xref: /openbsd-src/gnu/usr.bin/perl/installhtml (revision 8500990981f885cbe5e6a4958549cacc238b5ae6)
1#!./perl -Ilib -w
2
3# This file should really be extracted from a .PL file
4
5use strict;
6use Config;		# for config options in the makefile
7use Getopt::Long;	# for command-line parsing
8use Cwd;
9use Pod::Html;
10
11=head1 NAME
12
13installhtml - converts a collection of POD pages to HTML format.
14
15=head1 SYNOPSIS
16
17    installhtml  [--help] [--podpath=<name>:...:<name>] [--podroot=<name>]
18         [--htmldir=<name>] [--htmlroot=<name>]  [--norecurse] [--recurse]
19         [--splithead=<name>,...,<name>]   [--splititem=<name>,...,<name>]
20	 [--libpods=<name>,...,<name>]  [--verbose]
21
22=head1 DESCRIPTION
23
24I<installhtml> converts a collection of POD pages to a corresponding
25collection of HTML pages.  This is primarily used to convert the pod
26pages found in the perl distribution.
27
28=head1 OPTIONS
29
30=over 4
31
32=item B<--help> help
33
34Displays the usage.
35
36=item B<--podroot> POD search path base directory
37
38The base directory to search for all .pod and .pm files to be converted.
39Default is current directory.
40
41=item B<--podpath> POD search path
42
43The list of directories to search for .pod and .pm files to be converted.
44Default is `podroot/.'.
45
46=item B<--recurse> recurse on subdirectories
47
48Whether or not to convert all .pm and .pod files found in subdirectories
49too.  Default is to not recurse.
50
51=item B<--htmldir> HTML destination directory
52
53The base directory which all HTML files will be written to.  This should
54be a path relative to the filesystem, not the resulting URL.
55
56=item B<--htmlroot> URL base directory
57
58The base directory which all resulting HTML files will be visible at in
59a URL.  The default is `/'.
60
61=item B<--splithead> POD files to split on =head directive
62
63Comma-separated list of pod files to split by the =head directive.  The
64.pod suffix is optional. These files should have names specified
65relative to podroot.
66
67=item B<--splititem> POD files to split on =item directive
68
69Comma-separated list of all pod files to split by the =item directive.
70The .pod suffix is optional.  I<installhtml> does not do the actual
71split, rather it invokes I<splitpod> to do the dirty work.  As with
72--splithead, these files should have names specified relative to podroot.
73
74=item B<--splitpod> Directory containing the splitpod program
75
76The directory containing the splitpod program. The default is `podroot/pod'.
77
78=item B<--libpods> library PODs for LE<lt>E<gt> links
79
80Comma-separated list of "library" pod files.  This is the same list that
81will be passed to pod2html when any pod is converted.
82
83=item B<--verbose> verbose output
84
85Self-explanatory.
86
87=back
88
89=head1 EXAMPLE
90
91The following command-line is an example of the one we use to convert
92perl documentation:
93
94    ./installhtml --podpath=lib:ext:pod:vms   \
95			--podroot=/usr/src/perl     \
96			--htmldir=/perl/nmanual     \
97			--htmlroot=/perl/nmanual    \
98			--splithead=pod/perlipc     \
99			--splititem=pod/perlfunc    \
100			--libpods=perlfunc,perlguts,perlvar,perlrun,perlop \
101			--recurse \
102			--verbose
103
104=head1 AUTHOR
105
106Chris Hall E<lt>hallc@cs.colorado.eduE<gt>
107
108=head1 TODO
109
110=cut
111
112my $usage;
113
114$usage =<<END_OF_USAGE;
115Usage: $0 --help --podpath=<name>:...:<name> --podroot=<name>
116         --htmldir=<name> --htmlroot=<name> --norecurse --recurse
117         --splithead=<name>,...,<name> --splititem=<name>,...,<name>
118	 --libpods=<name>,...,<name> --verbose
119
120    --help      - this message
121    --podpath   - colon-separated list of directories containing .pod and
122                  .pm files to be converted (. by default).
123    --podroot   - filesystem base directory from which all relative paths in
124                  podpath stem (default is .).
125    --htmldir   - directory to store resulting html files in relative
126                  to the filesystem (\$podroot/html by default).
127    --htmlroot  - http-server base directory from which all relative paths
128                  in podpath stem (default is /).
129    --libpods   - comma-separated list of files to search for =item pod
130                  directives in as targets of C<> and implicit links (empty
131                  by default).
132    --norecurse - don't recurse on those subdirectories listed in podpath.
133                  (default behavior).
134    --recurse   - recurse on those subdirectories listed in podpath
135    --splithead - comma-separated list of .pod or .pm files to split.  will
136                  split each file into several smaller files at every occurrence
137                  of a pod =head[1-6] directive.
138    --splititem - comma-separated list of .pod or .pm files to split using
139                  splitpod.
140    --splitpod  - directory where the program splitpod can be found
141                  (\$podroot/pod by default).
142    --verbose   - self-explanatory.
143
144END_OF_USAGE
145
146my (@libpods, @podpath, $podroot, $htmldir, $htmlroot, $recurse, @splithead,
147    @splititem, $splitpod, $verbose, $pod2html);
148
149@libpods = ();
150@podpath = ( "." );	# colon-separated list of directories containing .pod
151			# and .pm files to be converted.
152$podroot = ".";		# assume the pods we want are here
153$htmldir = "";		# nothing for now...
154$htmlroot = "/";	# default value
155$recurse = 0;		# default behavior
156@splithead = ();	# don't split any files by default
157@splititem = ();	# don't split any files by default
158$splitpod = "";		# nothing for now.
159
160$verbose = 0; 	    	# whether or not to print debugging info
161
162$pod2html = "pod/pod2html";
163
164usage("") unless @ARGV;
165
166# Overcome shell's p1,..,p8 limitation.
167# See vms/descrip_mms.template -> descrip.mms for invokation.
168if ( $^O eq 'VMS' ) { @ARGV = split(/\s+/,$ARGV[0]); }
169
170use vars qw( %Options );
171
172# parse the command-line
173my $result = GetOptions( \%Options, qw(
174	help
175	podpath=s
176	podroot=s
177	htmldir=s
178	htmlroot=s
179	libpods=s
180	recurse!
181	splithead=s
182	splititem=s
183	splitpod=s
184	verbose
185));
186usage("invalid parameters") unless $result;
187parse_command_line();
188
189
190# set these variables to appropriate values if the user didn't specify
191#  values for them.
192$htmldir = "$htmlroot/html" unless $htmldir;
193$splitpod = "$podroot/pod" unless $splitpod;
194
195
196# make sure that the destination directory exists
197(mkdir($htmldir, 0755) ||
198	die "$0: cannot make directory $htmldir: $!\n") if ! -d $htmldir;
199
200
201# the following array will eventually contain files that are to be
202# ignored in the conversion process.  these are files that have been
203# process by splititem or splithead and should not be converted as a
204# result.
205my @ignore = ();
206my @splitdirs;
207
208# split pods.  its important to do this before convert ANY pods because
209#  it may effect some of the links
210@splitdirs = ();    # files in these directories won't get an index
211split_on_head($podroot, $htmldir, \@splitdirs, \@ignore, @splithead);
212split_on_item($podroot,           \@splitdirs, \@ignore, @splititem);
213
214
215# convert the pod pages found in @poddirs
216#warn "converting files\n" if $verbose;
217#warn "\@ignore\t= @ignore\n" if $verbose;
218foreach my $dir (@podpath) {
219    installdir($dir, $recurse, $podroot, \@splitdirs, \@ignore);
220}
221
222
223# now go through and create master indices for each pod we split
224foreach my $dir (@splititem) {
225    print "creating index $htmldir/$dir.html\n" if $verbose;
226    create_index("$htmldir/$dir.html", "$htmldir/$dir");
227}
228
229foreach my $dir (@splithead) {
230    $dir .= ".pod" unless $dir =~ /(\.pod|\.pm)$/;
231    # let pod2html create the file
232    runpod2html($dir, 1);
233
234    # now go through and truncate after the index
235    $dir =~ /^(.*?)(\.pod|\.pm)?$/sm;
236    my $file = "$htmldir/$1";
237    print "creating index $file.html\n" if $verbose;
238
239    # read in everything until what would have been the first =head
240    # directive, patching the index as we go.
241    open(H, "<$file.html") ||
242	die "$0: error opening $file.html for input: $!\n";
243    $/ = "";
244    my @data = ();
245    while (<H>) {
246	last if /NAME=/;
247	$_ =~ s{HREF="#(.*)">}{
248	    my $url = "$file/$1.html" ;
249            $url = Pod::Html::relativize_url( $url, "$file.html" )
250                if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' );
251            "HREF=\"$url\">" ;
252        }eg;
253	push @data, $_;
254    }
255    close(H);
256
257    # now rewrite the file
258    open(H, ">$file.html") ||
259	die "$0: error opening $file.html for output: $!\n";
260    print H "@data", "\n";
261    close(H);
262}
263
264##############################################################################
265
266
267sub usage {
268    warn "$0: @_\n" if @_;
269    die $usage;
270}
271
272
273sub parse_command_line {
274    usage() if defined $Options{help};
275    $Options{help} = ""; 	    	    # make -w shut up
276
277    # list of directories
278    @podpath   = split(":", $Options{podpath}) if defined $Options{podpath};
279
280    # lists of files
281    @splithead = split(",", $Options{splithead}) if defined $Options{splithead};
282    @splititem = split(",", $Options{splititem}) if defined $Options{splititem};
283    @libpods   = split(",", $Options{libpods}) if defined $Options{libpods};
284
285    $htmldir  = $Options{htmldir}	    if defined $Options{htmldir};
286    $htmlroot = $Options{htmlroot}	    if defined $Options{htmlroot};
287    $podroot  = $Options{podroot}	    if defined $Options{podroot};
288    $splitpod = $Options{splitpod}	    if defined $Options{splitpod};
289
290    $recurse  = $Options{recurse}	    if defined $Options{recurse};
291    $verbose  = $Options{verbose}	    if defined $Options{verbose};
292}
293
294
295sub absolute_path {
296    my($cwd, $path) = @_;
297	return "$cwd/$path" unless $path =~ m:/:;
298    # add cwd if path is not already an absolute path
299    $path = "$cwd/$path" if (substr($path,0,1) ne '/');
300    return $path;
301}
302
303
304sub create_index {
305    my($html, $dir) = @_;
306    my(@files, @filedata, @index, $file);
307    my($lcp1,$lcp2);
308
309
310    # get the list of .html files in this directory
311    opendir(DIR, $dir) ||
312	die "$0: error opening directory $dir for reading: $!\n";
313    @files = sort(grep(/\.html?$/, readdir(DIR)));
314    closedir(DIR);
315
316    open(HTML, ">$html") ||
317	die "$0: error opening $html for output: $!\n";
318
319    # for each .html file in the directory, extract the index
320    #	embedded in the file and throw it into the big index.
321    print HTML "<DL COMPACT>\n";
322    foreach $file (@files) {
323	$/ = "";
324
325	open(IN, "<$dir/$file") ||
326	    die "$0: error opening $dir/$file for input: $!\n";
327	@filedata = <IN>;
328	close(IN);
329
330	# pull out the NAME section
331  my $name;
332	($name) = grep(/NAME=/, @filedata);
333	($lcp1,$lcp2) = ($name =~ m,/H1>\s(\S+)\s[\s-]*(.*?)\s*$,sm);
334	if (defined $lcp1 and $lcp1 eq '<P>') { # Uninteresting.  Try again.
335	   ($lcp1,$lcp2) = ($name =~ m,/H1>\s<P>\s(\S+)\s[\s-]*(.*?)\s*$,sm);
336	}
337	my $url= "$dir/$file" ;
338	if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' ) {
339	    $url = Pod::Html::relativize_url( "$dir/$file", $html ) ;
340	}
341
342	print HTML qq(<A HREF="$url">);
343	print HTML "<DT>$lcp1</A><DD>$lcp2\n" if defined $lcp1;
344#	print HTML qq(<A HREF="$url">$lcp1</A><BR>\n") if defined $lcp1;
345
346	next;
347
348	@index = grep(/<!-- INDEX BEGIN -->.*<!-- INDEX END -->/s,
349		    @filedata);
350	for (@index) {
351	    s/<!-- INDEX BEGIN -->(\s*<!--)(.*)(-->\s*)<!-- INDEX END -->/$lcp2/s;
352	    s,#,$dir/$file#,g;
353	    # print HTML "$_\n";
354	    print HTML "$_\n<P><HR><P>\n";
355	}
356    }
357    print HTML "</DL>\n";
358
359    close(HTML);
360}
361
362
363sub split_on_head {
364    my($podroot, $htmldir, $splitdirs, $ignore, @splithead) = @_;
365    my($pod, $dirname, $filename);
366
367    # split the files specified in @splithead on =head[1-6] pod directives
368    print "splitting files by head.\n" if $verbose && $#splithead >= 0;
369    foreach $pod (@splithead) {
370	# figure out the directory name and filename
371	$pod      =~ s,^([^/]*)$,/$1,;
372	$pod      =~ m,(.*?)/(.*?)(\.pod)?$,;
373	$dirname  = $1;
374	$filename = "$2.pod";
375
376	# since we are splitting this file it shouldn't be converted.
377	push(@$ignore, "$podroot/$dirname/$filename");
378
379	# split the pod
380	splitpod("$podroot/$dirname/$filename", "$podroot/$dirname", $htmldir,
381	    $splitdirs);
382    }
383}
384
385
386sub split_on_item {
387    my($podroot, $splitdirs, $ignore, @splititem) = @_;
388    my($pwd, $dirname, $filename);
389
390    print "splitting files by item.\n" if $verbose && $#splititem >= 0;
391    $pwd = getcwd();
392	my $splitter = absolute_path($pwd, "$splitpod/splitpod");
393    foreach my $pod (@splititem) {
394	# figure out the directory to split into
395	$pod      =~ s,^([^/]*)$,/$1,;
396	$pod      =~ m,(.*?)/(.*?)(\.pod)?$,;
397	$dirname  = "$1/$2";
398	$filename = "$2.pod";
399
400	# since we are splitting this file it shouldn't be converted.
401	push(@$ignore, "$podroot/$dirname.pod");
402
403	# split the pod
404	push(@$splitdirs, "$podroot/$dirname");
405	if (! -d "$podroot/$dirname") {
406	    mkdir("$podroot/$dirname", 0755) ||
407		    die "$0: error creating directory $podroot/$dirname: $!\n";
408	}
409	chdir("$podroot/$dirname") ||
410	    die "$0: error changing to directory $podroot/$dirname: $!\n";
411	die "$splitter not found. Use '-splitpod dir' option.\n"
412	    unless -f $splitter;
413	system("perl", $splitter, "../$filename") &&
414	    warn "$0: error running '$splitter ../$filename'"
415		 ." from $podroot/$dirname";
416    }
417    chdir($pwd);
418}
419
420
421#
422# splitpod - splits a .pod file into several smaller .pod files
423#  where a new file is started each time a =head[1-6] pod directive
424#  is encountered in the input file.
425#
426sub splitpod {
427    my($pod, $poddir, $htmldir, $splitdirs) = @_;
428    my(@poddata, @filedata, @heads);
429    my($file, $i, $j, $prevsec, $section, $nextsec);
430
431    print "splitting $pod\n" if $verbose;
432
433    # read the file in paragraphs
434    $/ = "";
435    open(SPLITIN, "<$pod") ||
436	die "$0: error opening $pod for input: $!\n";
437    @filedata = <SPLITIN>;
438    close(SPLITIN) ||
439	die "$0: error closing $pod: $!\n";
440
441    # restore the file internally by =head[1-6] sections
442    @poddata = ();
443    for ($i = 0, $j = -1; $i <= $#filedata; $i++) {
444	$j++ if ($filedata[$i] =~ /^\s*=head[1-6]/);
445	if ($j >= 0) {
446	    $poddata[$j]  = "" unless defined $poddata[$j];
447	    $poddata[$j] .= "\n$filedata[$i]" if $j >= 0;
448	}
449    }
450
451    # create list of =head[1-6] sections so that we can rewrite
452    #  L<> links as necessary.
453    my %heads = ();
454    foreach $i (0..$#poddata) {
455	$heads{htmlize($1)} = 1 if $poddata[$i] =~ /=head[1-6]\s+(.*)/;
456    }
457
458    # create a directory of a similar name and store all the
459    #  files in there
460    $pod =~ s,.*/(.*),$1,;	# get the last part of the name
461    my $dir = $pod;
462    $dir =~ s/\.pod//g;
463    push(@$splitdirs, "$poddir/$dir");
464    mkdir("$poddir/$dir", 0755) ||
465	die "$0: could not create directory $poddir/$dir: $!\n"
466	unless -d "$poddir/$dir";
467
468    $poddata[0] =~ /^\s*=head[1-6]\s+(.*)/;
469    $section    = "";
470    $nextsec    = $1;
471
472    # for each section of the file create a separate pod file
473    for ($i = 0; $i <= $#poddata; $i++) {
474	# determine the "prev" and "next" links
475	$prevsec = $section;
476	$section = $nextsec;
477	if ($i < $#poddata) {
478	    $poddata[$i+1] =~ /^\s*=head[1-6]\s+(.*)/;
479	    $nextsec       = $1;
480	} else {
481	    $nextsec = "";
482	}
483
484	# determine an appropriate filename (this must correspond with
485	#  what pod2html will try and guess)
486	# $poddata[$i] =~ /^\s*=head[1-6]\s+(.*)/;
487	$file = "$dir/" . htmlize($section) . ".pod";
488
489	# create the new .pod file
490	print "\tcreating $poddir/$file\n" if $verbose;
491	open(SPLITOUT, ">$poddir/$file") ||
492	    die "$0: error opening $poddir/$file for output: $!\n";
493	$poddata[$i] =~ s,L<([^<>]*)>,
494			    defined $heads{htmlize($1)} ? "L<$dir/$1>" : "L<$1>"
495			 ,ge;
496	print SPLITOUT $poddata[$i]."\n\n";
497	print SPLITOUT "=over 4\n\n";
498	print SPLITOUT "=item *\n\nBack to L<$dir/\"$prevsec\">\n\n" if $prevsec;
499	print SPLITOUT "=item *\n\nForward to L<$dir/\"$nextsec\">\n\n" if $nextsec;
500	print SPLITOUT "=item *\n\nUp to L<$dir>\n\n";
501	print SPLITOUT "=back\n\n";
502	close(SPLITOUT) ||
503	    die "$0: error closing $poddir/$file: $!\n";
504    }
505}
506
507
508#
509# installdir - takes care of converting the .pod and .pm files in the
510#  current directory to .html files and then installing those.
511#
512sub installdir {
513    my($dir, $recurse, $podroot, $splitdirs, $ignore) = @_;
514    my(@dirlist, @podlist, @pmlist, $doindex);
515
516    @dirlist = ();	# directories to recurse on
517    @podlist = ();	# .pod files to install
518    @pmlist  = ();	# .pm files to install
519
520    # should files in this directory get an index?
521    $doindex = (grep($_ eq "$podroot/$dir", @$splitdirs) ? 0 : 1);
522
523    opendir(DIR, "$podroot/$dir")
524	|| die "$0: error opening directory $podroot/$dir: $!\n";
525
526    # find the directories to recurse on
527    @dirlist = map { if ($^O eq 'VMS') {/^(.*)\.dir$/i; "$dir/$1";} else {"$dir/$_";}}
528	grep(-d "$podroot/$dir/$_" && !/^\.{1,2}/, readdir(DIR)) if $recurse;
529    rewinddir(DIR);
530
531    # find all the .pod files within the directory
532    @podlist = map { /^(.*)\.pod$/; "$dir/$1" }
533	grep(! -d "$podroot/$dir/$_" && /\.pod$/, readdir(DIR));
534    rewinddir(DIR);
535
536    # find all the .pm files within the directory
537    @pmlist = map { /^(.*)\.pm$/; "$dir/$1" }
538	grep(! -d "$podroot/$dir/$_" && /\.pm$/, readdir(DIR));
539
540    closedir(DIR);
541
542    # recurse on all subdirectories we kept track of
543    foreach $dir (@dirlist) {
544	installdir($dir, $recurse, $podroot, $splitdirs, $ignore);
545    }
546
547    # install all the pods we found
548    foreach my $pod (@podlist) {
549	# check if we should ignore it.
550	next if grep($_ eq "$podroot/$pod.pod", @$ignore);
551
552	# check if a .pm files exists too
553	if (grep($_ eq "$pod.pm", @pmlist)) {
554	    print  "$0: Warning both `$podroot/$pod.pod' and "
555		. "`$podroot/$pod.pm' exist, using pod\n";
556	    push(@ignore, "$pod.pm");
557	}
558	runpod2html("$pod.pod", $doindex);
559    }
560
561    # install all the .pm files we found
562    foreach my $pm (@pmlist) {
563	# check if we should ignore it.
564	next if grep($_ eq "$pm.pm", @ignore);
565
566	runpod2html("$pm.pm", $doindex);
567    }
568}
569
570
571#
572# runpod2html - invokes pod2html to convert a .pod or .pm file to a .html
573#  file.
574#
575sub runpod2html {
576    my($pod, $doindex) = @_;
577    my($html, $i, $dir, @dirs);
578
579    $html = $pod;
580    $html =~ s/\.(pod|pm)$/.html/g;
581
582    # make sure the destination directories exist
583    @dirs = split("/", $html);
584    $dir  = "$htmldir/";
585    for ($i = 0; $i < $#dirs; $i++) {
586	if (! -d "$dir$dirs[$i]") {
587	    mkdir("$dir$dirs[$i]", 0755) ||
588		die "$0: error creating directory $dir$dirs[$i]: $!\n";
589	}
590	$dir .= "$dirs[$i]/";
591    }
592
593    # invoke pod2html
594    print "$podroot/$pod => $htmldir/$html\n" if $verbose;
595#system("./pod2html",
596        Pod::Html'pod2html(
597        #Pod::Html'pod2html($pod2html,
598        "--htmldir=$htmldir",
599	"--htmlroot=$htmlroot",
600	"--podpath=".join(":", @podpath),
601	"--podroot=$podroot", "--netscape",
602	"--header",
603	($doindex ? "--index" : "--noindex"),
604	"--" . ($recurse ? "" : "no") . "recurse",
605	($#libpods >= 0) ? "--libpods=" . join(":", @libpods) : "",
606	"--infile=$podroot/$pod", "--outfile=$htmldir/$html");
607    die "$0: error running $pod2html: $!\n" if $?;
608}
609
610sub htmlize { htmlify(0, @_) }
611