#!/usr/bin/perl
# 
# epub-rename -- renames epub-files according to metatags 
#
# Author:  Peter Keel <seegras@discordia.ch>
# Date:    19.01.2011
# Changed: 12.04.2013
# Version: 0.9
# License: Artistic License 2.0
# URL:     https://seegras.discordia.ch/Programs/
# 
# Incorporates TitleCase originally by John Gruber:
# https://daringfireball.net/2008/05/title_case
# Re-written and much improved by Aristotle Pagaltzis:
# https://plasmasturm.org/code/titlecase/
# which also lies under the MIT-License.
#

use Getopt::Long;
use Pod::Usage;
use Encode;
use open qw/:std :utf8/;

&Getopt::Long::Configure( 'pass_through', 'no_autoabbrev');
&Getopt::Long::GetOptions(
		'compat|c'		=> \$compat,
		'debug|d'		=> \$debug,
		'fixtitle|t'		=> \$fixtitle,
		'fixauthor|f'		=> \$fixauthor,
		'fixseries|s'		=> \$fixseries,
		'help|h'		=> \$needshelp,
		'rename|r'		=> \$rename,
		'exchange|x'		=> \$exchange,
		'verbose|v'		=> \$verbose,
);

if (!$ARGV[0]) {
    $dname = ".";
} else { 
    $dname=$ARGV[0]; 
}

if ($needshelp) {
pod2usage(1);
}

opendir(IN_DIR, $dname) || die "I am unable to access that directory...Sorry";
@dir_contents = readdir(IN_DIR);
closedir(IN_DIR);

@dir_contents = sort(@dir_contents);
    foreach $x (@dir_contents) {
	if ($x ne ".." and $x ne ".") {
	    if ($x =~m/.*\.epub$/i) {
		$y = $x;
		if ($compat) {
		    getmetacompat($y);
		} else {
		    getmeta($y);
		}
		if ($exchange) {
		    exchangetitle($y);
		}
		if ($fixtitle) {
		    fix_title($y);
		}
		if ($fixauthor) {
		    fix_author($y);
		}
		if ($fixseries) {
		    fix_series($y);
		}
		if ($rename) {
		    renepub();
		}
	    }
	}
    }

sub exchangetitle {
    my $fname = $_[0];
    # BTW: don't check return values, they're wrong
    $fname = decode("utf-8", $fname);
    system("export LC_CTYPE=en_GB.UTF-8; ebook-meta \"$fname\" --authors=\"$oldtitle\" --title=\"$oldauth\"");
    print "changed $fname\n";
}

sub fix_title {
    my $fname = $_[0];
    if (! ($oldtitle eq $title)) {
        $fname = decode("utf-8", $fname);
	#my $mytitle = encode("iso-8859-1", $title);
	my $mytitle = $title;
	system("export LC_CTYPE=en_GB.UTF-8; ebook-meta \"$fname\" --title=\"$mytitle\"");
	print "changed $y $oldtitle to $mytitle\n";
    }
}

sub fix_series {
    my $fname = $_[0];
    my $newstr = $title;
    if ($newstr =~ m/[^#]+#?[0-9]+ ?[-:] ?.*/) {
    ($newseries, $newseriesid, $newtitle) = $newstr =~ /([^#]+)#?([0-9\.]+) ?[-:] ?(.*)/i;
    } elsif ($newstr =~ m/[^\(]+\([^#]+#?[0-9\.]+\)/) {
    ($newtitle, $newseries, $newseriesid) = $newstr =~ /([^\(]+)\(([^#]+)#?([0-9\.]+)\)/i;
    } elsif ($newstr =~ m/[^-]+-[^#]+#?[0-9\.]+/) {
    ($newtitle, $newseries, $newseriesid) = $newstr =~ /([^-]+)-([^#]+)#?([0-9\.]+)/i;
    #} elsif ($newstr =~ m/[^0-9\[]+\[[0-9]+\] ?[-:]? ?.*/) {
    #($newseries, $newseriesid, $newtitle) = $newstr =~ /^([^0-9\[]+)\[([0-9]+)\] ?[-:]? ?(.*)$/i;
    #} elsif ($newstr =~ m/[^0-9]+[0-9]+ ?[-:]? ?.*/) {
    #($newseries, $newseriesid, $newtitle) = $newstr =~ /^([^0-9]+)([0-9]+) ?[-:]? ?(.*)$/i;
    } else {
    $newseries = "";
    $newtitle = $title;
    }
    if ((! $series) && ($newseries ne "")) {
    $newseries =~ s/^\s+//;
    $newseries =~ s/\s+$//;
    $newseries =~ s/\s+/ /g;
    $newseries = ucfirst($newseries);
    $series = $newseries;
    }
    if ((! $seriesid) && $newseriesid ) {
    $newseriesid =~ s/^\s+//;
    $newseriesid =~ s/\s+$//;
    $seriesid = $newseriesid;
    }
    $newtitle =~ s/^\s+//;
    $newtitle =~ s/\s+$//;
    $newtitle = ucfirst($newtitle);

#    print "$series | $seriesid | $newtitle\n";
    
    if (! ($oldtitle eq $newtitle)) {
        $fname = decode("utf-8", $fname);
	system("export LC_CTYPE=en_GB.UTF-8; ebook-meta \"$fname\" --series=\"$series\" --index=\"$seriesid\" --title=\"$newtitle\"");
    }
}

# this is intrusive, will change every file, title too
sub fix_author {
    my $fname = $_[0];
    my $newauth = "";
    my $cnt = 1;
    while ($cnt <= $authcount) {
	$newauth .= $author[$cnt];
	$cnt++;
	if ($cnt <= $authcount) {
	    $newauth .= " & ";
	}
    }
    #$newauth = encode("iso-8859-1", $newauth);
    $fname = decode("utf-8", $fname);
    #my $mytitle = encode("iso-8859-1", $title);
    $mytitle = $title;
    system("export LC_CTYPE=en_GB.UTF-8; ebook-meta \"$fname\" --title=\"$mytitle\" --authors=\"$newauth\"");
    if ($date) {
	system("export LC_CTYPE=en_GB.UTF-8; ebook-meta \"$fname\" --title=\"$mytitle\" --authors=\"$newauth\" --date=$date");
    }
    print "changed $fname $authors - $oldtitle to $newauth - $mytitle\n";
}

# FIXME: code deciding new name stupid

sub renepub {
my $counter=0;
my $exname = $newname;
$oldname = decode("utf-8", $y);
$newfullname = $dname . "/" . $newname;
$oldfullname = $dname . "/" . $oldname;
    if ($newfullname ne $oldfullname) {
	$exname = $newfullname;
	($exname,$suffix) = $exname =~ /^(.*)(\.[^.]*)$/;
	while (-e "$newfullname") {    
	    print ("WARN \"$newfullname\" exists\n");
	    $counter++;
	    $newfullname = $exname . "-" . $counter . $suffix;
	}
	rename ("$oldfullname", "$newfullname") unless ($newfullname eq $oldfullname);
	print ("renamed $oldfullname to \"$newfullname\"\n") unless ($newfullname eq $oldfullname);
    }
}

sub regex_author {
    my $string = $_[0];
    if ($debug) { print (stderr " string-in : '$string'\n"); };
    $string =~ s/\(Ebook By Undead\)//goi;
    $string =~ s/\[ss\]//goi;
    $string =~ s/\(ss\)//goi;
    $string =~ s/([A-Z][.])([^ ])/$1 $2/go;
    $string =~ s/([A-Z]){1,}([A-Z]){1,}\s+/$1\. $2\. /go;
    $string =~ s/([A-Z]){1,}\s+/$1\. /go;
    $string =~ s/\N{U+0060}/\'/go;
    $string =~ s/\N{U+2018}/\'/go;
    $string =~ s/\N{U+2019}/\'/go;
    $string =~ s/\N{U+201A}/\'/go;
    $string =~ s/\N{U+201B}/\'/go;
    $string =~ s/\N{U+2010}/-/go;
    $string =~ s/\N{U+2012}/-/go;
    $string =~ s/\N{U+2013}/-/go;
    $string =~ s/\N{U+2014}/-/go;
    $string =~ s/-/ /go;
    $string =~ s/\//_/go;
    $string =~ s/([A-Z]{3,})/\L$1/go;
    $string =~ s/(\w+)/\u$1/go;
    $string =~ s/  / /go;
    $string =~ s/^\s+//go;
    $string =~ s/\s+$//go;
    if ($debug) { print (stderr " string-out: '$string'\n"); };
    return $string; 
}

sub regex_title {
    my $string = $_[0];
    my $newstring = "";
    my @small_words = qw( (?<!q&)a an and as at(?!&t) but by en for if in of on or the to v[.]? via vs[.]? );
    my $small_re = join '|', @small_words;

    if ($debug) { print (stderr " string-in : '$string'\n"); };

    $string =~ s/\(\xc3\x9cbersetzung\)$//goi;
    $string =~ s/\(Übersetzung\)$//goi;
    $string =~ s/\(Original\)$//goi;
    $string =~ s/\(Dodo Press\)$//goi;
    $string =~ s/\(Epub\)$//goi;
    $string =~ s/\(German Edition\)$//goi;
    $string =~ s/\[ss\]//goi;
    $string =~ s/\(ss\)//goi;
    $string =~ s/\(Vintage Contemporaries\)$//goi;
    $string =~ s/\(Barnes & Noble Classics Series\)$//goi;
    $string =~ s/\(Yesterday\'s Classics\)$//goi;
    $string =~ s/\([Vv][0-9]\.[0-9]\)$//goi;
    $string =~ s/\([Vv][0-9]\)$//goi;
    $string =~ s/\- [Vv][0-9]$//goi;
    $string =~ s/\s+$//go;
    $string =~ s/\(html\)$//goi;
    $string =~ s/\s+$//go;
    $string =~ s/\([0-9][0-9][0-9][0-9]\)$//goi;
    $string =~ s/\N{U+0060}/\'/go;
    $string =~ s/\N{U+2018}/\'/go;
    $string =~ s/\N{U+2019}/\'/go;
    $string =~ s/\N{U+201A}/\'/go;
    $string =~ s/\N{U+201B}/\'/go;
    $string =~ s/\N{U+2010}/-/go;
    $string =~ s/\N{U+2012}/-/go;
    $string =~ s/\N{U+2013}/-/go;
    $string =~ s/\N{U+2014}/-/go;
    $string =~ s/\ - Roman$//goi;
    $string =~ s/\//_/go;
    $string =~ s/\!//go;
    $string =~ s/\"/\'/go;
    $string =~ s/\.$//go;
    $string =~ s/-$//go;
    #$string =~ s/([\w\']+)/\u\L$1/g;
    $string =~ s/([A-Z]{3,})/\L$1/go;
    # $string =~ s/([^ (]+)/\u$1/go;
    $string =~ s/\s+/ /go;
    $string =~ s/^\s+//go;
    $string =~ s/\s+$//go;
    
    my $apos = qr/ (?: [\'] [[:lower:]]* )? /x;
    foreach ($string) {
	s{
		\b (_*) (?:
			( [-_[:alpha:]]+ [@.:/] [-_[:alpha:]@.:/]+ $apos ) # URL, domain, or email
			|
			( (?i: $small_re ) $apos )                         # or small word (case-insensitive)
			|
			( [[:alpha:]] [[:lower:]'()\[\]{}]* $apos )       # or word w/o internal caps
			|
			( [[:alpha:]] [[:alpha:]'()\[\]{}]* $apos )       # or some other word
		) (_*) \b
	}{
		$1 . (
		  defined $2 ? $2         # preserve URL, domain, or email
		: defined $3 ? "\L$3"     # lowercase small word
		: defined $4 ? "\u\L$4"   # capitalize word w/o internal caps
		: $5                      # preserve other kinds of word
		) . $6
	}exgo;

	# exceptions for small words: capitalize at start and end of title
	s{
		(  \A [[:punct:]]*         # start of title...
		|  [:.;?!][ ]+             # or of subsentence...
		|  [ ][\'\"(\[][ ]*     )  # or of inserted subphrase...
		( $small_re ) \b           # ... followed by small word
	}{$1\u\L$2}xigo;

	s{
		\b ( $small_re )      # small word...
		(?= [[:punct:]]* \Z   # ... at the end of the title...
		|   [\'\")\]] [ ] )   # ... or of an inserted subphrase?
	}{\u\L$1}xigo;
	$newstring .= $_;
    }
    if ($debug) { print (stderr " string-out: '$newstring'\n"); };
    return $newstring;
}

sub getmeta {
my $fname = $_[0];
open(META,"epub-meta -atm \"$fname\"|") || die "Failed: $!\n";

$title ="";
$author ="";
$series ="";
$seriesid ="";
$authcount ="";
$auth = "";

    if ($debug) { print (stderr "$fname\n"); };

    while ( <META> ) {
	if ($_ =~ m/^Title: (.*)/i) {
	    $title = $1;
	    $oldtitle = $title;
	    if ($title =~ m/\([0-9][0-9][0-9][0-9]\)$/) {
		$date = $title;
		$date =~ s/.*\(([0-9][0-9][0-9][0-9])\)$/$1/;
	        if ($debug) { print (stderr " date: $date\n"); };
	    };
	    if ($title =~ m/(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?([0-9xX])$/) {
		# the bloody title is an ISBN instead.
		$isbn = $title;
	    };
	    $title = regex_title($title);
	} elsif ($_ =~ m/^Author: (aut|Author): ([^(,]*)?,([^(,]*)\(.*\(.*\).*\)/i) {
	    $authcount++;
	    $author[$authcount] = $3 . " " . $2;
	    $oldauthor[$authcount] = $2 . " " . $3;
	    $author[$authcount] = regex_author($author[$authcount]);
	} elsif ($_ =~ m/^Author: (aut|Author): ([^(,]*)?,([^(,]*)\(.*\)/i) {
	    $authcount++;
	    $author[$authcount] = $3 . " " . $2;
	    $oldauthor[$authcount] = $2 . " " . $3;
	    $author[$authcount] = regex_author($author[$authcount]);
	} elsif ($_ =~ m/^Author: (aut|Author): (.*)?\(.*\(.*\)\)/i) {
	    $authcount++;
	    $author[$authcount] = $2;
	    $oldauthor[$authcount] = $2;
	    $author[$authcount] = regex_author($author[$authcount]);
	} elsif ($_ =~ m/^Author: (aut|Author): (.*)?\(.*\)/i) {
	    $authcount++;
	    $author[$authcount] = $2;
	    $oldauthor[$authcount] = $2;
	    $author[$authcount] = regex_author($author[$authcount]);
	} elsif ($_ =~ m/^Author: (aut|Author): (.*)?,(.*)/i) {
	    $authcount++;
	    $author[$authcount] = $3 . " " . $2;
	    $oldauthor[$authcount] = $2 . " " . $3;
	    $author[$authcount] = regex_author($author[$authcount]);
	} elsif ($_ =~ m/^Author: (aut|Author): (.*)/i) {
	    $authcount++;
	    $author[$authcount] = $2;
	    $oldauthor[$authcount] = $2;
	    $author[$authcount] = regex_author($author[$authcount]);
	} elsif ($_ =~ m/^Meta: calibre:series: (.*)/i) {
	    $series = $1;
	    $series = regex_title($series);
	    $series =~ s/-/ /go;
	} elsif ($_ =~ m/^Meta: calibre:series_index: (.*)/i) {
	    $seriesid = $1;
	    $seriesid = sprintf("%02d", $seriesid);
	}
    }
    if ($authcount == 0) {
	$auth = "unknown";
	$oldauth = "unknown";
    } elsif ($authcount == 1) {
	$auth = $author[1];
	$oldauth = $oldauthor[1];
    } elsif ($authcount == 2) {
	$auth = $author[1] . " & " . $author[2];
	$oldauth = $oldauthor[1] . " & " . $oldauthor[2];
    } elsif ($authcount == 3) {
	$auth = $author[1] . " & " . $author[2] . " & " . $author[3];
	$oldauth = $oldauthor[1] . " & " . $oldauthor[2] . " & " . $oldauthor[3];
    } else {
	$auth = "anthology"; 
	$oldauth = "anthology"; 
    }

    if ($series) {
	$newname = $auth . " - " . $series . $seriesid . " - " . $title . ".epub";
    }
    else {
	$newname = $auth . " - " . $title . ".epub";
    }
    if ($verbose) {
	print "$fname \"$newname\"\n"; 
    }
}

sub getmetacompat {
my $fname = $_[0];
open(META,"ebook-meta \"$fname\"|") || die "Failed: $!\n";

$title ="";
$author ="";
$series ="";
$seriesid ="";

while ( <META> ) {
    if ($_ =~ m/Title               : (.*)/i) {
	$title = $1;
	$oldtitle = $title;
	$title = regex_title($title);
	if ($title =~ m/(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?(\d)[- ]?([0-9xX])$/) {
	# the bloody title is an ISBN instead.
	$isbn = $title;
	}
    } elsif ($_ =~ m/Author\(s\)           : (.*)?,(.*)\[.*\]/i) {
	$author = $2 . $1;
	$author =~ regex_author($author);
    } elsif ($_ =~ m/Author\(s\)           : (.*)\[.*\]/i) {
	$author = $1;
	$author =~ regex_author($author);
    } elsif ($_ =~ m/Author\(s\)           : (.*)?,(.*)/i) {
	$author = $2 . $1;
	$author =~ regex_author($author);
    } elsif ($_ =~ m/Author\(s\)           : (.*)/i) {
	$author = $1;
	$author =~ regex_author($author);
    } elsif ($_ =~ m/Series              : (.*) \#(\d.*)/i) {
	$series = $1;
	$series = regex_title($series);
	$seriesid = $2;
	$seriesid = sprintf("%02d", $seriesid);
    }
}
if ($series) {
    $newname = $author . " - " . $series . $seriesid . " - " . $title . ".epub";
}
else {
    $newname = $author . " - " . $title . ".epub";
}
if ($verbose) {
    print "$fname \"$newname\"\n"; 
}


}
__END__

=head1 NAME

epub-rename - Rename Epub-files according to metadata

=head1 SYNOPSIS

<This program> [options] [directory ...]

 Options:
   -c|--compat
   -t|--fixtitle
   -f|--fixauthor
   -s|--fixseries
   -h|--help
   -r|--rename
   -x|--exchange
   -v|--verbose

=head1 OPTIONS

=over 8


=item B<-c|--compat>

Use ebook-meta from calibre instead of epub-meta. Much slower.

=item B<-t|--title>

Fix title. This means the tag gets sanitized, as it would if destined 
for a filename, and then written back to the metadata. Uses ebook-meta.

=item B<-s|--fixseries>

Fix series. This tries to find out the series and series-index from the 
title, sets it, and strips it from the title. Uses ebook-meta. Works 
with titles in the format "series #index - title" and some variations
thereof. 

=item B<-f|--fixauthor>

Fix all tags: author, title, and in some cases date. Uses ebook-meta
and touches every file, even those that don't need fixing. Slow.

=item B<-h|--help>

Print a brief help message and exit.

=item B<-r|--rename>

rename files to the pattern "Author - Series SeriesIndex - Title"

=item B<-x|--exchange>

changes title for author-tag and vice versa. For all those files
that have the author in the title-field and the title in the author-
field. Uses ebook-meta, thus is slow. 

=item B<-v|--verbose>

Show how all files would be renamed, not just those really renamed.

=back

=head1 DESCRIPTION

B<This program> will rename all epub-files in the directory you're in, 
unless given another directory. Per default, it will only show what
it wants to do. 

=head1 ISSUES

B<This program> needs epub-meta (from me) and ebook-meta (from Calibre)

B<1> Rename with -r first, if you see any files named 
"unknown - unknown.epub" you need to fix them first with 
ebook-convert (also from Calibre). Don't run this program 
with -f, -x, -t switches on them. 

B<2> If you're not using an UTF-8 capable terminal, you will see 
garbled output of ebook-meta. Don't worry, the file contents will
be fine.

B<3> If some idiot put a title containing commas or brackets into 
the authors field, the title will be very wrong after <This program -x>.

=cut