screen.yahoo.com video grabber

yahoo.py

# Get a video url from screen.yahoo.com
# python3 yahoo.py 'http://screen.yahoo.com/raven-refused-sing-010647927.html' | xargs vlc
import sys
import contextlib
import json
import random
import re
import urllib.parse
import urllib.request

API_URL = "video.query.yahoo.com"
API_PATH = "/v1/public/yql"
QUERY = (
    'SELECT * FROM yahoo.media.video.streams WHERE id="{id}" AND format="{format}"' +
    ' AND protocol="{protocol}" AND plrs="{player_guid}" AND offnetwork="{onet}"' +
    ' AND site="{site}" AND lang="{lang}" AND region="{region}" AND override="";')

def create_guid():
    chars = "ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789._"
    s = ""
    while len(s) < 22:
        s += random.choice(chars)
    return s

def fatal(s):
    print(s, file=sys.stderr)
    sys.exit(1)

def urlopen(url):
    return contextlib.closing(urllib.request.urlopen(url))

def main():
    page_url = sys.argv[1]
    if not page_url.startswith("http://screen.yahoo.com/"):
        fatal("Doesn't look like a Yahoo! Screen url")

    with urlopen(page_url) as f:
        page_head = f.read(4096)

    m = re.search(rb'<meta property="og:video" content="([^"]*)"/?>', page_head)
    if m is None:
        fatal("Couldn't find video player url")

    player_url = m.group(1).decode('utf-8', errors='ignore').replace("&amp;", "&")
    #print(player_url)

    q = urllib.parse.parse_qs(
            urllib.parse.urlparse(player_url).query)

    query_params = {
        'id': q['uuid'][0],
        'format': "mp4, flv",
        'protocol': "http", # rtmp,http
        'player_guid': create_guid(),
        'lang': "en-US",
        'region': q['region'][0],
        'onet': q['onet'][0],
        'site': q['site'][0],
    }
    #print(query_params)

    info_url = "http://{API_URL}{API_PATH}?q={query}&env=prod&format={fmt}".format(
        API_URL=API_URL,
        API_PATH=API_PATH,
        query=urllib.parse.quote(QUERY.format(**query_params)),
        fmt="json", # xml
    )
    #print(info_url)

    with urlopen(info_url) as f:
        info = json.loads(f.read().decode('utf-8'))

    #print(json.dumps(info))

    streams = info['query']['results']['mediaObj'][0]['streams']
    stream = streams[0]
    video_url = stream['host'] + stream['path']

    print(video_url)

if __name__ == '__main__':
    main()