tweetset_loader.py 文件源码

def update_dataset_stats(dataset):
    search = Search(index=get_tweets_index_name(dataset.meta.id))
    search = search.query('term', dataset_id=dataset.meta.id)[0:0]
    search.aggs.metric('created_at_min', 'min', field='created_at')
    search.aggs.metric('created_at_max', 'max', field='created_at')
    search_response = search.execute()
    dataset.first_tweet_created_at = datetime.utcfromtimestamp(
        search_response.aggregations.created_at_min.value / 1000.0)
    dataset.last_tweet_created_at = datetime.utcfromtimestamp(
        search_response.aggregations.created_at_max.value / 1000.0)
    dataset.tweet_count = search_response.hits.total
    dataset.save()