package Anime::ANN;
+use POSIX;
+use Digest::MD5;
use LWP::UserAgent;
use HTML::Entities;
$ver = "0.1";
+$ua = LWP::UserAgent->new;
+$ua->agent("ANNData/$ver ");
+$usecache = 1;
sub _get
{
- my($uri, $ua, $res);
+ my($uri, $cname, $res);
($uri) = @_;
- $ua = LWP::UserAgent->new;
- $ua->agent("ANNData/$ver ");
-
+ $cname = $ENV{"HOME"} . "/.ann/";
+ mkdir $cname unless -e $cname;
+ $cname .= "cache/";
+ mkdir $cname unless -e $cname;
+ $cname .= Digest::MD5::md5_hex $uri;
+ if($usecache && -e $cname) {
+ my(@s);
+ @s = stat $cname;
+ if((time - $s[9]) < 86400) {
+ my($buf);
+ open CACHE, "<:utf8", $cname;
+ $res .= $buf while read CACHE, $buf, 1024;
+ close CACHE;
+ return $res;
+ }
+ }
+
$res = $ua->request(HTTP::Request->new("GET", "$uri"));
-
- die "could not fetch $uri\n" unless $res->is_success;
+
+ if(open CACHE, ">:utf8", $cname) {
+ print CACHE $res->content;
+ close CACHE;
+ }
+
+ return undef unless $res->is_success;
return $res->content;
}
my($name, $il, $html, @ret);
($name) = @_;
+ $name = ($name =~ /^(the\s+)?(.*)$/i)[1];
$il = uc(($name =~ /^(.)/)[0]);
$il = "9" if (!($il =~ /[A-Z]/));
- $html = _get "http://www.animenewsnetwork.com/encyclopedia/anime.php?list=$il";
+ if(!($html = _get "http://www.animenewsnetwork.com/encyclopedia/anime.php?list=$il")) {
+ return undef;
+ }
# The only way to recognize entries that seems sure is to look
# after the "HOVERLINE" class.
- push @ret, $1 while $html =~ /<A\s.*CLASS=HOVERLINE\s.*>.*<FONT.*>([^<>]*$name[^<>]*)<\/FONT/ig;
+ while($html =~ /<A\s.*CLASS=HOVERLINE\s.*HREF=\"([^\"]+)\".*>(\<small\>.*\<\/small\>)?([^<]+)<\//ig) {
+ if((substr "" . lc $3 , 0, length $name) eq lc $name) {
+ push @ret, $3;
+ }
+ }
+ # push @ret, $1 while $html =~ /<A\s.*CLASS=HOVERLINE\s.*>.*<FONT.*>([^<>]*$name[^<>]*)<\/FONT/ig;
- return(@ret);
+ return @ret;
}
sub getid
my($name, $il, $html, $url);
($name) = @_;
+ $name = ($name =~ /^(the\s+)?(.*)$/i)[1];
$il = uc(($name =~ /^(.)/)[0]);
$il = "9" if (!($il =~ /[A-Z]/));
- $html = _get "http://www.animenewsnetwork.com/encyclopedia/anime.php?list=$il";
+ if(!($html = _get "http://www.animenewsnetwork.com/encyclopedia/anime.php?list=$il")) {
+ return undef;
+ }
# The only way to recognize entries that seems sure is to look
# after the "HOVERLINE" class.
- (($url) = ($html =~ /<A\s.*CLASS=HOVERLINE\s.*HREF=\"([^\"]+)\".*$name/i)) || return;
+ while($html =~ /<A\s.*CLASS=HOVERLINE\s.*HREF=\"([^\"]+)\".*>(\<small\>.*\<\/small\>)?([^<]+)<\//ig) {
+ if((substr "" . lc $3 , 0, length $name) eq lc $name) {
+ return ($1 =~ /id=(\d+)$/)[0];
+ }
+ }
- return ($url =~ /\?id=(\d+)$/)[0];
+ return undef;
}
sub geturl
my($html, $kind, @ret);
($html, $kind) = @_;
- if($html =~ /$kind theme:<\/b>\n/igc) {
+ if($html =~ /$kind theme:<\/strong>\s*\n/igc) {
my(@parts, $ct, $buf);
- while($html =~ /\G\<br\> (([^<>]|\<i\>|<\/i>)+)/igc) {
+ while($html =~ /\G\s*\<div\>(([^<>]|\<i\>|<\/i>)+)<\/div>/igc) {
$buf = $1;
- # 0 1 2 3 4 5 6 7 8 9 10 11
- if(@parts = ($buf =~ /(\#(\d+):)?\s*\"([^\"\(]+)(\s+\((\<i\>(.*)<\/i>(;\s*)?)?([^<>]+)?\))?\"\s+by\s+([^\(]*[^\(\s])(\s*\(eps (\d+)-(\d+)?\))?/i)) {
+ # 0 1 2 3 4 5 6 7 8 9 10 11
+ if(@parts = ($buf =~ /(\#(\d+):)?\s*\"([^\"\(]+\S)(\s*\((\<i\>(.*)<\/i>( - \s*)?)?([^<>]+)?\))?\"\s+by\s+([^\(]*[^\(\s])(\s*\(eps (\d+)-(\d+)?\))?/i)) {
$ct = {};
$ct->{"num"} = $parts[1] if defined $parts[1];
if(defined $parts[5]) {
my($id, $buf, $html, %ret);
($id) = @_;
- $html = _get geturl $id;
+ if(!($html = _get geturl $id)) {
+ return undef;
+ }
$ret{"url"} = geturl $id;
- ($buf) = ($html =~ /\<TITLE\>Anime News Network - ([^<]*)<\/TITLE>/);
+ ($buf) = ($html =~ /\<title\>([^<]*) - Anime News Network<\/title>/);
if($buf =~ /\([^\)]+\)$/) {
($ret{"name"}, $ret{"type"}) = ($buf =~ /^(.*[^\s])\s*\(([^\)]+)\)$/);
} else {
$ret{"name"} = $buf;
}
- if(($buf) = ($html =~ /vintage:<\/b>\n([^<]+)</is)) {
+ if(($buf) = ($html =~ /vintage:<\/strong>\s*\n\s*<span>([^<]+)</is)) {
$ret{"vtg"} = $buf;
}
- if(($buf) = ($html =~ /number of episodes:<\/b>\n([^<]+)</is)) {
+ if(($buf) = ($html =~ /number of episodes:<\/strong>\s*\n\s*<span>([^<]+)</is)) {
$ret{"eps"} = $buf;
}
+ if(($buf) = ($html =~ /genres:<\/strong>\s*\n\s*([^<]+)</is)) {
+ $ret{"gnr"} = [split /, /, $buf];
+ }
$buf = getthemes $html, "opening";
$ret{"op"} = $buf if(@{$buf});
$buf = getthemes $html, "ending";