xref: /openbsd/gnu/usr.bin/perl/installhtml (revision 3cab2bb3) 
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 File::Path qw(remove_tree);
8use File::Spec::Functions qw(rel2abs no_upwards);
9use Getopt::Long;	# for command-line parsing
10use Cwd;
11use Pod::Html 1.23 qw(anchorify relativize_url);
12
13=head1 NAME
14
15installhtml - converts a collection of POD pages to HTML format.
16
17=head1 SYNOPSIS
18
19  installhtml  [--help] [--podpath=<name>:...:<name>] [--podroot=<name>]
20       [--htmldir=<name>] [--htmlroot=<name>]  [--norecurse] [--recurse]
21       [--splithead=<name>,...,<name>]   [--splititem=<name>,...,<name>]
22       [--ignore=<name>,...,<name>]  [--verbose]
23
24=head1 DESCRIPTION
25
26I<installhtml> converts a collection of POD pages to a corresponding
27collection of HTML pages.  This is used to convert the pod pages found in the
28perl distribution.  (It is not intended as a general-purpose
29converter/installer of POD pages in HTML format.  See L<Pod::Html>.)
30
31=head1 OPTIONS
32
33=over 4
34
35=item B<--help> help
36
37Displays the usage.
38
39=item B<--podroot> POD search path base directory
40
41The base directory to search for all .pod and .pm files to be converted.
42Default is current directory.
43
44=item B<--podpath> POD search path
45
46The list of directories to search for .pod and .pm files to be converted.
47Default is 'podroot/.'.
48
49=item B<--recurse> recurse on subdirectories
50
51Whether or not to convert all .pm and .pod files found in subdirectories
52too.  Default is to not recurse.
53
54=item B<--htmldir> HTML destination directory
55
56The base directory which all HTML files will be written to.  This should
57be a path relative to the filesystem, not the resulting URL.
58
59=item B<--htmlroot> URL base directory
60
61The base directory which all resulting HTML files will be visible at in
62a URL.  The default is '/'.
63
64=item B<--splithead> POD files to split on =head directive
65
66Comma-separated list of pod files to split by the =head directive.  The
67.pod suffix is optional. These files should have names specified
68relative to podroot.
69
70=item B<--splititem> POD files to split on =item directive
71
72Comma-separated list of all pod files to split by the =item directive.  The
73.pod suffix is optional.  I<installhtml> does not do the actual split, rather
74it invokes I<splitpod>, a separate program in the Perl 5 core distribution,
75to do the dirty work.  As with --splithead, these files should have names
76specified relative to podroot.
77
78=item B<--splitpod> Directory containing the splitpod program
79
80The directory containing the splitpod program. The default is 'podroot/pod'.
81
82=item B<--ignore> files to be ignored
83
84Comma-separated of files that shouldn't be installed, given relative
85to podroot.
86
87=item B<--verbose> verbose output
88
89Self-explanatory.
90
91=back
92
93=head1 EXAMPLE
94
95The following command-line is an example of the one we use to convert
96perl documentation:
97
98    ./installhtml --podpath=lib:ext:pod:vms   \
99			--podroot=/usr/src/perl     \
100			--htmldir=/perl/nmanual     \
101			--htmlroot=/perl/nmanual    \
102			--splithead=pod/perlipc     \
103			--splititem=pod/perlfunc    \
104			--recurse \
105			--verbose
106
107=head1 AUTHOR
108
109Chris Hall E<lt>hallc@cs.colorado.eduE<gt>
110
111=cut
112
113my $usage;
114
115$usage =<<END_OF_USAGE;
116Usage: $0 --help --podpath=<name>:...:<name> --podroot=<name>
117         --htmldir=<name> --htmlroot=<name> --norecurse --recurse
118         --splithead=<name>,...,<name> --splititem=<name>,...,<name>
119         --ignore=<name>,...,<name> --verbose
120
121    --help      - this message
122    --podpath   - colon-separated list of directories containing .pod and
123                  .pm files to be converted (. by default).
124    --podroot   - filesystem base directory from which all relative paths in
125                  podpath stem (default is .).
126    --htmldir   - directory to store resulting html files in relative
127                  to the filesystem (\$podroot/html by default).
128    --htmlroot  - http-server base directory from which all relative paths
129                  in podpath stem (default is /).
130    --norecurse - don't recurse on those subdirectories listed in podpath.
131                  (default behavior).
132    --recurse   - recurse on those subdirectories listed in podpath
133    --splithead - comma-separated list of .pod or .pm files to split.  will
134                  split each file into several smaller files at every occurrence
135                  of a pod =head[1-6] directive.
136    --splititem - comma-separated list of .pod or .pm files to split using
137                  splitpod.
138    --splitpod  - directory where the program splitpod can be found
139                  (\$podroot/pod by default).
140    --ignore    - comma-separated list of files that shouldn't be installed.
141    --verbose   - self-explanatory.
142
143END_OF_USAGE
144
145my (@podpath, $podroot, $htmldir, $htmlroot, $recurse, @splithead,
146    @splititem, $splitpod, $verbose, $pod2html, @ignore);
147
148@podpath = ( "." );	# colon-separated list of directories containing .pod
149			# and .pm files to be converted.
150$podroot = ".";		# assume the pods we want are here
151$htmldir = "";		# nothing for now...
152$htmlroot = "/";	# default value
153$recurse = 0;		# default behavior
154@splithead = ();	# don't split any files by default
155@splititem = ();	# don't split any files by default
156$splitpod = "";		# nothing for now.
157
158$verbose = 0;		# whether or not to print debugging info
159
160$pod2html = "pod/pod2html";
161
162usage("") unless @ARGV;
163
164# Overcome shell's p1,..,p8 limitation.
165# See vms/descrip_mms.template -> descrip.mms for invocation.
166if ( $^O eq 'VMS' ) { @ARGV = split(/\s+/,$ARGV[0]); }
167
168our %Options;
169
170# parse the command-line
171my $result = GetOptions( \%Options, qw(
172	help
173	podpath=s
174	podroot=s
175	htmldir=s
176	htmlroot=s
177	ignore=s
178	recurse!
179	splithead=s
180	splititem=s
181	splitpod=s
182	verbose
183));
184usage("invalid parameters") unless $result;
185parse_command_line();
186
187
188# set these variables to appropriate values if the user didn't specify
189#  values for them.
190$htmldir = "$htmlroot/html" unless $htmldir;
191$splitpod = "$podroot/pod" unless $splitpod;
192
193
194# make sure that the destination directory exists
195(mkdir($htmldir, 0755) ||
196	die "$0: cannot make directory $htmldir: $!\n") if ! -d $htmldir;
197
198
199# the following array will eventually contain files that are to be
200# ignored in the conversion process.  these are files that have been
201# process by splititem or splithead and should not be converted as a
202# result.
203my @splitdirs;
204
205# split pods. It's important to do this before convert ANY pods because
206# it may affect some of the links
207@splitdirs = ();    # files in these directories won't get an index
208split_on_head($podroot, $htmldir, \@splitdirs, \@ignore, @splithead);
209split_on_item($podroot,           \@splitdirs, \@ignore, @splititem);
210
211
212# convert the pod pages found in @poddirs
213#warn "converting files\n" if $verbose;
214#warn "\@ignore\t= @ignore\n" if $verbose;
215foreach my $dir (@podpath) {
216    installdir($dir, $recurse, $podroot, \@splitdirs, \@ignore);
217}
218
219
220# now go through and create master indices for each pod we split
221foreach my $dir (@splititem) {
222    print "creating index $htmldir/$dir.html\n" if $verbose;
223    create_index("$htmldir/$dir.html", "$htmldir/$dir");
224}
225
226foreach my $dir (@splithead) {
227    (my $pod = $dir) =~ s,^.*/,,;
228    $dir .= ".pod" unless $dir =~ /(\.pod|\.pm)$/;
229    # let pod2html create the file
230    runpod2html($dir, 1);
231
232    # now go through and truncate after the index
233    $dir =~ /^(.*?)(\.pod|\.pm)?$/sm;
234    my $file = "$htmldir/$1";
235    print "creating index $file.html\n" if $verbose;
236
237    # read in everything until what would have been the first =head
238    # directive, patching the index as we go.
239    open(H, '<', "$file.html") ||
240	die "$0: error opening $file.html for input: $!\n";
241    $/ = "";
242    my @data = ();
243    while (<H>) {
244	last if m!<h1 id="NAME">NAME</h1>!;
245	$_ =~ s{href="#(.*)">}{
246	    my $url = "$file/@{[anchorify(qq($1))]}.html" ;
247	    $url = relativize_url( $url, "$file.html" )
248	    if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' );
249	    "href=\"$url\">" ;
250	}egi;
251	push @data, $_;
252    }
253    close(H);
254
255    # now rewrite the file
256    open(H, '>', "$file.html") ||
257	die "$0: error opening $file.html for output: $!\n";
258    print H @data, "</body>\n\n</html>\n\n\n";
259    close(H);
260}
261
262remove_tree(@splitdirs, {safe=>1});
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
284    $htmldir  = $Options{htmldir}	    if defined $Options{htmldir};
285    $htmlroot = $Options{htmlroot}	    if defined $Options{htmlroot};
286    $podroot  = $Options{podroot}	    if defined $Options{podroot};
287    $splitpod = $Options{splitpod}	    if defined $Options{splitpod};
288
289    $recurse  = $Options{recurse}	    if defined $Options{recurse};
290    $verbose  = $Options{verbose}	    if defined $Options{verbose};
291
292    @ignore = map "$podroot/$_", split(",", $Options{ignore}) if defined $Options{ignore};
293}
294
295
296sub create_index {
297    my($html, $dir) = @_;
298    (my $pod = $dir) =~ s,^.*/,,;
299
300    # get the list of .html files in this directory
301    opendir(DIR, $dir) ||
302	die "$0: error opening directory $dir for reading: $!\n";
303    my @files = sort(grep(/\.html?$/, readdir(DIR)));
304    closedir(DIR);
305
306    open(HTML, '>', $html) ||
307	die "$0: error opening $html for output: $!\n";
308
309    # for each .html file in the directory, extract the index
310    #	embedded in the file and throw it into the big index.
311    print HTML "<DL COMPACT>\n";
312    foreach my $file (@files) {
313
314	my $filedata = do {
315	    open(my $in, '<', "$dir/$file") ||
316		die "$0: error opening $dir/$file for input: $!\n";
317	    local $/ = undef;
318	    <$in>;
319	};
320
321	# pull out the NAME section
322	my($lcp1, $lcp2) =
323	    ($filedata =~
324		m#<h1 id="NAME">NAME</h1>\s*<p>\s*(\S+)\s+-\s+(\S.*?\S)</p>#);
325	defined $lcp1 or die "$0: can't find NAME section in $dir/$file\n";
326
327	my $url= "$dir/$file" ;
328	if ( ! defined $Options{htmlroot} || $Options{htmlroot} eq '' ) {
329	    $url = relativize_url( $url, $html ) ;
330	}
331
332	print HTML qq(<DT><A HREF="$url">);
333	print HTML "$lcp1</A></DT><DD>$lcp2</DD>\n";
334    }
335    print HTML "</DL>\n";
336
337    close(HTML);
338}
339
340
341sub split_on_head {
342    my($podroot, $htmldir, $splitdirs, $ignore, @splithead) = @_;
343    my($pod, $dirname, $filename);
344
345    # split the files specified in @splithead on =head[1-6] pod directives
346    print "splitting files by head.\n" if $verbose && $#splithead >= 0;
347    foreach $pod (@splithead) {
348	# figure out the directory name and filename
349	$pod      =~ s,^([^/]*)$,/$1,;
350	$pod      =~ m,(.*)/(.*?)(\.pod)?$,;
351	$dirname  = $1;
352	$filename = "$2.pod";
353
354	# since we are splitting this file it shouldn't be converted.
355	push(@$ignore, "$podroot/$dirname/$filename");
356
357	# split the pod
358	splitpod("$podroot/$dirname/$filename", "$podroot/$dirname", $htmldir,
359	    $splitdirs);
360    }
361}
362
363
364sub split_on_item {
365    my($podroot, $splitdirs, $ignore, @splititem) = @_;
366    my($pwd, $dirname, $filename);
367
368    print "splitting files by item.\n" if $verbose && $#splititem >= 0;
369    $pwd = getcwd();
370    my $splitter = rel2abs("$splitpod/splitpod", $pwd);
371    my $perl = rel2abs($^X, $pwd);
372    foreach my $pod (@splititem) {
373	# figure out the directory to split into
374	$pod      =~ s,^([^/]*)$,/$1,;
375	$pod      =~ m,(.*)/(.*?)(\.pod)?$,;
376	$dirname  = "$1/$2";
377	$filename = "$2.pod";
378
379	# since we are splitting this file it shouldn't be converted.
380	push(@$ignore, "$podroot/$dirname.pod");
381
382	# split the pod
383	push(@$splitdirs, "$podroot/$dirname");
384	-d "$podroot/$dirname" and remove_tree("$podroot/$dirname", {safe=>1});
385	mkdir("$podroot/$dirname", 0755) ||
386		    die "$0: error creating directory $podroot/$dirname: $!\n";
387	chdir("$podroot/$dirname") ||
388	    die "$0: error changing to directory $podroot/$dirname: $!\n";
389	die "$splitter not found. Use '-splitpod dir' option.\n"
390	    unless -f $splitter;
391	system($perl, $splitter, "../$filename") &&
392	    warn "$0: error running '$splitter ../$filename'"
393		 ." from $podroot/$dirname";
394    }
395    chdir($pwd);
396}
397
398
399#
400# splitpod - splits a .pod file into several smaller .pod files
401#  where a new file is started each time a =head[1-6] pod directive
402#  is encountered in the input file.
403#
404sub splitpod {
405    my($pod, $poddir, $htmldir, $splitdirs) = @_;
406    my(@poddata, @filedata, @heads);
407    my($file, $i, $j, $prevsec, $section, $nextsec);
408
409    print "splitting $pod\n" if $verbose;
410
411    # read the file in paragraphs
412    $/ = "";
413    open(SPLITIN, '<', $pod) ||
414	die "$0: error opening $pod for input: $!\n";
415    @filedata = <SPLITIN>;
416    close(SPLITIN) ||
417	die "$0: error closing $pod: $!\n";
418
419    # restore the file internally by =head[1-6] sections
420    @poddata = ();
421    for ($i = 0, $j = -1; $i <= $#filedata; $i++) {
422	$j++ if ($filedata[$i] =~ /^\s*=head[1-6]/);
423	if ($j >= 0) {
424	    $poddata[$j]  = "" unless defined $poddata[$j];
425	    $poddata[$j] .= "\n$filedata[$i]" if $j >= 0;
426	}
427    }
428
429    # create list of =head[1-6] sections so that we can rewrite
430    #  L<> links as necessary.
431    my %heads = ();
432    foreach $i (0..$#poddata) {
433	$heads{anchorify($1)} = 1 if $poddata[$i] =~ /=head[1-6]\s+(.*)/;
434    }
435
436    # create a directory of a similar name and store all the
437    #  files in there
438    $pod =~ s,.*/(.*),$1,;	# get the last part of the name
439    my $dir = $pod;
440    $dir =~ s/\.pod//g;
441    push(@$splitdirs, "$poddir/$dir");
442    -d "$poddir/$dir" and remove_tree("$poddir/$dir", {safe=>1});
443    mkdir("$poddir/$dir", 0755) ||
444	die "$0: could not create directory $poddir/$dir: $!\n";
445
446    $poddata[0] =~ /^\s*=head[1-6]\s+(.*)/;
447    $section    = "";
448    $nextsec    = $1;
449
450    # for each section of the file create a separate pod file
451    for ($i = 0; $i <= $#poddata; $i++) {
452	# determine the "prev" and "next" links
453	$prevsec = $section;
454	$section = $nextsec;
455	if ($i < $#poddata) {
456	    $poddata[$i+1] =~ /^\s*=head[1-6]\s+(.*)/;
457	    $nextsec       = $1;
458	} else {
459	    $nextsec = "";
460	}
461
462	# determine an appropriate filename (this must correspond with
463	#  what pod2html will try and guess)
464	# $poddata[$i] =~ /^\s*=head[1-6]\s+(.*)/;
465	$file = "$dir/" . anchorify($section) . ".pod";
466
467	# create the new .pod file
468	print "\tcreating $poddir/$file\n" if $verbose;
469	open(SPLITOUT, '>', "$poddir/$file") ||
470	    die "$0: error opening $poddir/$file for output: $!\n";
471	$poddata[$i] =~ s,L<([^<>]*)>,
472			defined $heads{anchorify($1)} ? "L<$dir/$1>" : "L<$1>"
473		     ,ge;
474	print SPLITOUT $poddata[$i]."\n\n";
475	print SPLITOUT "=over 4\n\n";
476	print SPLITOUT "=item *\n\nBack to L<$dir/\"$prevsec\">\n\n" if $prevsec;
477	print SPLITOUT "=item *\n\nForward to L<$dir/\"$nextsec\">\n\n" if $nextsec;
478	print SPLITOUT "=item *\n\nUp to L<$dir>\n\n";
479	print SPLITOUT "=back\n\n";
480	close(SPLITOUT) ||
481	    die "$0: error closing $poddir/$file: $!\n";
482    }
483}
484
485
486#
487# installdir - takes care of converting the .pod and .pm files in the
488#  current directory to .html files and then installing those.
489#
490sub installdir {
491    my($dir, $recurse, $podroot, $splitdirs, $ignore) = @_;
492
493    my @dirlist; # directories to recurse on
494    my @podlist; # .pod files to install
495    my @pmlist;  # .pm files to install
496
497    # should files in this directory get an index?
498    my $doindex = (grep($_ eq "$podroot/$dir", @$splitdirs) ? 0 : 1);
499
500    opendir(DIR, "$podroot/$dir")
501	|| die "$0: error opening directory $podroot/$dir: $!\n";
502
503    while(readdir DIR) {
504	no_upwards($_) or next;
505	my $is_dir = -d "$podroot/$dir/$_";
506	next if $is_dir and not $recurse;
507	my $target
508	    = $is_dir    ? \@dirlist
509	    : s/\.pod$// ? \@podlist
510	    : s/\.pm$//  ? \@pmlist
511	    : undef
512	    ;
513	push @$target, "$dir/$_" if $target;
514    }
515
516    closedir(DIR);
517
518    if ($^O eq 'VMS') { s/\.dir$//i for @dirlist }
519
520    # recurse on all subdirectories we kept track of
521    foreach $dir (@dirlist) {
522	installdir($dir, $recurse, $podroot, $splitdirs, $ignore);
523    }
524
525    # install all the pods we found
526    foreach my $pod (@podlist) {
527	# check if we should ignore it.
528	next if $pod =~ m(/t/); # comes from a test file
529	next if grep($_ eq "$pod.pod", @$ignore);
530
531	# check if a .pm files exists too
532	if (grep($_ eq $pod, @pmlist)) {
533	    print  "$0: Warning both '$podroot/$pod.pod' and "
534		. "'$podroot/$pod.pm' exist, using pod\n";
535	    push(@ignore, "$pod.pm");
536	}
537	runpod2html("$pod.pod", $doindex);
538    }
539
540    # install all the .pm files we found
541    foreach my $pm (@pmlist) {
542	# check if we should ignore it.
543	next if $pm =~ m(/t/); # comes from a test file
544	next if grep($_ eq "$pm.pm", @ignore);
545
546	runpod2html("$pm.pm", $doindex);
547    }
548}
549
550
551#
552# runpod2html - invokes pod2html to convert a .pod or .pm file to a .html
553#  file.
554#
555sub runpod2html {
556    my($pod, $doindex) = @_;
557    my($html, $i, $dir, @dirs);
558
559    $html = $pod;
560    $html =~ s/\.(pod|pm)$/.html/g;
561
562    # make sure the destination directories exist
563    @dirs = split("/", $html);
564    $dir  = "$htmldir/";
565    for ($i = 0; $i < $#dirs; $i++) {
566	if (! -d "$dir$dirs[$i]") {
567	    mkdir("$dir$dirs[$i]", 0755) ||
568		die "$0: error creating directory $dir$dirs[$i]: $!\n";
569	}
570	$dir .= "$dirs[$i]/";
571    }
572
573    # invoke pod2html
574    print "$podroot/$pod => $htmldir/$html\n" if $verbose;
575    Pod::Html::pod2html(
576        "--htmldir=$htmldir",
577	"--htmlroot=$htmlroot",
578	"--podpath=".join(":", @podpath),
579	"--podroot=$podroot",
580	"--header",
581	($doindex ? "--index" : "--noindex"),
582	"--" . ($recurse ? "" : "no") . "recurse",
583	"--infile=$podroot/$pod", "--outfile=$htmldir/$html");
584    die "$0: error running $pod2html: $!\n" if $?;
585}
586