#!/usr/bin/perl
# IBM(c) 2007 EPL license http://www.eclipse.org/legal/epl-v10.html

# Downloads/converts the xCAT docs on the sourceforge wiki to local HTML and PDF.
# This script is not dependent on other xCAT code, so you can copy it to a machine
# that has internet access to run it.  Before running this command, you must have
# wget, python, and pisa installed.  See: http://sourceforge.net/apps/mediawiki/xcat/index.php?title=Editing_xCAT_Documentation_Pages#Converting_Wiki_Pages_to_HTML_and_PDFs .

# Note: do not use the --upload option, unless your machine has authority to write to http://xcat.sourceforge.net/doc/ .
#       You also need to set $UPLOADUSER to your sourceforge user.

#BEGIN
#{
#    $::XCATROOT = $ENV{'XCATROOT'} ? $ENV{'XCATROOT'} : -d '/opt/xcat' ? '/opt/xcat' : '/usr';
#}
use strict;
#use lib "$::XCATROOT/lib/perl";
#use xCAT::Utils;
use Getopt::Long;
#use File::Path;
use Cwd;
#use Data::Dumper;

# Update this list if you group any xcat docs on a separate page such that they are no longer linked from the
# main doc page.
my @indexdocs = ('XCAT_Documentation', 'Power_775_Cluster_Documentation', 'Highly_Available_Management_Node', 'Mixed_Cluster_Support','IBM_HPC_Stack_in_an_xCAT_Cluster');

#my $VERSION;
my $HELP;
my $UPLOAD;
my $UPLOADONLY;
my $VERBOSE;
 
my $usage = sub {
   	my $exitcode = shift @_;
   	print "Usage: getxcatdocs [-?|-h|--help] [-v|--verbose] [-u|--upload] [--uploadonly] [<destination-dir>]\n";
	exit $exitcode;
};

# Process the cmd line args
Getopt::Long::Configure("bundling");
#Getopt::Long::Configure("pass_through");
Getopt::Long::Configure("no_pass_through");
if (!GetOptions('h|?|help'  => \$HELP, 'v|verbose' => \$VERBOSE, 'u|upload' => \$UPLOAD, 'uploadonly' => \$UPLOADONLY )) { $usage->(1); }

if ($HELP) { $usage->(0); }

#if ($VERSION) {
    #print xCAT::Utils->Version(), "\n";
#    exit;
#}

if ($^O =~ /^aix/i) { die "Error: this command is not yet supported on AIX.\n"; }

my $destdir = scalar(@ARGV) ? $ARGV[0] : '.';
chdir($destdir) or die "Can not cd to $destdir: $!\n";
#my $docdir = $ENV{'PWD'};

# Download the HTML docs and convert them all to pdfs
my @dir;
if (!$UPLOADONLY) {
	@dir = gethtmldocs('html');
	convert2pdf('pdf', \@dir);
}


# tar/compress
my $date=`date +%Y%m%d%H%M`;
chop $date;
my $docname="xcat-docs-snap$date.tar.gz";
#system('pwd');
my $cmd = "tar -zcf $docname html pdf 2>&1";
verbose($cmd);
system($cmd) == 0 or die "Error running $cmd: $!, rc=$?";

# Optionally upload the tarball to sourceforge
if ($UPLOAD || $UPLOADONLY) {
	my $UPLOADUSER = 'bp-sawyers';
	my $count = 1;
	#my $cmd = "rsync -v $docname $UPLOADUSER," . 'xcat@web.sourceforge.net:htdocs/doc/';
	my $cmd = "rsync -v $docname $UPLOADUSER," . 'xcat@web.sourceforge.net:/home/frs/project/x/xc/xcat/doc/';
	print "$cmd\n";
	while ($count<=5 && system("$cmd 2>&1")) { $count++; }
}
exit 0;


sub verbose { if ($VERBOSE) { print shift, "\n"; } }


# Download all of the html docs from several "index" docs
sub gethtmldocs {
	my $dir = shift;
	my $savedir = getcwd();
	#File::Path::make_path($dir);
	mkdir($dir);
	chdir($dir);
	#system('pwd');
	unlink <*>;		# delete all the files in the dir, in case they previously ran this
	#system('ls');
	
	my $indexes = '';
	foreach my $index (@indexdocs) {
		$indexes .= qq('http://sourceforge.net/apps/mediawiki/xcat/index.php?title=$index&printable=yes' );
	}
	print "Downloading the xCAT wiki documentation to $dir, from: $indexes ...\n";
	runwget($indexes);
	
	# Remove the funny chars from the links to other docs and rename the docs
	#my $sedcmd = q(sed -i 's/<a href="\/apps\/mediawiki\/xcat\/index.php?title/<a href="index.php%3Ftitle/' *);
	# sed -i 's/href="index.php%3Ftitle=/href="/g' index.php\?title\=
	# sed -i 's/<a href="\([^"]*\)"/<a href="\1.html"/'
	# This searches for '<a href="index.php?title=' and then all text before a '"' or '#', and then removes the front part and add .html on the end
	# Note: this does not convert the 'MediaWiki:*' files because they are used in <link> tags, but converting them does not seem to do any good anyway.
	my $cmd = q(sed -i 's/<a href="index.php?title=\\([^"#]*\\)\\("\|#\\)/<a href="\1.html\2/g' *);
	verbose($cmd);
	system($cmd) == 0 or die "Error running $cmd: $!, rc=$?";
	# get the list of docs
	opendir(DIR, '.') or die "Error: could not read the just created html directory.\n";
	#my @docs = grep /^index.php\?title=/, readdir(DIR);		# /
	my @docs;
	foreach my $f (readdir(DIR)) {
		if ($f !~ /^index.php\?title=/ || $f =~ /^index.php\?title=MediaWiki:/) { next; }
		my $newf = $f;
		$newf =~ s/^index.php\?title=//;
		if ($newf !~ /\./) { $newf .= '.html'; }
		verbose("Renaming $f to $newf");
		rename($f, $newf);
		push @docs, $newf;
	}
	close(DIR);
	chdir($savedir);
	return @docs;
}


# Convert to pdf
sub convert2pdf {
	my ($dir, $files) = @_;
	my $savedir = getcwd();
	#File::Path::make_path($dir);
	mkdir($dir);
	chdir($dir);
	if (system('which xhtml2pdf >/dev/null 2>&1')) { die "xhtml2pdf is not installed.  See http://sourceforge.net/apps/mediawiki/xcat/index.php?title=Editing_xCAT_Documentation_Pages#Converting_Wiki_Pages_to_HTML_and_PDFs .\n"; }
	unlink <*>;		# delete all the files in the dir, in case they previously ran this
	foreach my $file (@$files) {
		#if ($file =~ /^index.php\?title=MediaWiki:/ || $file eq 'index.php?title=XCAT_Documentation') { next; }
		if ($file eq 'XCAT_Documentation') { next; }
		#my ($docname) = $file =~ /^index.php\?title=(.+)$/;
		$file =~ s/\.html$//;
		print "Converting $file to PDF format...\n";
		my $url = 'http://sourceforge.net/apps/mediawiki/xcat/index.php?title=' . $file . '&printable=yes';
		my $destfile = "$file.pdf";
		my $cmd = "xhtml2pdf '$url' '$destfile' ";
		runh2p($cmd);
	}
	chdir($savedir);
}


# Run the wget cmd and filter out some of the silly output
sub runwget {
	my $index = shift;
	# options we might consider: --html-extension --restrict-file-names=windows  --cut-dirs=3
	# options that do not work:  --relative
	#my $rejectlist = q('*title=Special:*,*title=Talk:*,*title=-&*,*title=HowTos,*title=Main_Page,*title=MediaWiki:*,*title=Release_Notes,*title=Wish_List_for_xCAT_2,*&action=edit*,*&action=history*,*&printable=yes*,*&oldid=*,index.html,opensearch_desc.php,xcat,login.php,support');
	my $rejectlist = q('*title=Special:*,*title=Talk:*,*title=-&*,*title=HowTos,*title=Main_Page,*title=Release_Notes,*title=Wish_List_for_xCAT_2,*&action=edit*,*&action=history*,*&printable=yes*,*&oldid=*,index.html,opensearch_desc.php,xcat,login.php,support');
	my $cmd = qq(wget --recursive --convert-links --no-verbose --progress=bar --level=1 --page-requisites --no-parent --no-host-directories --no-directories --no-clobber --execute robots=off --post-data='printable=yes' --reject $rejectlist $index);
	verbose($cmd);
	open(OUT, "$cmd 2>&1 |") || die "can't fork $cmd: $!\n";
	while (<OUT>) {
		if (/URL:https*:\/\/sourceforge\.net.+\s+->\s+\"(\S+)\"\s+\[/) { print "Downloaded $1.\n"; }
		else { print; }
	}
	close OUT || die "Error running $cmd: $! $?";
}

# Run the xhtml2pdf cmd and filter out some of the silly output
sub runh2p {
	my $cmd = shift;
	verbose($cmd);
	open(OUT, "$cmd 2>&1 |") || die "can't fork $cmd: $!\n";
	while (<OUT>) {
		next if /DeprecationWarning:\sthe sets module is deprecated/;
		next if /from sets import ImmutableSet/;
		next if /^Converting\ssourceforge.net/;
		print;
	}
	close OUT || die "Error running $cmd: $! $?";
}