NLP_program.py

#Natural Language Processing using wit.ai library
from wit import Wit
from gnewsclient import gnewsclient

#REFER https://wit.ai/ generate access-token
access_token="ACCESS-TOKEN"

client=Wit(access_token=access_token)
def wit_response(message_text):
	resp=client.message(message_text)

	categories={'newstype':None,'location':None}
	entities=list(resp['entities'])

	for entity in entities:
		categories[entity]=resp['entities'][entity][0]['value']
	return categories

def get_news_elements(categories):
	news_client=gnewsclient()
	news_client.query=''
	if categories['newstype'] != None:
		news_client.query+=categories['newstype']+' '
	if categories['location'] != None:
		news_client.query+=categories['location']

	news_items=news_client.get_news()
	elements=[]
	for item in news_items:
		element={
			'title':item['title'],
			'buttons':[{
				'type':'web_url',
				'title':"Read more",
				'url':item['link']
			}],
			'image_url':item['img']
		}
		elements.append(element)

	return elements

#Type any news in 'wit_response' function
categories=wit_response('ndtv news')
result=get_news_elements(categories)
print(result)