xref: /openbsd-src/gnu/usr.bin/perl/installhtml (revision b2ea75c1b17e1a9a339660e7ed45cd24946b230e)
1#!./perl -w
2
3# This file should really be extracted from a .PL file
4
5use lib 'lib';		# use source library if present
6
7use Config;		# for config options in the makefile
8use Getopt::Long;	# for command-line parsing
9use Cwd;
10use Pod::Html;
11
12=head1 NAME
13
14installhtml - converts a collection of POD pages to HTML format.
15
16=head1 SYNOPSIS
17
18    installhtml  [--help] [--podpath=<name>:...:<name>] [--podroot=<name>]
19         [--htmldir=<name>] [--htmlroot=<name>]  [--norecurse] [--recurse]
20         [--splithead=<name>,...,<name>]   [--splititem=<name>,...,<name>]
21	 [--libpods=<name>,...,<name>]  [--verbose]
22
23=head1 DESCRIPTION
24
25I<installhtml> converts a collection of POD pages to a corresponding
26collection of HTML pages.  This is primarily used to convert the pod
27pages found in the perl distribution.
28
29=head1 OPTIONS
30
31=over 4
32
33=item B<--help> help
34
35Displays the usage.
36
37=item B<--podroot> POD search path base directory
38
39The base directory to search for all .pod and .pm files to be converted.
40Default is current directory.
41
42=item B<--podpath> POD search path
43
44The list of directories to search for .pod and .pm files to be converted.
45Default is `podroot/.'.
46
47=item B<--recurse> recurse on subdirectories
48
49Whether or not to convert all .pm and .pod files found in subdirectories
50too.  Default is to not recurse.
51
52=item B<--htmldir> HTML destination directory
53
54The base directory which all HTML files will be written to.  This should
55be a path relative to the filesystem, not the resulting URL.
56
57=item B<--htmlroot> URL base directory
58
59The base directory which all resulting HTML files will be visible at in
60a URL.  The default is `/'.
61
62=item B<--splithead> POD files to split on =head directive
63
64Comma-separated list of pod files to split by the =head directive.  The
65.pod suffix is optional. These files should have names specified
66relative to podroot.
67
68=item B<--splititem> POD files to split on =item directive
69
70Comma-separated list of all pod files to split by the =item directive.
71The .pod suffix is optional.  I<installhtml> does not do the actual
72split, rather it invokes I<splitpod> to do the dirty work.  As with
73--splithead, these files should have names specified relative to podroot.
74
75=item B<--splitpod> Directory containing the splitpod program
76
77The directory containing the splitpod program. The default is `podroot/pod'.
78
79=item B<--libpods> library PODs for LE<lt>E<gt> links
80
81Comma-separated list of "library" pod files.  This is the same list that
82will be passed to pod2html when any pod is converted.
83
84=item B<--verbose> verbose output
85
86Self-explanatory.
87
88=back
89
90=head1 EXAMPLE
91
92The following command-line is an example of the one we use to convert
93perl documentation:
94
95    ./installhtml --podpath=lib:ext:pod:vms   \
96			--podroot=/usr/src/perl     \
97			--htmldir=/perl/nmanual     \
98			--htmlroot=/perl/nmanual    \
99			--splithead=pod/perlipc     \
100			--splititem=pod/perlfunc    \
101			--libpods=perlfunc,perlguts,perlvar,perlrun,perlop \
102			--recurse \
103			--verbose
104
105=head1 AUTHOR
106
107Chris Hall E<lt>hallc@cs.colorado.eduE<gt>
108
109=head1 TODO
110
111=cut
112
113$usage =<<END_OF_USAGE;
114Usage: $0 --help --podpath=<name>:...:<name> --podroot=<name>
115         --htmldir=<name> --htmlroot=<name> --norecurse --recurse
116         --splithead=<name>,...,<name> --splititem=<name>,...,<name>
117	 --libpods=<name>,...,<name> --verbose
118
119    --help      - this message
120    --podpath   - colon-separated list of directories containing .pod and
121                  .pm files to be converted (. by default).
122    --podroot   - filesystem base directory from which all relative paths in
123                  podpath stem (default is .).
124    --htmldir   - directory to store resulting html files in relative
125                  to the filesystem (\$podroot/html by default).
126    --htmlroot  - http-server base directory from which all relative paths
127                  in podpath stem (default is /).
128    --libpods   - comma-separated list of files to search for =item pod
129                  directives in as targets of C<> and implicit links (empty
130                  by default).
131    --norecurse - don't recurse on those subdirectories listed in podpath.
132                  (default behavior).
133    --recurse   - recurse on those subdirectories listed in podpath
134    --splithead - comma-separated list of .pod or .pm files to split.  will
135                  split each file into several smaller files at every occurrence
136                  of a pod =head[1-6] directive.
137    --splititem - comma-separated list of .pod or .pm files to split using
138                  splitpod.
139    --splitpod  - directory where the program splitpod can be found
140                  (\$podroot/pod by default).
141    --verbose   - self-explanatory.
142
143END_OF_USAGE
144
145@libpods = ();
146@podpath = ( "." );	# colon-separated list of directories containing .pod
147			# and .pm files to be converted.
148$podroot = ".";		# assume the pods we want are here
149$htmldir = "";		# nothing for now...
150$htmlroot = "/";	# default value
151$recurse = 0;		# default behavior
152@splithead = ();	# don't split any files by default
153@splititem = ();	# don't split any files by default
154$splitpod = "";		# nothing for now.
155
156$verbose = 0; 	    	# whether or not to print debugging info
157
158$pod2html = "pod/pod2html";
159
160usage("") unless @ARGV;
161
162# Overcome shell's p1,..,p8 limitation.
163# See vms/descrip_mms.template -> descrip.mms for invokation.
164if ( $^O eq 'VMS' ) { @ARGV = split(/\s+/,$ARGV[0]); }
165
166# parse the command-line
167$result = GetOptions( qw(
168	help
169	podpath=s
170	podroot=s
171	htmldir=s
172	htmlroot=s
173	libpods=s
174	recurse!
175	splithead=s
176	splititem=s
177	splitpod=s
178	verbose
179));
180usage("invalid parameters") unless $result;
181parse_command_line();
182
183
184# set these variables to appropriate values if the user didn't specify
185#  values for them.
186$htmldir = "$htmlroot/html" unless $htmldir;
187$splitpod = "$podroot/pod" unless $splitpod;
188
189
190# make sure that the destination directory exists
191(mkdir($htmldir, 0755) ||
192	die "$0: cannot make directory $htmldir: $!\n") if ! -d $htmldir;
193
194
195# the following array will eventually contain files that are to be
196# ignored in the conversion process.  these are files that have been
197# process by splititem or splithead and should not be converted as a
198# result.
199@ignore = ();
200
201
202# split pods.  its important to do this before convert ANY pods because
203#  it may effect some of the links
204@splitdirs = ();    # files in these directories won't get an index
205split_on_head($podroot, $htmldir, \@splitdirs, \@ignore, @splithead);
206split_on_item($podroot,           \@splitdirs, \@ignore, @splititem);
207
208
209# convert the pod pages found in @poddirs
210#warn "converting files\n" if $verbose;
211#warn "\@ignore\t= @ignore\n" if $verbose;
212foreach $dir (@podpath) {
213    installdir($dir, $recurse, $podroot, \@splitdirs, \@ignore);
214}
215
216
217# now go through and create master indices for each pod we split
218foreach $dir (@splititem) {
219    print "creating index $htmldir/$dir.html\n" if $verbose;
220    create_index("$htmldir/$dir.html", "$htmldir/$dir");
221}
222
223foreach $dir (@splithead) {
224    $dir .= ".pod" unless $dir =~ /(\.pod|\.pm)$/;
225    # let pod2html create the file
226    runpod2html($dir, 1);
227
228    # now go through and truncate after the index
229    $dir =~ /^(.*?)(\.pod|\.pm)?$/sm;
230    $file = "$htmldir/$1";
231    print "creating index $file.html\n" if $verbose;
232
233    # read in everything until what would have been the first =head
234    # directive, patching the index as we go.
235    open(H, "<$file.html") ||
236	die "$0: error opening $file.html for input: $!\n";
237    $/ = "";
238    @data = ();
239    while (<H>) {
240	last if /NAME=/;
241	$_ =~ s{HREF="#(.*)">}{
242	    my $url = "$file/$1.html" ;
243            $url = Pod::Html::relativize_url( $url, "$file.html" )
244                if ( ! defined $opt_htmlroot || $opt_htmlroot eq '' ) ;
245            "HREF=\"$url\">" ;
246        }eg;
247	push @data, $_;
248    }
249    close(H);
250
251    # now rewrite the file
252    open(H, ">$file.html") ||
253	die "$0: error opening $file.html for output: $!\n";
254    print H "@data\n";
255    close(H);
256}
257
258##############################################################################
259
260
261sub usage {
262    warn "$0: @_\n" if @_;
263    die $usage;
264}
265
266
267sub parse_command_line {
268    usage() if defined $opt_help;
269    $opt_help = ""; 	    	    # make -w shut up
270
271    # list of directories
272    @podpath   = split(":", $opt_podpath) if defined $opt_podpath;
273
274    # lists of files
275    @splithead = split(",", $opt_splithead) if defined $opt_splithead;
276    @splititem = split(",", $opt_splititem) if defined $opt_splititem;
277    @libpods   = split(",", $opt_libpods) if defined $opt_libpods;
278
279    $htmldir  = $opt_htmldir	    if defined $opt_htmldir;
280    $htmlroot = $opt_htmlroot	    if defined $opt_htmlroot;
281    $podroot  = $opt_podroot	    if defined $opt_podroot;
282    $splitpod = $opt_splitpod	    if defined $opt_splitpod;
283
284    $recurse  = $opt_recurse	    if defined $opt_recurse;
285    $verbose  = $opt_verbose	    if defined $opt_verbose;
286}
287
288
289sub absolute_path {
290    my($cwd, $path) = @_;
291	return "$cwd/$path" unless $path =~ m:/:;
292    # add cwd if path is not already an absolute path
293    $path = "$cwd/$path" if (substr($path,0,1) ne '/');
294    return $path;
295}
296
297
298sub create_index {
299    my($html, $dir) = @_;
300    my(@files, @filedata, @index, $file);
301    my($lcp1,$lcp2);
302
303
304    # get the list of .html files in this directory
305    opendir(DIR, $dir) ||
306	die "$0: error opening directory $dir for reading: $!\n";
307    @files = sort(grep(/\.html?$/, readdir(DIR)));
308    closedir(DIR);
309
310    open(HTML, ">$html") ||
311	die "$0: error opening $html for output: $!\n";
312
313    # for each .html file in the directory, extract the index
314    #	embedded in the file and throw it into the big index.
315    print HTML "<DL COMPACT>\n";
316    foreach $file (@files) {
317	$/ = "";
318
319	open(IN, "<$dir/$file") ||
320	    die "$0: error opening $dir/$file for input: $!\n";
321	@filedata = <IN>;
322	close(IN);
323
324	# pull out the NAME section
325	($name) = grep(/NAME=/, @filedata);
326	($lcp1,$lcp2) = ($name =~ m,/H1>\s(\S+)\s[\s-]*(.*?)\s*$,sm);
327	if (defined $lcp1 and $lcp1 eq '<P>') { # Uninteresting.  Try again.
328	   ($lcp1,$lcp2) = ($name =~ m,/H1>\s<P>\s(\S+)\s[\s-]*(.*?)\s*$,sm);
329	}
330	my $url= "$dir/$file" ;
331	if ( ! defined $opt_htmlroot || $opt_htmlroot eq '' ) {
332	    $url = Pod::Html::relativize_url( "$dir/$file", $html ) ;
333	}
334
335	print HTML qq(<A HREF="$url">);
336	print HTML "<DT>$lcp1</A><DD>$lcp2\n" if defined $lcp1;
337#	print HTML qq(<A HREF="$url">$lcp1</A><BR>\n") if defined $lcp1;
338
339	next;
340
341	@index = grep(/<!-- INDEX BEGIN -->.*<!-- INDEX END -->/s,
342		    @filedata);
343	for (@index) {
344	    s/<!-- INDEX BEGIN -->(\s*<!--)(.*)(-->\s*)<!-- INDEX END -->/$lcp2/s;
345	    s,#,$dir/$file#,g;
346	    # print HTML "$_\n";
347	    print HTML "$_\n<P><HR><P>\n";
348	}
349    }
350    print HTML "</DL>\n";
351
352    close(HTML);
353}
354
355
356sub split_on_head {
357    my($podroot, $htmldir, $splitdirs, $ignore, @splithead) = @_;
358    my($pod, $dirname, $filename);
359
360    # split the files specified in @splithead on =head[1-6] pod directives
361    print "splitting files by head.\n" if $verbose && $#splithead >= 0;
362    foreach $pod (@splithead) {
363	# figure out the directory name and filename
364	$pod      =~ s,^([^/]*)$,/$1,;
365	$pod      =~ m,(.*?)/(.*?)(\.pod)?$,;
366	$dirname  = $1;
367	$filename = "$2.pod";
368
369	# since we are splitting this file it shouldn't be converted.
370	push(@$ignore, "$podroot/$dirname/$filename");
371
372	# split the pod
373	splitpod("$podroot/$dirname/$filename", "$podroot/$dirname", $htmldir,
374	    $splitdirs);
375    }
376}
377
378
379sub split_on_item {
380    my($podroot, $splitdirs, $ignore, @splititem) = @_;
381    my($pwd, $dirname, $filename);
382
383    print "splitting files by item.\n" if $verbose && $#splititem >= 0;
384    $pwd = getcwd();
385	my $splitter = absolute_path($pwd, "$splitpod/splitpod");
386    foreach $pod (@splititem) {
387	# figure out the directory to split into
388	$pod      =~ s,^([^/]*)$,/$1,;
389	$pod      =~ m,(.*?)/(.*?)(\.pod)?$,;
390	$dirname  = "$1/$2";
391	$filename = "$2.pod";
392
393	# since we are splitting this file it shouldn't be converted.
394	push(@$ignore, "$podroot/$dirname.pod");
395
396	# split the pod
397	push(@$splitdirs, "$podroot/$dirname");
398	if (! -d "$podroot/$dirname") {
399	    mkdir("$podroot/$dirname", 0755) ||
400		    die "$0: error creating directory $podroot/$dirname: $!\n";
401	}
402	chdir("$podroot/$dirname") ||
403	    die "$0: error changing to directory $podroot/$dirname: $!\n";
404	die "$splitter not found. Use '-splitpod dir' option.\n"
405	    unless -f $splitter;
406	system("perl", $splitter, "../$filename") &&
407	    warn "$0: error running '$splitter ../$filename'"
408		 ." from $podroot/$dirname";
409    }
410    chdir($pwd);
411}
412
413
414#
415# splitpod - splits a .pod file into several smaller .pod files
416#  where a new file is started each time a =head[1-6] pod directive
417#  is encountered in the input file.
418#
419sub splitpod {
420    my($pod, $poddir, $htmldir, $splitdirs) = @_;
421    my(@poddata, @filedata, @heads);
422    my($file, $i, $j, $prevsec, $section, $nextsec);
423
424    print "splitting $pod\n" if $verbose;
425
426    # read the file in paragraphs
427    $/ = "";
428    open(SPLITIN, "<$pod") ||
429	die "$0: error opening $pod for input: $!\n";
430    @filedata = <SPLITIN>;
431    close(SPLITIN) ||
432	die "$0: error closing $pod: $!\n";
433
434    # restore the file internally by =head[1-6] sections
435    @poddata = ();
436    for ($i = 0, $j = -1; $i <= $#filedata; $i++) {
437	$j++ if ($filedata[$i] =~ /^\s*=head[1-6]/);
438	if ($j >= 0) {
439	    $poddata[$j]  = "" unless defined $poddata[$j];
440	    $poddata[$j] .= "\n$filedata[$i]" if $j >= 0;
441	}
442    }
443
444    # create list of =head[1-6] sections so that we can rewrite
445    #  L<> links as necessary.
446    %heads = ();
447    foreach $i (0..$#poddata) {
448	$heads{htmlize($1)} = 1 if $poddata[$i] =~ /=head[1-6]\s+(.*)/;
449    }
450
451    # create a directory of a similar name and store all the
452    #  files in there
453    $pod =~ s,.*/(.*),$1,;	# get the last part of the name
454    $dir = $pod;
455    $dir =~ s/\.pod//g;
456    push(@$splitdirs, "$poddir/$dir");
457    mkdir("$poddir/$dir", 0755) ||
458	die "$0: could not create directory $poddir/$dir: $!\n"
459	unless -d "$poddir/$dir";
460
461    $poddata[0] =~ /^\s*=head[1-6]\s+(.*)/;
462    $section    = "";
463    $nextsec    = $1;
464
465    # for each section of the file create a separate pod file
466    for ($i = 0; $i <= $#poddata; $i++) {
467	# determine the "prev" and "next" links
468	$prevsec = $section;
469	$section = $nextsec;
470	if ($i < $#poddata) {
471	    $poddata[$i+1] =~ /^\s*=head[1-6]\s+(.*)/;
472	    $nextsec       = $1;
473	} else {
474	    $nextsec = "";
475	}
476
477	# determine an appropriate filename (this must correspond with
478	#  what pod2html will try and guess)
479	# $poddata[$i] =~ /^\s*=head[1-6]\s+(.*)/;
480	$file = "$dir/" . htmlize($section) . ".pod";
481
482	# create the new .pod file
483	print "\tcreating $poddir/$file\n" if $verbose;
484	open(SPLITOUT, ">$poddir/$file") ||
485	    die "$0: error opening $poddir/$file for output: $!\n";
486	$poddata[$i] =~ s,L<([^<>]*)>,
487			    defined $heads{htmlize($1)} ? "L<$dir/$1>" : "L<$1>"
488			 ,ge;
489	print SPLITOUT $poddata[$i]."\n\n";
490	print SPLITOUT "=over 4\n\n";
491	print SPLITOUT "=item *\n\nBack to L<$dir/\"$prevsec\">\n\n" if $prevsec;
492	print SPLITOUT "=item *\n\nForward to L<$dir/\"$nextsec\">\n\n" if $nextsec;
493	print SPLITOUT "=item *\n\nUp to L<$dir>\n\n";
494	print SPLITOUT "=back\n\n";
495	close(SPLITOUT) ||
496	    die "$0: error closing $poddir/$file: $!\n";
497    }
498}
499
500
501#
502# installdir - takes care of converting the .pod and .pm files in the
503#  current directory to .html files and then installing those.
504#
505sub installdir {
506    my($dir, $recurse, $podroot, $splitdirs, $ignore) = @_;
507    my(@dirlist, @podlist, @pmlist, $doindex);
508
509    @dirlist = ();	# directories to recurse on
510    @podlist = ();	# .pod files to install
511    @pmlist  = ();	# .pm files to install
512
513    # should files in this directory get an index?
514    $doindex = (grep($_ eq "$podroot/$dir", @$splitdirs) ? 0 : 1);
515
516    opendir(DIR, "$podroot/$dir")
517	|| die "$0: error opening directory $podroot/$dir: $!\n";
518
519    # find the directories to recurse on
520    @dirlist = map { if ($^O eq 'VMS') {/^(.*)\.dir$/i; "$dir/$1";} else {"$dir/$_";}}
521	grep(-d "$podroot/$dir/$_" && !/^\.{1,2}/, readdir(DIR)) if $recurse;
522    rewinddir(DIR);
523
524    # find all the .pod files within the directory
525    @podlist = map { /^(.*)\.pod$/; "$dir/$1" }
526	grep(! -d "$podroot/$dir/$_" && /\.pod$/, readdir(DIR));
527    rewinddir(DIR);
528
529    # find all the .pm files within the directory
530    @pmlist = map { /^(.*)\.pm$/; "$dir/$1" }
531	grep(! -d "$podroot/$dir/$_" && /\.pm$/, readdir(DIR));
532
533    closedir(DIR);
534
535    # recurse on all subdirectories we kept track of
536    foreach $dir (@dirlist) {
537	installdir($dir, $recurse, $podroot, $splitdirs, $ignore);
538    }
539
540    # install all the pods we found
541    foreach $pod (@podlist) {
542	# check if we should ignore it.
543	next if grep($_ eq "$podroot/$pod.pod", @$ignore);
544
545	# check if a .pm files exists too
546	if (grep($_ eq "$pod.pm", @pmlist)) {
547	    print  "$0: Warning both `$podroot/$pod.pod' and "
548		. "`$podroot/$pod.pm' exist, using pod\n";
549	    push(@ignore, "$pod.pm");
550	}
551	runpod2html("$pod.pod", $doindex);
552    }
553
554    # install all the .pm files we found
555    foreach $pm (@pmlist) {
556	# check if we should ignore it.
557	next if grep($_ eq "$pm.pm", @ignore);
558
559	runpod2html("$pm.pm", $doindex);
560    }
561}
562
563
564#
565# runpod2html - invokes pod2html to convert a .pod or .pm file to a .html
566#  file.
567#
568sub runpod2html {
569    my($pod, $doindex) = @_;
570    my($html, $i, $dir, @dirs);
571
572    $html = $pod;
573    $html =~ s/\.(pod|pm)$/.html/g;
574
575    # make sure the destination directories exist
576    @dirs = split("/", $html);
577    $dir  = "$htmldir/";
578    for ($i = 0; $i < $#dirs; $i++) {
579	if (! -d "$dir$dirs[$i]") {
580	    mkdir("$dir$dirs[$i]", 0755) ||
581		die "$0: error creating directory $dir$dirs[$i]: $!\n";
582	}
583	$dir .= "$dirs[$i]/";
584    }
585
586    # invoke pod2html
587    print "$podroot/$pod => $htmldir/$html\n" if $verbose;
588#system("./pod2html",
589        Pod::Html'pod2html(
590        #Pod::Html'pod2html($pod2html,
591        "--htmldir=$htmldir",
592	"--htmlroot=$htmlroot",
593	"--podpath=".join(":", @podpath),
594	"--podroot=$podroot", "--netscape",
595	"--header",
596	($doindex ? "--index" : "--noindex"),
597	"--" . ($recurse ? "" : "no") . "recurse",
598	($#libpods >= 0) ? "--libpods=" . join(":", @libpods) : "",
599	"--infile=$podroot/$pod", "--outfile=$htmldir/$html");
600    die "$0: error running $pod2html: $!\n" if $?;
601}
602
603sub htmlize { htmlify(0, @_) }
604