BING APIサーチを使って画像をダウンロードします

image_downloader.py

#!/usr/bin/env python
# coding: utf-8

import asyncio
import aiohttp
import requests
import json
import os
import click
from hashlib import md5

ENDPOINT = 'https://api.datamarket.azure.com/Bing/Search/v1/Image'

SAVE_PATH = os.path.join(os.environ['HOME'], 'image')

# 並列数
semaphore = asyncio.Semaphore(4)

CHUNK_SIZE = 1024

@asyncio.coroutine
def fetch_and_save(url):
    name, ext = os.path.splitext(url)
    m = md5()
    m.update(url.encode('utf-8'))
    file_name = os.path.join(SAVE_PATH, m.hexdigest() + ext)

    r = yield from aiohttp.get(url)
    if r.status != 200:
        return
    with open(file_name, 'wb') as f:
        while True:
            chunk = yield from r.content.read(CHUNK_SIZE)
            if not chunk:
                break
            f.write(chunk)

@asyncio.coroutine
def get(url):
    with (yield from semaphore):
        yield from fetch_and_save(url)

CONTEXT_SETTINGS = dict(help_option_names=['-h', '--help'])

@click.command(context_settings=CONTEXT_SETTINGS)
@click.option('-a', '--adult', flag_value=True, default=False, help='Adult filter OFF')
@click.argument('word')
def main(adult, word):
    if not 'BING_ACCESS_KEY' in os.environ:
        print('BINGのアクセスキーを設定してください')
        print('  ex) export BING_ACCESS_KEY="Your access key"')
        return
    access_key = os.environ['BING_ACCESS_KEY']

    adult_filter = 'Moderate'
    if adult:
        adult_filter = 'Off'

    params = {'Query': "'{}'".format(word),
              'Market': "'ja-JP'",
              'Adult': "'{}'".format(adult_filter),
              '$format': 'json'}

    r = requests.get(ENDPOINT, params=params, auth=('', access_key))
    r.raise_for_status()

    tasks = [get(x['MediaUrl']) for x in json.loads(r.text)['d']['results']]

    try:
        loop = asyncio.get_event_loop()
        loop.run_until_complete(asyncio.wait(tasks))
    finally:
        loop.close()

if __name__ == '__main__':
    main()

requirements.txt

aiohttp
requests
click

• Python3.4
• Python3.3 # pip install asyncio

実行方法

$ pip install -r requirements.txt
$ export BING_ACCESS_KEY="Your access key"'
$ ./image_downloader.py -h

BINGのアクセスキー

https://datamarket.azure.com/dataset/bing/search

ログインして、マイアカウントのプライマリ アカウント キーから取得する