fakenews.py

import machine
import article_topics
import comparison_topic
import sentiment_analysis
import additional_points
from article_scraper import article_scrape
import re

#testurl = "http://www.bbc.co.uk/news/world-europe-41785292/"

testurl = "http://www.bbc.co.uk/news/world-europe-41785292"

def blackList(url):
    socialNetworks = ["facebook.com", "twitter.com", "tumblr.com", "instagram.com", "pinterest.com", "4chan.org","reddit.com", "myspace.com","linkedin.com"]
    for network in socialNetworks:
        if network in url:
            return True
    return False

def whiteList(url):
    trusted = ["independent.co.uk", "thetimes.co.uk", "nytimes.com", "washingtonpost.com", "edition.cnn.com"]
    for address in trusted:
        if address in url:
            return True
    return False

def invalid(url):
    if url.startswith("https://"):
        return False
    elif url.startswith("http://"):
        return False
    else:
        return True



def fakeNews(url):

    score = 0

    if machine.SatireCheck(url):
        return -0.05 # If satire, it's fake news

    # get the srs words
    titlekeywords = machine.FindTitle(url)
    print(titlekeywords)


    # TRUMP CHECK START
    trump = machine.TrumpCheck(titlekeywords)

    if trump >= 2:
        return "Trump" # if it's Trump, it's almost guaranteed fake news
    else:
        pass
    # TRUMP CHECK OVER

    # initiate the class
    topics = article_topics.OnStart()

    namestocheck = article_topics.OnStart.FindNames()

    

    # get article you are evaluating text
    try:
        mainArticleContent = article_scrape(url)
    except:
        return "oops"

    with open("testeroo.txt", 'w') as file:
        file.write(mainArticleContent)
    
    with open("testeroo.txt", 'r') as file:
        mainArticleContent = file.readlines()

    # get topics of the main article you are evaluating
    mainArticleTopics = topics.FindTopics(mainArticleContent)
    for elem in mainArticleTopics:
        elem = re.sub(r'\W+', '', elem)

    # gets a list of lists.
    # each list includes:
    # 0. matched words
    # 1. the title/description
    # 2. article url
    # 3. topics
    additional_fake_others = []

    articlecompare = machine.NewsCheck(titlekeywords)
    topicothers = []
    titles = [] 
    articletxt = ""
    for key in articlecompare:
        try:
            articletxt = article_scrape(articlecompare[key])
        except:
            return "oops"
        with open("test2.txt", 'w') as file:
            file.write(articletxt)
        with open("test2.txt", 'r') as file:
            articletxt = file.readlines()
        topicothers.append(topics.FindTopics(articletxt))
        titles.append(key)
        additional_fake_others.append(additional_points.MoreCheck.checkMore(articletxt))

    scoring = comparison_topic.Comparison
    #print(titles)
    negativity = sentiment_analysis.Model

    additional_fake = additional_points.MoreCheck.checkMore(mainArticleContent)

    res = scoring.compare(mainArticleTopics, topicothers, titlekeywords, titles, additional_fake, additional_fake_others)
     
    #TODO: get rokas or karolis to remove this properly bc it just ain't workin'
    
    #PersonBlacklist = [("Alex", "Jones"), ("Tom", "Cruise"), ("Adolf", "Hitler"), ("Findlay", "Smith")]
    #for person in PersonBlacklist:
    #    if (person[0] == namestocheck[0]) and (person[1] == namestocheck[1]):
    #        score += 20.0
    print(score)
    print (score+res)
    return score+res 

def whySoNegative():

    with open("testeroo.txt", 'r') as file:
        mainArticleContent = file.readlines()

    negativity = sentiment_analysis.Model


    return negativity.startModel(mainArticleContent)




if __name__ == "__main__":
    fakeNews(testurl)