def craft_newsletter():
'''
Craft the newsletter. Returns JSON.
:return: the newsletter json
'''
a = Articles(API_KEY=os.environ["NEWSAPI_KEY"])
top_results = a.get_by_top(source="google-news")
breaking = requests.get("https://librenews.io/api").json()["latest"]
period = "AM"
greeting = "It's 5:30 ZULU time."
if datetime.datetime.now(tz=None).time() > datetime.time(12):
period = "PM"
greeting = "It's 17:30 ZULU time."
name = period + " - " + datetime.date.today().strftime("%A, %d %B %Y")
for story in top_results["articles"]:
story["source"] = urlparse(story["url"]).netloc
return {
"top_stories":
top_results["articles"][:3],
"breaking":
[story for story in breaking
if story["channel"] == "Breaking News"][:5],
"name":
name,
"greeting":
greeting
}
class NewsDataIngestion(DataIngestionInterface):
def __init__(self, api_key=keys.news['api_key']):
self.api_key = api_key
self.article = Articles(self.api_key)
self.source = Sources(self.api_key)
self.base_url = keys.news['base_everything_url']
self.logger = logutils.get_logger('News Data Ingestion')
def get_data(self,
query,
from_date=None,
to_date=None,
page_size=100,
sort_by='publishedAt',
language='en',
**kwargs):
key_value_params = {
'apiKey': self.api_key,
'q': query,
'from': from_date,
'to': to_date,
'sortBy': sort_by,
'pageSize': page_size,
'language': language
}
url = self.data_config(self.base_url,
query_separator='?',
key_value_params=key_value_params)
response = requests.get(url)
return response.json()
def data_config(self, base_url, **kwargs):
query_separator = None
key_value_params = None
join_sep = '&'
url = base_url
if 'query_separator' in kwargs.keys():
query_separator = kwargs['query_separator']
if 'key_value_params' in kwargs.keys():
key_value_params = kwargs['key_value_params']
if query_separator is not None:
url = base_url + str(query_separator)
if key_value_params is not None:
for key in key_value_params.keys():
if key_value_params[key] is not None:
url = url + str(key) + '=' + str(
key_value_params[key]) + join_sep
return url[:-1]
def store_data(self, data_list, connection_object):
connection_object.set_collection(constants.NEWS_COLLECTION_NAME)
for data_dict in data_list:
try:
if newsutils.check_duplicate_document(data_dict) is False:
connection_object.insert_document(data_dict)
except:
self.logger.error('Error while inserting data')
connection_object.close_connection()
def parse_data(self, news_response_json, product='default'):
article_list = list([])
if news_response_json['status'] == 'ok':
article_list = news_response_json['articles']
for article in article_list:
try:
article['source'] = article['source']['name']
article['product'] = product
article[
'human_date'] = newsutils.convert_string_timestamp_to_python_date(
article['publishedAt'])
except:
self.logger.error("error while parsing data")
return article_list
def get_articles(self, source_id, selection_type="popular"):
if selection_type == 'latest':
return self.article.get_by_latest(source_id)
elif selection_type == 'top':
return self.article.get_by_top(source_id)
else:
return self.article.get_by_popular(source_id)
def data_ingestion_pipeline(self, query, product='default'):
news_json = self.get_data(query)
self.logger.info("News data fetched for product %s " % product)
parsed_news_list = self.parse_data(news_json, product=product)
self.logger.info("News Data parsed for product %s" % product)
mongo_connector = dbutils.get_mongodb_connection()
self.store_data(parsed_news_list, mongo_connector)
self.logger.info("News data stored for product %s " % product)
class News:
def __init__(self, api_key=keys.news['api_key']):
self.api_key = api_key
self.article = Articles(self.api_key)
self.source = Sources(self.api_key)
self.base_url = keys.news['base_everything_url']
def get_data(self,
query,
from_date=None,
to_date=None,
page_size=100,
sort_by='publishedAt',
language='en',
**kwargs):
key_value_params = {
'apiKey': self.api_key,
'q': query,
'from': from_date,
'to': to_date,
'sortBy': sort_by,
'pageSize': page_size,
'language': language
}
url = self._data_config(self.base_url,
query_separator='?',
key_value_params=key_value_params)
response = requests.get(
url,
headers={
'User-Agent':
'Mozilla/5.0 (Macintosh; Intel Mac OS X 10_10_1) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/39.0.2171.95 Safari/537.36'
})
return self._parse_data(response.json())
@staticmethod
def _data_config(base_url, **kwargs):
query_separator = None
key_value_params = None
join_sep = '&'
url = base_url
if 'query_separator' in kwargs.keys():
query_separator = kwargs['query_separator']
if 'key_value_params' in kwargs.keys():
key_value_params = kwargs['key_value_params']
if query_separator is not None:
url = base_url + str(query_separator)
if key_value_params is not None:
for key in key_value_params.keys():
if key_value_params[key] is not None:
url = url + str(key) + '=' + str(
key_value_params[key]) + join_sep
return url[:-1]
@staticmethod
def _parse_data(news_response_json):
article_list = list([])
if news_response_json['status'] == 'ok':
article_list = news_response_json['articles']
for article in article_list:
try:
article['source'] = article['source']['name']
except:
pass
return article_list
def get_articles(self, source_id, selection_type="popular"):
if selection_type == 'latest':
return self.article.get_by_latest(source_id)
elif selection_type == 'top':
return self.article.get_by_top(source_id)
else:
return self.article.get_by_popular(source_id)