use URI::Escape;
use Plack::Request;
use Data::Dump qw(dump);
-use File::Path;
+use File::Path qw(make_path remove_tree);
use Graphics::Magick;
use File::Slurp;
-use Image::Size;
use JSON;
-use Time::Piece ();
-use Time::Seconds 'ONE_YEAR';
+use autodie;
+use Time::HiRes qw(time);
+use Encode;
sub make_basedir {
my $path = shift;
return if -e $path;
$path =~ s{/[^/]+$}{} || die "no dir/file in $path";
- File::Path::make_path $path;
+ warn "# make_basedir $path\n";
+ -e $path ? 0 : File::Path::make_path $path;
}
# Stolen from rack/directory.rb
// Book title and the URL used for the book title link
br.bookTitle= '%s';
-br.bookUrl = 'http://openlibrary.org';
+br.bookUrl = '%s';
// Override the path used to find UI images
br.imagesBaseURL = '/BookReader/images/';
];
}
+sub convert { gm('convert',@_) }
+sub montage { gm('montage',@_) }
+
+sub gm {
+ my $command = shift;
+ warn "# $command ",dump(@_);
+ my $t = time();
+ system 'gm', $command, @_;
+ $t = time() - $t;
+ warn sprintf("## $command %d bytes in %.2f s %s\n", -s $_[-1], $t, $_[-1]);
+}
+
+sub longest_common_prefix {
+ my $prefix = shift;
+ for (@_) {
+ chop $prefix while (! /^\Q$prefix\E/i);
+ }
+ warn "# longest_common_prefix [$prefix]\n";
+ return $prefix;
+}
+
+sub sort_pages {
+ my $prefix = longest_common_prefix @_;
+ sort {
+ my ( $an,$bn ) = ( $a,$b );
+ $an =~ s/^\Q$prefix\E//i; $an =~ s/\D+//g;
+ $bn =~ s/^\Q$prefix\E//i; $bn =~ s/\D+//g;
+ warn "## sort [$a] $an <=> $bn [$b]\n";
+ $an <=> $bn;
+ } @_;
+}
+
+sub convert_pdf_page {
+ my ($pdf, $page, $path) = @_;
+ my $t = time();
+
+ make_path $path;
+
+ warn "# pdfimages $page $pdf -> $path/\n";
+ system 'pdfimages', '-f', $page, '-l', $page, '-q', '-j', '-p', $pdf, "$path/p";
+
+ my @parts = ();
+ # glob split on spaces!
+ opendir(my $dh, $path);
+ while (readdir($dh)) {
+ my $full = "$path/$_";
+ warn "## readdir $full\n";
+ next unless -f $full; # skip . ..
+ push @parts, $_;
+ }
+ closedir $dh;
+
+ die "can't find images for $pdf in $path" unless $#parts >= 0;
+
+ @parts = sort_pages @parts;
+
+ my $image = "$path.jpg";
+
+ if ( $#parts == 0 ) { # single image
+ my $part = "$path/$parts[0]";
+ convert( $part => $image );
+ } else {
+ my @full = map { "$path/$_" } @parts;
+ montage( @full, '-tile', '1x'.scalar(@full), '-geometry', '+1+1' => $image );
+ }
+
+ die "$image: $!" unless -r $image;
+
+ remove_tree $path;
+
+ $t = time() - $t;
+ warn sprintf("## page: %d in %.2f s for %s\n", $page, $t, $image);
+ return $image;
+}
+
sub serve_path {
my($self, $env, $path, $fullpath) = @_;
my $req = Plack::Request->new($env);
- if (-f $path) {
+ my $dir_url = $env->{SCRIPT_NAME} . $env->{PATH_INFO};
+ my @files = ();
+ my @page_files;
+
+ if ( -f $path && $path =~ s{/([^/]+\.pdf)$}{} ) {
+ push @page_files, $1;
+ warn "# single pdf: $path / $1\n";
+ } elsif (-f $path ) {
if ( my $reduce = $req->param('reduce') ) {
$reduce = int($reduce); # BookReader javascript somethimes returns float
warn "# reduce $reduce $path\n";
- my $cache_path = "cache/$path.reduce.$reduce.jpg";
+ my $cache_path = "cache/$dir_url.reduce.$reduce.jpg";
if ( $reduce <= 1 && $path =~ m/\.jpe?g$/ ) {
$cache_path = $path;
} elsif ( ! -e $cache_path ) {
- my $image = Graphics::Magick->new( magick => 'jpg' );
- $image->Read($path);
- my ( $w, $h ) = $image->Get('width','height');
- $image->Resize(
- width => $w / $reduce,
- height => $h / $reduce
- );
make_basedir $cache_path;
- $image->Write( filename => $cache_path );
- warn "# created $cache_path ", -s $cache_path, " bytes\n";
+ convert( '-scale', ( 100 / $reduce ) .'%', $path => $cache_path );
}
return $self->SUPER::serve_path($env, $cache_path, $fullpath);
}
return $self->SUPER::serve_path($env, $path, $fullpath);
- }
+ } elsif ( -d $path ) {
- my $dir_url = $env->{SCRIPT_NAME} . $env->{PATH_INFO};
+ if ($dir_url !~ m{/$}) {
+ return $self->return_dir_redirect($env);
+ }
- if ($dir_url !~ m{/$}) {
- return $self->return_dir_redirect($env);
- }
+ my $dh = DirHandle->new($path);
+ my @children;
+ while (defined(my $ent = $dh->read)) {
+ next if $ent eq '.';
+ push @children, $ent;
+ }
- my @files = ([ "../", "Parent Directory", '', '', '' ]);
+ for my $basename (sort { $a cmp $b } @children) {
+ push @page_files, $basename if $basename =~ m/\d+\D?\.(jpg|gif|pdf)$/;
+ my $file = "$path/$basename";
+ my $url = $dir_url . $basename;
- my $dh = DirHandle->new($path);
- my @children;
- while (defined(my $ent = $dh->read)) {
- next if $ent eq '.';
- push @children, $ent;
- }
+ my $is_dir = -d $file;
+ my @stat = stat _;
- my @page_files;
-
- for my $basename (sort { $a cmp $b } @children) {
- push @page_files, $basename if $basename =~ m/\.(jpg|gif)$/;
- my $file = "$path/$basename";
- my $url = $dir_url . $basename;
- my $is_dir = -d $file;
- my @stat = stat _;
+ $url = join '/', map {uri_escape($_)} split m{/}, $url;
+ if ($is_dir) {
+ $basename .= "/";
+ $url .= "/";
+ }
- $url = join '/', map {uri_escape($_)} split m{/}, $url;
+ my $mime_type = $is_dir ? 'directory' : ( Plack::MIME->mime_type($file) || 'text/plain' );
+ push @files, [ $url, $basename, $stat[7], $mime_type, HTTP::Date::time2str($stat[9]) ];
+ }
- if ($is_dir) {
- $basename .= "/";
- $url .= "/";
- }
+ } else {
+ die "Unsupported format: $path";
+ }
- my $mime_type = $is_dir ? 'directory' : ( Plack::MIME->mime_type($file) || 'text/plain' );
- push @files, [ $url, $basename, $stat[7], $mime_type, HTTP::Date::time2str($stat[9]) ];
- }
+ if ( @page_files ) {
+ @page_files = sort_pages @page_files;
+ warn "# page_files = ",dump( @page_files );
+ }
my $dir = Plack::Util::encode_html( $env->{PATH_INFO} );
my $page = 'empty';
if ( $req->param('bookreader') ) {
- my $pages;
+ my $pages; # []
my $pages_path = "cache/$dir_url/bookreader.json";
- if ( 0 && -e $pages_path ) {
+ if ( -e $pages_path ) {
$pages = decode_json read_file $pages_path;
} else {
- $pages = [
- grep { defined $_ }
- map {
- my ( $w, $h ) = imgsize("$path/$_");
- $w && $h ? [ "$dir_url/$_", $w, $h ] : undef
- } sort { $a <=> $b } @page_files
- ];
+ foreach my $page ( @page_files ) {
+ my $image = Graphics::Magick->new;
+ if ( $page =~ m/\.pdf$/ ) {
+ die "$path/$page: $!" unless -r "$path/$page";
+
+ my $info = `pdfinfo $path/$page`;
+ warn "# pdfinfo $path/$page\n$info\n";
+ my $pdf_pages = $1 if ( $info =~ m/Pages:\s*(\d+)/s );
+ die "can't find number of pages for $path/$page in:\n$pdf_pages\n" unless $pdf_pages;
+
+
+ $pdf_pages = $ENV{MAX_PAGES} if $pdf_pages > $ENV{MAX_PAGES}; # FIXME
+
+ foreach my $nr ( 1 .. $pdf_pages ) {
+ my $cache_path = "cache/$dir_url/$page";
+ my $page_url = convert_pdf_page( "$path/$page", $nr, $cache_path . '.' . $nr );
+ warn "## ping $page_url\n";
+ my ( $w, $h, $size, $format ) = $image->ping($page_url);
+ warn "## image size $w*$h $size $format $page_url\n";
+ my $url = decode('utf-8',"/$page_url");
+ push @$pages, [ $url, $w, $h ] if $w && $h;
+ }
+
+ } else {
+ die "$path/$page: $!" unless -r "$path/$page";
+ my ( $w, $h, $size, $format ) = $image->ping("$path/$page");
+ warn "# image size $w*$h $size $format $path/$page\n";
+ my $url = decode('utf-8',"$dir_url/$page");
+ push @$pages, [ $url, $w, $h ] if $w && $h;
+ }
+ }
make_basedir $pages_path;
- write_file $pages_path => encode_json( $pages );
+ write_file $pages_path, => encode_json( $pages );
warn "# created $pages_path ", -s $pages_path, " bytes\n";
}
warn "# pages = ",dump($pages);
- $page = sprintf $reader_page, $dir, encode_json( $pages ), $dir; # FIXME: title
+ $page = sprintf $reader_page, $dir, encode_json( $pages ), $dir, '..';
} else {
sprintf $dir_file, map Plack::Util::encode_html($_), @$f;
} @files;
- $page = sprintf $dir_page, $dir, $dir, $files;
+ $page = sprintf $dir_page, $dir, $dir, $files,
+ @page_files ? '<form><input type=submit name=bookreader value="Read"></form>' . dump( [ @page_files ] ) : '';
}