利用perl语言抓取脚本的时候,我是拒绝的。因为用perl语言太繁琐。空闲时间在听音乐,突然发现有些歌没办法听,就想起用perl抓个缓存音乐听听。看代码好了。
usestrict;
usewarnings;
useLWP::Simple;
useData::Dumper;
useHTTP::Request::Common;
useJSON;
useDigest::SHA;
useHTTP::Cookies;
useMIME::Base64;
useLWP::UserAgent;
useEncodeqw(decode encode);
my$V= LWP::UserAgent->new();
push@{ $V->requests_redirectable},’POST’;
$V->protocols_allowed([‘http’,’https’]);
$V->agent(“Mozilla/5.0 (Windows NT 6.1; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/38.0.2125.122 UBrowser/4.0.3214.0 Safari/537.36”);
#=========================================================
sub search{
my$n=15;
my$M=shift();
chomp($M);
my$P=shift();
#P是换页;
#N是每以页显示的歌曲数;最多30首;
#M是搜索条件;
my$data;
my$sda=decode(‘gb2312’,$M);
my$rrr=$V->get(“https://c.y.qq.com/soso/fcgi-bin/client_search_cp?ct=24&qqmusic_ver=1298&new_json=1&remoteplace=txt.yqq.center&searchid=52186464057672763&t=0&aggr=1&cr=1&catZhida=1&lossless=0&flag_qc=0&p=”.$P.”&n=”.$n.”&w=”.$sda.”&g_tk=2036544290&jsonpCallback=searchCallbacksong9565&loginUin=854411658&hostUin=0&format=jsonp&inCharset=utf8&outCharset=utf-8¬ice=0&platform=yqq&needNewCode=0″)->content;
my($ssjson) =$rrr=~ /^.*?\((.*)\)$/g;
#print encode(“gb2312”, decode(“utf-8”, $ssjson));
my$uir=decodeJSON($ssjson);
my$digit=0;
foreach my$element(@{$uir->{data}->{song}->{list}})
{
$digit++;
my$sng;
my$name=$element->{title};
my$list=$element->{file}->{media_mid};
print$digit.”. “;
printencode(“gb2312”,decode(“utf-8″,”歌曲名字:”));
printencode(“gb2312”,decode(“utf-8”,$name));
my$song;
print”\n “;
printencode(“gb2312”,decode(“utf-8″,”歌手:”));
foreach my$singer(@{$element->{singer}})
{
my$se=encode(“gb2312”,decode(“utf-8”,$singer->{title}));
push(@{$song},$se);
print$se;
print” “;
}
print”\n”;
#print Dumper $list;
print”\n”;
print”\n”;
$sng->{url} =$list;
$sng->{title} =$song;
$sng->{singer} =encode(“gb2312”,decode(“utf-8”,$name));
push(@{$data},$sng);
}
printencode(“gb2312”,decode(“utf-8″,”本页有”.$digit.”首歌,这是第”.$P.”页\n”));
print”=========================================================\n\n”;
return$data;
}
#=========================================================
sub download{
my$int=shift();
my$data=shift();
#print Dumper $data;
my$h=$V->get(“https://c.y.qq.com/base/fcgi-bin/fcg_music_express_mobile3.fcg?g_tk=737013179&jsonpCallback=MusicJsonCallback778146979323697&loginUin=0&hostUin=0&format=json&inCharset=utf8&outCharset=utf-8¬ice=0&platform=yqq&needNewCode=0&cid=205361747&callback=MusicJsonCallback778146979323697&uin=854411658&songmid=0020UhY82Uhzvq&filename=C400″.@{$data}[$int]->{url}.”.m4a&guid=4605006430″)->content;
my($sssjson) =$h=~ /^.*?\((.*)\)$/g;
my$sss=decodeJSON($sssjson);
my$id=@{$sss->{data}->{items}}[0]->{vkey};
my$e=$V->get(“http://dl.stream.qqmusic.qq.com/C400″.@{$data}[$int]->{url}.”.m4a?vkey=”.$id.”&guid=4605006430&uin=854411658&fromtag=66″)->content;
open(FILE,”>”.@{$data}[$int]->{singer}.”.mp3″);
binmodeFILE;
printFILE$e;
close(FILE);
printencode(“gb2312”,decode(“utf-8″,”下载完成\n\n”));
}
#=============================================================
printencode(“gb2312”,decode(“utf-8″,”请输入搜索关键词 :”));
my$dde= ;
chomp($dde);
my$name=$dde;
my$int=1;
my$data=search($dde,$int);
for(my$r=1;$r>0;$r++){
printencode(“gb2312”,decode(“utf-8″,”请进行编辑,下一页输入(next),上一页请输入(previous),下载请输入(dow+下载的歌曲序号),重新搜索请输入(again),停止运行请输入(exit)\n\n”));
my$dede= ;
print”\n”;
chomp($dede);
if($dede=~ /next/g){
$int++;
$data=search($name,$int);
}elsif($dede=~ /previous/g){
$int–;
if($int<=0){
printencode(“gb2312”,decode(“utf-8″,”输入错误。\n”));
}else{
$data=search($name,$int);
}
}elsif($dede=~ /dow(\d+?)/g){
if($1>15){
printencode(“gb2312”,decode(“utf-8″,”输入错误。\n”));
}else{
download($1-1,$data);
}
}elsif($dede=~ /exit/g){
printencode(“gb2312”,decode(“utf-8″,”正在停止……..\n”));
last;
}elsif($dede=~ /again/g){
printencode(“gb2312”,decode(“utf-8″,”请输入搜索关键词 :”));
my$de= ;
print”\n”;
chomp($de);
$name=$de;
$data=search($de,$int)
}else{
printencode(“gb2312”,decode(“utf-8″,”输入错误。\n”));
}
}
printencode(“gb2312”,decode(“utf-8″,”已退出程序”));
sub sprintJSON{
my$json= JSON->new->allow_nonref;
return$json->encode(shift());
}
sub decodeJSON{
my$json= JSON->new->allow_nonref;
return$json->decode(shift());
}
脚本是通过控制台运行。只有一个搜索和下载,其他的没完善。