#!/usr/bin/env perl # use strict; use warnings; use CSS::Minifier::XS 0.07; use File::Basename; use File::Copy; use File::Find; use File::Path qw(make_path); use File::Slurp; use File::Spec; use Getopt::Long; use JavaScript::Minifier::XS; use Pod::Usage; use XML::LibXML; use XML::LibXSLT; use FindBin qw'$Bin'; my $canonical_url = 'https://www.exim.org/'; ## Parse arguments my %opt = parse_arguments(); ## setup static root location ## TODO: for doc generation only this should be within the docs dir $opt{staticroot} = File::Spec->catdir( $opt{docroot}, 'static' ); ## Generate the pages my %cache; # General cache object do_doc( 'spec', $_ ) foreach @{ $opt{spec} || [] }; do_doc( 'filter', $_ ) foreach @{ $opt{filter} || [] }; do_web() if ( $opt{web} ); do_static() if ( $opt{web} or !$opt{localstatic} ); # need this for all other pages generated ## Add the exim-html-current symlink foreach my $type (qw(html pdf)) { print "Symlinking exim-$type-current to exim-$type-$opt{latest}\n" if ( $opt{verbose} ); unlink("$opt{docroot}/exim-$type-current"); symlink( "exim-$type-$opt{latest}", "$opt{docroot}/exim-$type-current" ) || warn "symlink to $opt{docroot}/exim-$type-current failed"; } # ------------------------------------------------------------------ ## Generate the website files sub do_web { ## copy these templates to docroot... copy_transform_files( "$opt{tmpl}/web", $opt{docroot}, 0 ); } # ------------------------------------------------------------------ ## Generate the static file set sub do_static { my $staticroot = shift || $opt{staticroot}; ## make sure I have a directory mkdir($staticroot) or die "Unable to make staticroot: $!\n" unless ( -d $staticroot ); ## copy these templates to docroot... copy_transform_files( "$opt{tmpl}/static", $staticroot, 1 ); } # ------------------------------------------------------------------ ## Generate the website files sub copy_transform_files { my $source = shift; my $target = shift; my $static = shift; ## Make sure the template web directory exists die "No such directory: $source\n" unless ( -d $source ); ## Scan the web templates find( sub { my ($path) = substr( $File::Find::name, length("$source"), length($File::Find::name) ) =~ m#^/*(.*)$#; if ( -d "$source/$path" ) { ## Create the directory in the target if it doesn't exist if ( !-d "$target/$path" ) { mkdir("$target/$path") or die "Unable to make $target/$path: $!\n"; } } else { ## Build HTML from XSL files and simply copy static files which have changed if ( ( !$static ) and ( $path =~ /(.+)\.xsl$/ ) ) { print "Generating : /$1.html\n" if ( $opt{verbose} ); transform( undef, "$source/$path", "$target/$1.html" ); } elsif ( -f "$source/$path" ) { ## Skip if the file hasn't changed (mtime/size based) return if (( -f "$target/$path" ) and ( ( stat("$source/$path") )[9] == ( stat("$target/$path") )[9] ) and ( ( stat("$source/$path") )[7] == ( stat("$target/$path") )[7] ) ); if ( $path =~ /(.+)\.css$/ ) { print "CSS to : /$path\n" if ( $opt{verbose} ); my $content = read_file("$source/$path"); write_file( "$target/$path", $opt{minify} ? CSS::Minifier::XS::minify($content) : $content ); } elsif ( $path =~ /(.+)\.js$/ ) { print "JS to : /$path\n" if ( $opt{verbose} ); my $content = read_file("$source/$path"); write_file( "$target/$path", $opt{minify} ? JavaScript::Minifier::XS::minify($content) : $content ); } else { ## Copy print "Copying to : /$path\n" if ( $opt{verbose} ); copy( "$source/$path", "$target/$path" ) or die "$path: $!"; } ## Set mtime utime( time, ( stat("$source/$path") )[9], "$target/$path" ); } } }, "$source" ); } # ------------------------------------------------------------------ ## Generate index/chapter files for a doc sub do_doc { my ( $type, $xml_path ) = @_; ## Read and validate the XML file my $xml = XML::LibXML->new(expand_entities => 1)->parse_file($xml_path) or die $!; ## Get the version number my $version = do { my $version = $xml->findvalue('/book/bookinfo/revhistory/revision/revnumber'); die "Unable to get version number\n" unless defined $version and $version =~ /^ (? \d+ # major (?:\.\d+(?:\.\d+)?)? # (minor(.patch)) (?:\.\d+(?:\.\d+(?:\.\d+)?)?)? # (minor(.patch.(fixes))) ) (?:-RC\d+)?$/x; # -RCX $+{version}; }; ## Prepend chapter filenames? my $prepend_chapter = $type eq 'filter' ? 'filter_' : ''; ## Add the canonical url for this document $xml->documentElement() ->appendTextChild( 'canonical_url', "${canonical_url}exim-html-current/doc/html/spec_html/" . ( $type eq 'spec' ? 'index' : 'filter' ) . ".html" ); ## Add a url for the latest version of this document if ( $version ne $opt{latest} ) { $xml->documentElement() ->appendTextChild( 'current_url', "../../../../exim-html-current/doc/html/spec_html/" . ( $type eq 'spec' ? 'index' : 'filter' ) . ".html" ); } ## Fixup the XML xref_fixup( $xml, $prepend_chapter ); ## set the staticroot my $staticroot = $opt{localstatic} ? File::Spec->catdir( $opt{docroot}, "exim-html-$version", 'doc', 'html', 'static' ) : $opt{staticroot}; unless ( -d $staticroot ) { make_path( $staticroot, { verbose => $opt{verbose} } ); do_static($staticroot); } ## Generate the front page { my $path = "exim-html-$version/doc/html/spec_html/" . ( $type eq 'filter' ? $type : 'index' ) . ".html"; print "Generating : docroot:/$path\n" if ( $opt{verbose} ); transform( $xml, "$opt{tmpl}/doc/index.xsl", "$opt{docroot}/$path", $staticroot ); } ## Generate a Table of Contents XML file { my $path = "exim-html-$version/doc/html/spec_html/" . ( $type eq 'filter' ? 'filter_toc' : 'index_toc' ) . ".xml"; print "Generating : docroot:/$path\n" if ( $opt{verbose} ); transform( $xml, "$opt{tmpl}/doc/toc.xsl", "$opt{docroot}/$path", $staticroot ); } ## Generate the chapters my $counter = 0; my @chapters = map { $_->cloneNode(1) } $xml->findnodes('/book/chapter'); my( $chapter_title, $chapter_title_prev, $chapter_title_next ); foreach my $chapter (@chapters) { ## Add a N node for the stylesheet to use $chapter->appendTextChild( 'chapter_id', ++$counter ); ## Get the current and surrounding chapter titles $chapter_title_prev = $chapter_title; $chapter_title = $chapter_title_next || $chapter->findvalue('title_uri'); $chapter_title_next = $chapters[$counter]->findvalue('title_uri') if $counter < int(@chapters); ## Add previous/next/canonical urls for nav { $chapter->appendTextChild( 'prev_url', $counter == 1 ? $type eq 'filter' ? 'filter.html' : 'index.html' : sprintf( '%sch-%s.html', $prepend_chapter, $chapter_title_prev ) ); $chapter->appendTextChild( 'this_url', sprintf( '%sch-%s.html', $prepend_chapter, $chapter_title ) ); $chapter->appendTextChild( 'next_url', sprintf( '%sch-%s.html', $prepend_chapter, $chapter_title_next ) ) unless int(@chapters) == $counter; $chapter->appendTextChild( 'toc_url', ( $type eq 'filter' ? 'filter' : 'index' ) . '.html' ); $chapter->appendTextChild( 'canonical_url', sprintf( 'https://www.exim.org/exim-html-current/doc/html/spec_html/%sch-%s.html', $prepend_chapter, $chapter_title ) ); if ( $version ne $opt{latest} ) { $chapter->appendTextChild( 'current_url', sprintf( '../../../../exim-html-current/doc/html/spec_html/%sch-%s.html', $prepend_chapter, $chapter_title ) ); } } ## Create an XML document from the chapter my $doc = XML::LibXML::Document->createDocument( '1.0', 'UTF-8' ); $doc->setDocumentElement($chapter); ## Transform the chapter into html { my $real_path = sprintf( 'exim-html-%s/doc/html/spec_html/%sch-%s.html', $version, $prepend_chapter, $chapter_title ); my $link_path = sprintf( 'exim-html-%s/doc/html/spec_html/%sch%02d.html', $version, $prepend_chapter, $counter ); print "Generating : docroot:/$real_path\n" if ( $opt{verbose} ); transform( $doc, "$opt{tmpl}/doc/chapter.xsl", "$opt{docroot}/$real_path", $staticroot ); # Making a relative symlink to a file in the same directory. # Extract just the filename portion of $real_path. my $real_file = basename($real_path); print "Symlinking : docroot:/$link_path to $real_file\n" if ( $opt{verbose} ); if ( -f "$opt{docroot}/$link_path" ) { unlink("$opt{docroot}/$link_path") or die "failed removing $opt{docroot}/$link_path: $!"; } symlink( "$real_file", "$opt{docroot}/$link_path" ) || die "symlink to $opt{docroot}/$link_path failed: $!"; } } } # ------------------------------------------------------------------ ## Fixup xref tags sub xref_fixup { my ( $xml, $prepend_chapter ) = @_; my %index = (); ## Add the "prepend_chapter" info ( $xml->findnodes('/book') )[0]->appendTextChild( 'prepend_chapter', $prepend_chapter ); ## Iterate over each chapter my $chapter_counter = 0; foreach my $chapter ( $xml->findnodes('/book/chapter') ) { ++$chapter_counter; my $chapter_id = $chapter->getAttribute('id'); unless ($chapter_id) { # synthesise missing id $chapter_id = sprintf( 'chapter_noid_%04d', $chapter_counter ); $chapter->setAttribute( 'id', $chapter_id ); } my $chapter_title = $chapter->findvalue('title'); ## Set title_uri so we can use eg ch-introduction.html instead of ch01.html $chapter->appendTextChild( 'title_uri', title_to_uri($chapter_title) ); $index{$chapter_id} = { chapter_id => $chapter_counter, chapter_title => $chapter_title }; ## Iterate over each section my $section_counter = 0; foreach my $section ( $chapter->findnodes('section') ) { ++$section_counter; $section->setAttribute( 'sectprefix', $section_counter ); my $section_id = $section->getAttribute('id'); unless ($section_id) { # synthesise missing id $section_id = sprintf( 'section_noid_%04d_%04d', $chapter_counter, $section_counter ); $section->setAttribute( 'id', $section_id ); } my $section_title = $section->findvalue('title'); $index{$section_id} = { chapter_id => $chapter_counter, chapter_title => $chapter_title, section_id => $section_counter, section_title => $section_title }; # 2022/07/07 jgh: added loop for sections under sections, which are resulting from the .subsection macro # Add a "level" attribute to these nodes ## Iterate over each subsection my $subsec_counter = 0; foreach my $subsection ( $section->findnodes('section') ) { ++$subsec_counter; $subsection->setAttribute( 'level', "2" ); $subsection->setAttribute( 'sectprefix', sprintf("%d.%d", $section_counter, $subsec_counter) ); my $subsec_id = $subsection->getAttribute('id'); unless ($subsec_id) { # synthesise missing id $subsec_id = sprintf( 'section_noid_%04d_%04d_%04d', $chapter_counter, $section_counter, $subsec_counter ); $subsection->setAttribute( 'id', $subsec_id ); } my $subsec_title = $subsection->findvalue('title'); $index{$subsec_id} = { chapter_id => $chapter_counter, chapter_title => $chapter_title, section_id => $subsec_counter, section_title => $subsec_title }; } } } ## Build indexes as new chapters build_indexes( $xml, $prepend_chapter, \%index ); ## Replace all of the xrefs in the XML foreach my $xref ( $xml->findnodes('//xref') ) { my $linkend = $xref->getAttribute('linkend'); if ( exists $index{$linkend} ) { $xref->setAttribute( 'chapter_id', $index{$linkend}{'chapter_id'} ) if ( $index{$linkend}{'chapter_id'} ); $xref->setAttribute( 'chapter_title', $index{$linkend}{'chapter_title'} ); $xref->setAttribute( 'section_id', $index{$linkend}{'section_id'} ) if ( $index{$linkend}{'section_id'} ); $xref->setAttribute( 'section_title', $index{$linkend}{'section_title'} ) if ( $index{$linkend}{'section_title'} ); $xref->setAttribute( 'url', sprintf( '%sch-%s.html', $prepend_chapter, title_to_uri($index{$linkend}{'chapter_title'}) ) . ( $index{$linkend}{'section_id'} ? '#' . $linkend : '' ) ); } } } # ------------------------------------------------------------------ ## Build indexes sub build_indexes { my ( $xml, $prepend_chapter, $xref ) = @_; my $index_hash = {}; my $seealso_hash = {}; my $current_id; my $verterm_counter = 0; foreach my $node ( $xml->findnodes('//section | //chapter | //varlistentry | //indexterm') ) { if ( $node->nodeName eq 'indexterm' ) { my $role = $node->getAttribute('role') || 'concept'; my $primary = $node->findvalue('child::primary'); my $first = ( $primary =~ /^[A-Za-z]/ ) ? uc( substr( $primary, 0, 1 ) ) : ''; # first letter or marker my $secondary = $node->findvalue('child::secondary') || ''; my $see = $node->findvalue('child::see'); my $see_also = $node->findvalue('child::seealso'); next unless ( $primary || $secondary ); # skip blank entries for now... $index_hash->{$role}{$first}{$primary}{$secondary} ||= []; if ( $see || $see_also ) { # The scalar value being written here assumes only one seealso on an indeed term # It would be nice to have the $see displayed in bold rather than in quotes $seealso_hash->{$role}{$first}{$primary}{$secondary} = 'see "' . $see .'"' if ($see); $seealso_hash->{$role}{$first}{$primary}{$secondary} = 'see also "' . $see_also .'"' if ($see_also); } else { push @{ $index_hash->{$role}{$first}{$primary}{$secondary} }, $current_id; } } elsif ( $node->nodeName eq 'varlistentry' ) { foreach my $vitem ( $node->findnodes('listitem') ) { # Add an anchorname xml attribute. # chapter.xsl spots this and places a " " my $anchorname = sprintf("vi%d", $verterm_counter++); $vitem->setAttribute( 'anchorname', $anchorname ); $current_id = $anchorname; # Set the latest indexable id to be picked up by the next indexterm, # which should be in the content of the listitem my ($chapter_title, $sec_id, $sec_title); foreach my $chap ( $node->findnodes('ancestor::chapter') ) { $chapter_title = $chap->findvalue('title'); } next unless ($chapter_title); # Search upward to find a subsection or section id & title foreach my $ssec ( $node->findnodes("ancestor::section[\@level='2']") ) { $sec_id = $ssec->getAttribute('id'); $sec_title = $ssec->findvalue('title'); last; } if (!defined($sec_id)) { foreach my $sec ( $node->findnodes('ancestor::section') ) { $sec_id = $sec->getAttribute('id'); $sec_title = $sec->findvalue('title'); last; } } $xref->{$anchorname}{'chapter_title'} = $chapter_title; $xref->{$anchorname}{'section_id'} = $anchorname; $xref->{$anchorname}{'section_title'} = $sec_title if ($sec_title); } } else { $current_id = $node->getAttribute('id'); } } # now we build a set of new chapters with the index data in my $book = ( $xml->findnodes('/book') )[0]; foreach my $role ( sort { $a cmp $b } keys %{$index_hash} ) { my $chapter = XML::LibXML::Element->new('chapter'); $book->appendChild($chapter); $chapter->setAttribute( 'id', join( '_', 'index', $role ) ); $chapter->setAttribute( 'class', 'index' ); $chapter->appendTextChild( 'title', ( ucfirst($role) . ' Index' ) ); $chapter->appendTextChild( 'title_uri', title_to_uri(ucfirst($role) . ' Index') ); foreach my $first ( sort { $a cmp $b } keys %{ $index_hash->{$role} } ) { my $section = XML::LibXML::Element->new('section'); my $list = XML::LibXML::Element->new('variablelist'); $chapter->appendChild($section); $section->setAttribute( 'id', join( '_', 'index', $role, $first ) ); $section->setAttribute( 'class', 'index' ); $section->appendTextChild( 'title', $first ? $first : 'Symbols' ); $section->appendChild($list); foreach my $primary ( sort { $a cmp $b } keys %{ $index_hash->{$role}{$first} } ) { my $entry = XML::LibXML::Element->new('varlistentry'); my $item = XML::LibXML::Element->new('listitem'); $list->appendChild($entry)->appendTextChild( 'term', $primary ); $entry->appendChild($item); my $slist; foreach my $secondary ( sort { $a cmp $b } keys %{ $index_hash->{$role}{$first}{$primary} } ) { my $para = XML::LibXML::Element->new('para'); if ( $secondary eq '' ) { $item->appendChild($para); # skip having extra layer of heirarchy } else { unless ($slist) { $slist = XML::LibXML::Element->new('variablelist'); $item->appendChild($slist); } my $sentry = XML::LibXML::Element->new('varlistentry'); my $sitem = XML::LibXML::Element->new('listitem'); $slist->appendChild($sentry)->appendTextChild( 'term', $secondary ); $sentry->appendChild($sitem)->appendChild($para); } my $seealso = $seealso_hash->{$role}{$first}{$primary}{$secondary}; $para->appendText($seealso) if ($seealso); my $count = 0; foreach my $ref ( @{ $index_hash->{$role}{$first}{$primary}{$secondary} } ) { $para->appendText(', ') if ( $count++ ); my $xrefel = XML::LibXML::Element->new('xref'); $xrefel->setAttribute( linkend => $ref ); $xrefel->setAttribute( longref => 1 ); $para->appendChild($xrefel); } } } } } } # ------------------------------------------------------------------ ## Handle the transformation sub transform { my ( $xml, $xsl_path, $out_path, $staticroot_abs ) = @_; ## make sure $staticroot is set $staticroot_abs ||= $opt{staticroot}; ## Build an empty XML structure if an undefined $xml was passed unless ( defined $xml ) { $xml = XML::LibXML::Document->createDocument( '1.0', 'UTF-8' ); $xml->setDocumentElement( $xml->createElement('content') ); } ## Add the current version of Exim to the XML $xml->documentElement()->appendTextChild( 'current_version', $opt{latest} ); ## Add the old versions of Exim to the XML $xml->documentElement()->appendTextChild( 'old_versions', $_ ) foreach old_docs_versions(); ## Parse the ".xsl" file as XML my $xsl = XML::LibXML->new(expand_entities => 1)->parse_file($xsl_path) or die $!; ## Generate a stylesheet from the ".xsl" XML. my $stylesheet = XML::LibXSLT->new()->parse_stylesheet($xsl); ## work out the static root relative to the target my $target_dir = ( File::Spec->splitpath($out_path) )[1]; my $staticroot = File::Spec->abs2rel( $staticroot_abs, $target_dir ); ## Generate a doc from the XML transformed with the XSL my $doc = $stylesheet->transform( $xml, staticroot => sprintf( "'%s'", $staticroot ) ); ## Make the containing directory if it doesn't exist make_path( ( $out_path =~ /^(.+)\/.+$/ )[0], { verbose => $opt{verbose} } ); ## Write out the document open my $out, '>', $out_path or die "Unable to write $out_path - $!"; print $out $stylesheet->output_as_bytes($doc); close $out; } # ------------------------------------------------------------------ ## Takes a chapter title and fixes it up so it is suitable for use in a URI sub title_to_uri { my $title = lc(shift); $title =~ s/[^a-z0-9\s]+//gi; # Only allow spaces, numbers and letters $title =~ s/\s+/_/g; # Replace spaces with underscores so URLs are easier to copy about return $title; } # ------------------------------------------------------------------ ## Look in the docroot for old versions of the documentation sub old_docs_versions { if ( !exists $cache{old_docs_versions} ) { my @versions; foreach ( glob("$opt{docroot}/exim-html-*") ) { push @versions, $1 if /-(\d+(?:\.\d+)?)$/ && $1 lt $opt{latest} && -d $_; } $cache{old_docs_versions} = [ reverse sort { $a cmp $b } @versions ]; } return @{ $cache{old_docs_versions} }; } # ------------------------------------------------------------------ ## error_help sub error_help { my $msg = shift; warn $msg; pod2usage( -exitval => 1, -verbose => 0 ); } # ------------------------------------------------------------------ ## Parse arguments sub parse_arguments { my %opt = ( spec => [], filter => [], help => 0, man => 0, web => 0, minify => 1, verbose => 0, localstatic => 0, tmpl => "$Bin/../templates" ); GetOptions( \%opt, 'help|h!', 'man!', 'web!', 'spec=s{1,}', 'filter=s{1,}', 'latest=s', 'tmpl=s', 'docroot=s', 'minify!', 'verbose!', 'localstatic!' ) || pod2usage( -exitval => 1, -verbose => 0 ); ## --help pod2usage(0) if ( $opt{help} ); pod2usage( -verbose => 2 ) if ( $opt{man} ); ## --spec and --filter lists foreach my $set (qw[spec filter]) { $opt{$set} = [ map { my $f = File::Spec->rel2abs($_); error_help( 'No such file: ' . $_ ) unless -f $f; $f } @{ $opt{$set} } ]; } ## --latest error_help('Missing value for latest') unless ( exists( $opt{latest} ) && defined( $opt{latest} ) ); error_help('Invalid value for latest') unless $opt{latest} =~ /^\d+(?:\.\d+)*$/; ## --tmpl and --docroot foreach my $set (qw[tmpl docroot]) { error_help( 'Missing value for ' . $set ) unless ( exists( $opt{$set} ) && defined( $opt{$set} ) ); my $f = File::Spec->rel2abs( $opt{$set} ); error_help( 'No such directory: ' . $opt{$set} ) unless -d $f; $opt{$set} = $f; } error_help('Excess arguments') if ( scalar(@ARGV) ); error_help('Must include at least one of --web, --spec or --filter') unless ( $opt{web} || scalar( @{ $opt{spec} || [] } ) || scalar( @{ $opt{filter} || [] } ) ); return %opt; } # ------------------------------------------------------------------ 1; __END__ =head1 NAME gen - Generate exim html documentation and website =head1 SYNOPSIS gen [options] Options: --help display this help and exits --man displays man page --spec file... spec docbook/XML source files --filter file... filter docbook/XML source files --web Generate the general website pages --latest VERSION Required. Specify the latest stable version of Exim. --tmpl PATH Required. Path to the templates directory --docroot PATH Required. Path to the website document root --[no-]minify [Don't] minify CSS and Javascript --localstatic Makes the static files local to each doc ver =head1 OPTIONS =over 4 =item B<--help> Display help and exits =item B<--man> Display man page =item B<--spec> I List of files that make up the specification documentation docbook/XML source files. =item B<--filter> I List of files that make up the filter documentation docbook/XML source files. =item B<--web> Generate the website from the template files. =item B<--latest> I Specify the current exim version. This is used to create links to the current documentation. This option is I =item B<--tmpl> I Specify the directory that the templates are kept in. This option is I =item B<--docroot> I Specify the directory that the output should be generated into. This is the website C directory. This option is I =item B<--minify> If this option is set then both the CSS and Javascript files processed are minified using L and L respectively. This option is set by default - to disable it specify C<--no-minify> =item B<--localstatic> Makes the static files (CSS, images etc), local for each version of the documentation. This is more suitable for packaged HTML documentation. =back =head1 DESCRIPTION Generates the exim website and HTML documentation. =head1 EXAMPLE script/gen \ --web \ --spec docbook/*/spec.xml \ --filter docbook/*/filter.xml \ --latest 4.72 \ --tmpl templates \ --docroot /tmp/website =head1 AUTHOR Mike Cardwell Nigel Metheringham - mostly broke the framework Mike produced. =head1 COPYRIGHT Copyright 2010-2012 Exim Maintainers. All rights reserved. =cut