Exemplo n.º 1
0
 def __init__(self, market):
     self.market = market
     resource = Resource(market)
     rs0 = resource.get_mongo_rs0()
     self.col_seed_table = rs0['crawl']['seedTable']
Exemplo n.º 2
0
            logger.info('invalid type {} '.format(tp))
            return
        metadata = {
            'type': tp,
            'handle': handle,
            'category': category
        }
        self.create(url, 'youtubev2_social_seed', 'youtubev2.{}.{}'.format(tp, handle), 'schedule', **metadata)


if __name__ == '__main__':
    logging.basicConfig(format='%(asctime)s, %(name)s %(levelname)s %(message)s',
                        datefmt="%Y-%m-%d %H:%M:%S",
                        level=logging.INFO)
    market = 'en-us'
    resource = Resource(market)
    rs0 = resource.get_mongo_rs0()
    col = rs0['crawl']['categories']
    lder = Loader(market)
    import json

    with open(sys.argv[1]) as f:
        for line in f:
            if line.startswith('#'):
                continue
            items = line.strip().split(',', 2)
            tp, handle, ch = items
            if not ch:
                ch = handle
            lder.inject_youtube(tp, handle, ch)