#! /usr/bin/perl
#
#     $Id: evlookup,v 1.9 2013/02/12 07:30:07 doi Exp doi $
#
#     evlookup - look up event information
#
#     this program will look up the information in the anime event
#     database, which is maintained by Hitoshi Doi <doi@usagi.org>,
#     and print out information for each voice actor.
#     it is mainly used for creating parts of the WWW seiyuu pages.
#
#     the seiyuu information is available via WWW at:
#             http://www.usagi.org/doi/seiyuu/
#
#     original hack by Hitoshi Doi (doi@usagi.org), 1995.02.05
#
#     this program is public domain.  do whatever you want with it.
#     this program is offered "as is", and does not have any warrantee.
#
#     ---------------------------------------------------------------
#     modifications
#
#     1995.06.30        Hitoshi Doi
#             use associative arrays to speed things up
#
#     1996.03.07        Hitoshi Doi
#             modifications to process non-event databases
#             the video DB and CD DB might also use this format
#     ---------------------------------------------------------------

    $me = $0;
    $me =~ s/.*\///;

    # the file containing the voice actor information
    $evfile = $ENV{'EVENT'};

    $s_count = 0;
    @s_name = ();
    @s_url = ();
    @s_events = ();
    @s_date = ();
    $www_all = 0;
    @ev_lookup = ();
    $ev_ct = 0;
    $nodate = 0;
    $listall = '';
    $listall2 = '';
    $match = '';

    $WWW_subdir = 'EV';                # subdirectory for WWW data

    while ($ai = shift(@ARGV)) {
        if ($ai =~ m/^-db/i) {
            $evfile = shift(@ARGV);
        }
        elsif ($ai =~ m/^-sub/i) {
            $WWW_subdir = shift(@ARGV);
        }
        elsif ($ai =~ m/^-www/i) {
            $www_all = 1;
        }
        elsif ($ai =~ m/^-printall/i) { # special option for DB.event
            $printall = 1;
        }
        elsif ($ai =~ m/^-nodate/i) {
            $nodate = 1;
        }
        elsif ($ai eq '-listall') {
            $listall = shift(@ARGV);
        }
        elsif ($ai eq '-listall2') {
            $listall2 = shift(@ARGV);
        }
        elsif ($ai =~ m/^-match/i) {
            $match = shift(@ARGV);
        }
        elsif ($ai =~ m/^-help/i) {
            print "$me - event information lookup
usage: $me [options]
options:  NAME       display all events for the actor NAME
         -www        create the WWW event data for all voice actors
         -sub SUB    set the WWW subdirectory
         -nodate     do not write out the date for the WWW data
         -listall P  list all of the events with URLs, preceeded by Path
         -listall2 P list all of the events (HTML) with URLs, preceeded by Path
         -db FILE    use the file FILE as the event database\n";
            exit 0;
        }
        elsif ($ai =~ m/^-debug/i) {
            $debug = 1;
        }
        else {
            $ev_lookup[$ev_ct++] = $ai;
        }
    }

    if (!$evfile) {
        print "$me: the event database is undefined.
please set the environment variable EVENT,
or specify the database with the -db option.\n";
        exit -1;
    }
    $ev_db = "event database $evfile";
    if (!open(EV, "<$evfile")) {
        print "$me: can't open the $ev_db.\n";
        exit -1;
    }

    &read_profile;
    $style = " onMouseOver=\"this.style.background='pink'\"";
    $style .= " onMouseOut=\"this.style.background=''\"";

    #
    # go through the event file and do processing
    #
    $read_done = 0;
    $ev_proc = 0;
    while (1) {
        if (!$read_done) {
            $l = <EV>;
            last if (!$l);        # end of file
            chop($l);
        }
        $read_done = 0;
        next if ($l =~ m/^$/);
        next if ($l =~ m/^#/);

        if ($l =~ m/^[1-9][0-9]/) {        # new event
            $event_OK = ($match) ? ($l =~ m/$match/i) : 1;
            ($ev_date, $ev_name) = split(/[ ]+/, $l, 2);
            $ev_url = '';
            if ($debug) {
                print "event $ev_name, date $ev_date\n";
            }
            next;
        }
        $l =~ s/^[ ]+//i;

        #
        # process keywords
        #
        if ($l =~ m/^URL/i) {                # URL keyword
            $l =~ s/^URL[ ]*//i;
            $ev_url = $l;
            if ($debug) {
                print "event URL $ev_url\n";
            }
            if ($listall && $event_OK) {
                print "<LI>$ev_date ";
                print "<A HREF=\"$listall/$ev_url\">";
                print "$ev_name</A>\n";
            }
            if ($listall2 && $event_OK) {
                print "<TR $style valign=top><TD>$ev_date</TD><TD>";
                print "<A HREF=\"$listall2/$ev_url\">$ev_name</A></TD></TR>\n";
            }
            next;
        }
        #
        # others are seiyuu names
        #
        $role = '';
        if ($l =~ m/\{.*\}/) {
            ($nn, $role) = split(/\{/, $l, 2);
            $role =~ s/\}//;
            $nn =~ s/[ ]+$//;
            $l = $nn
        }
        if ($printall) {
            print "$ev_date - ";
            if ($ev_url) {
                print "<A HREF=\"/doi/$ev_url\">$ev_name</A>";
            }
            else {
                print "$ev_name";
            }
            print " {$role}" if ($role);
            my $va = $l;
            $va .= " ($kanji{$va})" if ($kanji{$va});
            print " - $va\n";
            next;
        }
        $s_id = &set_seiyuu($l);
        if ($debug) {
            print "seiyuu ($s_id) $s_name[$s_id]\n";
        }
        $s_events[$s_id] .= "$ev_name\t";
        $s_url[$s_id] .= "$ev_url\t";
        $s_date[$s_id] .= "$ev_date\t";
    }

    #
    # print out information
    #
    if ($www_all) {
        &write_www();
    }
    if ($ev_ct) {
        for ($i = 0; $i < $ev_ct; $i++) {
            print "\n";
            $s_id = &find_seiyuu($ev_lookup[$i]);
            if ($s_id == $s_count - 1) {
                $s_count--;
                print "cannot find $ev_lookup[$i]\n";
                next;
            }
            print "$s_name[$s_id]\n";
            @e_name = split(/\t/, $s_events[$s_id]);
            @e_date = split(/\t/, $s_date[$s_id]);
            while ($en = shift(@e_name)) {
                $ed = shift(@e_date);
                print " $ed  $en\n";
            }
        }
    }
    exit 0;

#
# subroutines
#

sub set_seiyuu
{
    my $name = $_[0];
    my $i;

    if (!($i = $seiyuu_id{$name})) {
        $s_count++;
        $seiyuu_id{$name} = $s_count;
        $i = $s_count;
        $s_name[$i - 1] = $name;
    }
    return $i - 1;
}

sub find_seiyuu
{
    my $name = $_[0];
    my $k;

    foreach $k (sort keys(%seiyuu_id)) {
        if ($k =~ m/$name/i) {
            return($seiyuu_id{$k} - 1);
        }
    }
    return $s_count + 1;
}

#
# write data for WWW use
#
sub write_www
{
    my($i, @e_name, @e_url, @e_date);
    my($en, $eu, $ed);
    my $www_index = "$WWW_subdir/Index";
    my $www_out;

    system "/bin/mv $WWW_subdir $WWW_subdir.old";
    mkdir($WWW_subdir, 0755);
    open(INDEX, ">$www_index");
    for ($i = 0; $i < $s_count; $i++) {
        printf INDEX "%05d:%s\n", $i, $s_name[$i];
        $www_out = sprintf("$WWW_subdir/%05d", $i);
        open(WWWOUT, ">$www_out");
        @e_name = split(/\t/, $s_events[$i]);
        @e_url = split(/\t/, $s_url[$i]);
        @e_date = split(/\t/, $s_date[$i]);
        while ($en = shift(@e_name)) {
            $ed = shift(@e_date);
            $eu = shift(@e_url);
            print WWWOUT "<TR valign=top><TD>";
            print WWWOUT "$ed</TD><TD>" if (!$nodate);
            if ($eu) {
                print WWWOUT "<A HREF=\"../$eu\">";
            }
            print WWWOUT "$en";
            if ($eu) {
                print WWWOUT "</A>";
            }
            print WWWOUT "</TD></TR>\n";
        }
        close(WWWOUT);
    }
    close(INDEX);
}

sub read_profile
{
    $pf = '/home/doi/www/seiyuu/db/DB.profile';
    open(PF, "<$pf");
    my @all = <PF>;
    close(PF);
    %kanji = ();
    foreach $pl (@all) {
        chop($pl);
        ($r, $k) = split(/\t/, $pl, 2);
        $kanji{$r} = $k;
    }
}