megaporn.py

# -*- coding: utf-8 -*-
# expressing my strong respect for mopemope

import re
import urllib
from urlparse import urlunparse, urlparse
import gzip
import eventlet
from eventlet.green import urllib2
import cookielib
from pyquery import PyQuery as pq
from functools import partial
from werkzeug import secure_filename
import os.path

pool = eventlet.GreenPool()

search_url = ur"http://www.megaporn.com/video/"
method = "GET"

query = [u"creampie"]
charset = "utf-8"

# flvがあるページへのURL
# http://cdn.empflix.com/empflv/xxxxxxxxxx
target_url_ptn = re.compile("""
\s*flashvars\.embed\ =\ "
(?P<embed>\S+)
";
""", re.VERBOSE)

file_title_ptn = re.compile("""
\s*flashvars\.title\ =\ "
(?P<title>\S+)
";
""", re.VERBOSE)

flv_url_ptn = re.compile("""
<!--<file>
(?P<url>http://\S+\.flv)
</file>-->
""", re.VERBOSE)


download_dir = "./download/megaporn"

def build_opener():
    jar = cookielib.CookieJar()
    opener = urllib2.build_opener(urllib2.HTTPCookieProcessor(jar),
                                  urllib2.HTTPRedirectHandler())
    opener.addheaders = [("User-Agent", "Mozilla/5.0 (compatible; python)"),
                         ("Connection", "keep-alive")]
    return opener


def get_search_result(opener, query):
    view_urls = []
    for q in query:
        form_dict = {u'c': u'videos',
                     u'setlang': u'jp',
                     u's': unicode(q)}
        params = urllib.urlencode(form_dict)
        print params
        conn = None
        if method.upper() == "GET":
            conn = opener.open(search_url + u'?' + params)
        elif method.upper() == "POST":
            conn = opener.open(search_url, params)
        else:
            raise ValueError, method
        
        if conn:
            page = conn.read().decode(charset)
            d = pq(page)
            for a in d(".vid_link_layer"):
                # linkは "?v=xxxxxxx" の相対パスで取れる
                view_url = pq(a).attr.href
                print view_url
                view_urls.append(search_url + view_url)

    return view_urls


def _get_download_url(opener, view_url):
    conn = opener.open(view_url)
    data = conn.read()

    # ファイル名
    tmatch = file_title_ptn.search(data)
    if tmatch:
        td = tmatch.groupdict()
        filename = secure_filename(urllib.unquote(td['title'])) + ".flv"

    # 次のページに行ってflvの直リンクを取得
    ematch = target_url_ptn.search(data)
    if ematch:
        ed = ematch.groupdict()
        extra_page_url = urllib.unquote(ed['embed'])
        download_url = extra_page_url

    return download_url, filename


if __name__ == "__main__":
    opener = build_opener()
    view_urls = get_search_result(opener, query)
    
    print _get_download_url(opener, view_urls[0])