[Python]抓取豆瓣电台里被标记喜欢的歌曲名称

一段时间前开始使用豆瓣电台,很喜欢一个功能就是听到一个好听的歌就可以标记上.以后就可以从我的音乐里看到按标记时间排列出的所有曾经标记过的歌曲.但是有个功能豆瓣电台还不支持,就是按照专辑来对所有标记的歌曲分类.因为我想看一下哪一张专辑我标记的喜欢的歌曲最多. 所以就用python写了下面的脚本. 使用的时候不要忘记把下面源码里 current_url里面的apc字串换成你的用户名, 并且登录豆瓣.

#!/usr/bin/python
from urlparse import urlparse, urljoin
import urllib, sgmllib
from HTMLParser import HTMLParser
import re, sys

class MyParser(HTMLParser):
    def __init__(self):
        HTMLParser.__init__(self)
        self.num = 0
        self.albums = {}
        
    def parse(self, url):
        req = urllib.urlopen(url)
        self.state       = 0
        self.raw_text = req.read()
        self.feed(self.raw_text)
        
    def handle_starttag(self, tag, attrs):
        #print "Encountered the beginning of a %s tag" % tag
        try:
            if self.state == 0:
                if  tag == "table" and dict(attrs)["class"] =="olts":
                    self.state = 1
                    self.row = 0

            elif self.state == 1:
                if  tag == "tr":
                    self.row += 1
                    self.state = 2

            elif self.state == 2:
                if  tag == "td":
                    self.state = 3

            elif self.state == 3:
                if tag == "a":
                    self.state = 4
                else:
                    self.state = 1

            elif self.state == 4:
                self.state = 1
        except KeyError:
            pass

    def handle_endtag(self, tag):
        #print "Encountered the end of a %s tag" % tag
        if self.state >= 1 and tag == "table":
            self.state = 0

    def handle_data(self, data):
        if self.state == 3:
            if self.row > 1:
                print "%d title"%(self.num), data[:-2],
                self.num += 1
                self.title = data[:-2]
        elif self.state == 4:
            if data.strip():
                print "album", data
                try:
                    self.albums[data].append(self.title)
                except KeyError:
                    self.albums[data] =  [self.title]
            
        
numsongs   = int(raw_input("How many songs do you have?"))
myparser    = MyParser()
base = 0
while base < numsongs: 
    current_url = "http://www.douban.com/people/apc/songs?start=%d"%(base)
    print current_url
    myparser.parse(current_url)
    base += 20
print "done, all together %d songs"%(myparser.num)
al = myparser.albums.keys()[:]
al.sort(cmp= lambda x,y: len(myparser.albums[x]) - len(myparser.albums[y]))
for a  in al:
    print a, "%d songs"%(len(myparser.albums[a]))

我输出的结果是这两张专辑我标记的最多:
Schindler's List 6 songs
Le Fabuleux destin d'Amélie Poulain 9 songs
辛德勒的名单的OST以及天使爱美丽的OST. 我果然是个电影控.