# split the pod
360	push(@$splitdirs, "$podroot/$dirname");
361	if (! -d "$podroot/$dirname") {
362	mkdir("$podroot/$dirname", 0755) \|\|
363	die "$0: error creating directory $podroot/$dirname: $!\n";
364	}
365	chdir("$podroot/$dirname") \|\|
366	die "$0: error changing to directory $podroot/$dirname: $!\n";
367	system("./splitpod", "../$filename");
368	}
369	chdir($pwd);
370	}
371
372
373	#
374	# splitpod - splits a .pod file into several smaller .pod files
375	# where a new file is started each time a =head[1-6] pod directive
376	# is encountered in the input file.
377	#
378	sub splitpod {
379	my($pod, $poddir, $htmldir, $splitdirs) = @_;
380	my(@poddata, @filedata, @heads);
381	my($file, $i, $j, $prevsec, $section, $nextsec);
382
383	print "splitting $pod\n" if $verbose;
384
385	# read the file in paragraphs
386	$/ = "";
387	open(SPLITIN, "<$pod") \|\|
388	die "$0: error opening $pod for input: $!\n";
389	@filedata = <SPLITIN>;
390	close(SPLITIN) \|\|
391	die "$0: error closing $pod: $!\n";
392
393	# restore the file internally by =head[1-6] sections
394	@poddata = ();
395	for ($i = 0, $j = -1; $i <= $#filedata; $i++) {
396	$j++ if ($filedata[$i] =~ /^\s*=head[1-6]/);
397	if ($j >= 0) {
398	$poddata[$j] = "" unless defined $poddata[$j];
399	$poddata[$j] .= "\n$filedata[$i]" if $j >= 0;
400	}
401	}
402
403	# create list of =head[1-6] sections so that we can rewrite
404	# L<> links as necessary.
405	%heads = ();
406	foreach $i (0..$#poddata) {
407	$heads{htmlize($1)} = 1 if $poddata[$i] =~ /=head[1-6]\s+(.*)/;
408	}
409
410	# create a directory of a similar name and store all the
411	# files in there
412	$pod =~ s,./(.),$1,; # get the last part of the name
413	$dir = $pod;
414	$dir =~ s/\.pod//g;
415	push(@$splitdirs, "$poddir/$dir");
416	mkdir("$poddir/$dir", 0755) \|\|
417	die "$0: could not create directory $poddir/$dir: $!\n"
418	unless -d "$poddir/$dir";
419
420	$poddata[0] =~ /^\s=head[1-6]\s+(.)/;
421	$section = "";
422	$nextsec = $1;
423
424	# for each section of the file create a separate pod file
425	for ($i = 0; $i <= $#poddata; $i++) {
426	# determine the "prev" and "next" links
427	$prevsec = $section;
428	$section = $nextsec;
429	if ($i < $#poddata) {
430	$poddata[$i+1] =~ /^\s=head[1-6]\s+(.)/;
431	$nextsec = $1;
432	} else {
433	$nextsec = "";
434	}
435
436	# determine an appropriate filename (this must correspond with
437	# what pod2html will try and guess)
438	# $poddata[$i] =~ /^\s=head[1-6]\s+(.)/;
439	$file = "$dir/" . htmlize($section) . ".pod";
440
441	# create the new .pod file
442	print "\tcreating $poddir/$file\n" if $verbose;
443	open(SPLITOUT, ">$poddir/$file") \|\|
444	die "$0: error opening $poddir/$file for output: $!\n";
445	$poddata[$i] =~ s,L<([^<>]*)>,
446	defined $heads{htmlize($1)} ? "L<$dir/$1>" : "L<$1>"
447	,ge;
448	print SPLITOUT $poddata[$i]."\n\n";
449	print SPLITOUT "=over 4\n\n";
450	print SPLITOUT "=item *\n\nBack to L<$dir/\"$prevsec\">\n\n" if $prevsec;
451	print SPLITOUT "=item *\n\nForward to L<$dir/\"$nextsec\">\n\n" if $nextsec;
452	print SPLITOUT "=item *\n\nUp to L<$dir>\n\n";
453	print SPLITOUT "=back\n\n";
454	close(SPLITOUT) \|\|
455	die "$0: error closing $poddir/$file: $!\n";
456	}
457	}
458
459
460	#
461	# installdir - takes care of converting the .pod and .pm files in the
462	# current directory to .html files and then installing those.
463	#
464	sub installdir {
465	my($dir, $recurse, $podroot, $splitdirs, $ignore) = @_;
466	my(@dirlist, @podlist, @pmlist, $doindex);
467
468	@dirlist = (); # directories to recurse on
469	@podlist = (); # .pod files to install
470	@pmlist = (); # .pm files to install
471
472	# should files in this directory get an index?
473	$doindex = (grep($_ eq "$podroot/$dir", @$splitdirs) ? 0 : 1);
474
475	opendir(DIR, "$podroot/$dir")
476	\|\| die "$0: error opening directory $podroot/$dir: $!\n";
477
478	# find the directories to recurse on
479	@dirlist = map { "$dir/$_" }
480	grep(-d "$podroot/$dir/$_" && !/^\.{1,2}/, readdir(DIR)) if $recurse;
481	rewinddir(DIR);
482
483	# find all the .pod files within the directory
484	@podlist = map { /^(.*)\.pod$/; "$dir/$1" }
485	grep(! -d "$podroot/$dir/$_" && /\.pod$/, readdir(DIR));
486	rewinddir(DIR);
487
488	# find all the .pm files within the directory
489	@pmlist = map { /^(.*)\.pm$/; "$dir/$1" }
490	grep(! -d "$podroot/$dir/$_" && /\.pm$/, readdir(DIR));
491
492	closedir(DIR);
493
494	# recurse on all subdirectories we kept track of
495	foreach $dir (@dirlist) {
496	installdir($dir, $recurse, $podroot, $splitdirs, $ignore);
497	}
498
499	# install all the pods we found
500	foreach $pod (@podlist) {
501	# check if we should ignore it.
502	next if grep($_ eq "$podroot/$pod.pod", @$ignore);
503
504	# check if a .pm files exists too
505	if (grep($_ eq "$pod.pm", @pmlist)) {
506	print "$0: Warning both `$podroot/$pod.pod' and "
507	. "`$podroot/$pod.pm' exist, using pod\n";
508	push(@ignore, "$pod.pm");
509	}
510	runpod2html("$pod.pod", $doindex);
511	}
512
513	# install all the .pm files we found
514	foreach $pm (@pmlist) {
515	# check if we should ignore it.
516	next if grep($_ eq "$pm.pm", @ignore);
517
518	runpod2html("$pm.pm", $doindex);
519	}
520	}
521
522
523	#
524	# runpod2html - invokes pod2html to convert a .pod or .pm file to a .html
525	# file.
526	#
527	sub runpod2html {
528	my($pod, $doindex) = @_;
529	my($html, $i, $dir, @dirs);
530
531	$html = $pod;
532	$html =~ s/\.(pod\|pm)$/.html/g;
533
534	# make sure the destination directories exist
535	@dirs = split("/", $html);
536	$dir = "$htmldir/";
537	for ($i = 0; $i < $#dirs; $i++) {
538	if (! -d "$dir$dirs[$i]") {
539	mkdir("$dir$dirs[$i]", 0755) \|\|
540	die "$0: error creating directory $dir$dirs[$i]: $!\n";
541	}
542	$dir .= "$dirs[$i]/";
543	}
544
545	# invoke pod2html
546	print "$podroot/$pod => $htmldir/$html\n" if $verbose;
547	#system("./pod2html",
548	Pod::Html'pod2html(
549	#Pod::Html'pod2html($pod2html,
550	"--htmlroot=$htmlroot",
551	"--podpath=".join(":", @podpath),
552	"--podroot=$podroot", "--netscape",
553	($doindex ? "--index" : "--noindex"),
554	"--" . ($recurse ? "" : "no") . "recurse",
555	($#libpods >= 0) ? "--libpods=" . join(":", @libpods) : "",
556	"--infile=$podroot/$pod", "--outfile=$htmldir/$html");
557	die "$0: error running $pod2html: $!\n" if $?;
558	}
559
560	sub htmlize { htmlify(0, @_) }