いかにしておっぱい画像をダウンロードするか〜2012 for Python

op.py

# -*- coding: utf-8 -*-

import os.path, re, hashlib, urllib, json

appid = ''
uri = 'http://api.bing.net/json.aspx'
dir = './data'

page_count = 0
download_count = 0

while True:
    offset = page_count * 50
    query = urllib.urlencode({
        'Appid': appid,
        'Version': '2.2',
        'Market': 'ja-JP',
        'Sources': 'Image',
        'Image.Count': 50,
        'Image.Offset': offset,
        'Adult': 'off',
        'Query': 'おっぱい'
    })

    res = urllib.urlopen(uri + '?' + query).read()
    ref = json.loads(res)
    try:
        ref['SearchResponse']['Image']['Results']
    except:
        break

    for entry in ref['SearchResponse']['Image']['Results']:
        if re.compile('\.jpg$').match(entry['MediaUrl']):
            continue
        download_count += 1
        filename = hashlib.sha224(entry['MediaUrl']).hexdigest() + '.jpg'
        filepath = dir + '/' + filename
        if os.path.exists(filepath):
            continue
        print str(download_count) + ": Download... " + entry['MediaUrl']
        try:
            res = urllib.urlopen(entry['MediaUrl'])
        except:
            continue
        if re.compile('^image').match(res.info()['Content-Type']):
            f = open(filepath, 'w')
            f.write(res.read())
            f.close()
    page_count += 1