tests/scraper_test.py

import re
import unittest

import pandas as pd
from datetime import datetime
import sys
from pathlib import Path
sys.path.insert(0, str(Path(__file__).parents[1]) + "/twitterscraper")

from TwitterScraper import TwitterScraper # Detta är inget problem, den hittar filen

class MyTestCase(unittest.TestCase):

    @classmethod
    def setUpClass(cls):
        print('Super set up')
        num_tweets = 40
        from_date = "2022-05-01"
        to_date = "2022-07-31"
        user = 'jimmieakesson'
        user_list = ['jimmieakesson', 'BuschEbba', 'annieloof', 'JohanPehrson', 'bolund', 'martastenevi', 'SwedishPM', 'dadgostarnooshi']
        sc1 = TwitterScraper(from_date=from_date, to_date=to_date, num_tweets=num_tweets)
        sc2 = TwitterScraper(from_date=from_date, to_date=to_date, num_tweets=num_tweets)
        sc3 = TwitterScraper(from_date=from_date, to_date=to_date, num_tweets=num_tweets)
        search_string = 'miljö'
        cls._df_uni = sc1.scrape_by_user(user)
        cls._df_poly = sc2.scrape_by_several_users(user_list)
        cls._df_by_string = sc3.scrape_by_string(search_string)

        nbr_of_cols = 9

    def setUp(self):
        print('set up')
        self.num_tweets = 40
        self.from_date = "2022-05-01"
        self.to_date = "2022-07-31"
        self.user = 'jimmieakesson'
        self.user_list = ['jimmieakesson', 'BuschEbba', 'annieloof', 'JohanPehrson', 'bolund', 'martastenevi', 'SwedishPM',
                     'dadgostarnooshi']
        self.search_string  = 'miljö'
        # self.sc = TwitterScraper(from_date=from_date, to_date=to_date, num_tweets=num_tweets)
        nbr_of_cols = 9

    # Checks that the returned datatype is pandas DataFrame
    def test_correct_type(self):
        print('Checking type...')
        self.assertEqual(type(self._df_uni), type(pd.DataFrame()))
        self.assertEqual(type(self._df_poly), type(pd.DataFrame()))
        self.assertEqual(type(self._df_by_string), type(pd.DataFrame()))

    # Checks that we get the correct number of tweets
    # OBS, FOR NOW IT ONLY CHECKS THAT WE DON'T OVER SAMPLE
    def test_correct_nbr_tweets(self):
        print('Checking number of tweets...')
        self.assertTrue(self._df_uni.shape[0] <= self.num_tweets)
        self.assertTrue(self._df_poly.shape[0] <= self.num_tweets)
        self.assertTrue(self._df_by_string.shape[0] <= self.num_tweets)

    # Checks that all dates are between the start date and the end date
    def test_dates(self):
        print('Checking dates...')
        d_start = datetime.fromisoformat(self.from_date)
        d_end = datetime.fromisoformat(self.to_date)
        correct_date_uni = True
        correct_date_poly = True
        correct_date_by_string = True
        for date in self._df_uni.date:
            d = datetime.fromisoformat(date)
            if not (d >= d_start and d <= d_end):
                correct_date_uni = False
                break
        for date in self._df_poly.date:
            d = datetime.fromisoformat(date)
            if not (d >= d_start and d <= d_end):
                correct_date_poly = False
                break
        for date in self._df_by_string.date:
            d = datetime.fromisoformat(date)
            if not (d >= d_start and d <= d_end):
                correct_date_by_string = False
                break

        self.assertTrue(correct_date_uni)
        self.assertTrue(correct_date_poly)
        self.assertTrue(correct_date_by_string)

    # Checks that all tweets are from the correct user
    def test_user(self):
        print('Checking user...')
        same_user = True
        for username in self._df_uni.username:
            if not username == self.user:
                same_user = False
                break

        self.assertTrue(same_user)

    # Checks that all user_ids are correct
    def test_user_id(self):
        print('Checking user ids...')
        same_user = True
        first_id = self._df_uni['user_id'][0]
        for user_id in self._df_uni.user_id:
            if not user_id == first_id:
                same_user = False
                break
        self.assertTrue(same_user)

    # Checks if there are tweets that have been sampled several times
    def test_no_doubles(self):
        print('Checking doubles...')
        id_set_uni = set(self._df_uni.id)
        id_set_poly = set(self._df_poly.id)
        id_set_by_string = set(self._df_by_string.id)
        self.assertTrue(len(id_set_uni) == self._df_uni.shape[0])
        self.assertTrue(len(id_set_poly) == self._df_poly.shape[0])
        self.assertTrue(len(id_set_by_string) == self._df_by_string.shape[0])

    # Checks that we have no None entries
    def test_none(self):
        print('Checking Nones...')
        self.assertFalse(any(b == True for b in self._df_uni.isnull()))
        self.assertFalse(any(b == True for b in self._df_poly.isnull()))
        self.assertFalse(any(b == True for b in self._df_by_string.isnull()))

    def test_no_url_tweets(self):
        print('Checking url tweets...')
        only_url_uni = False
        only_url_poly = False
        only_url_by_string = False
        for tweet in self._df_uni.tweet:
            if len(tweet.split()) == 1 and tweet.split()[0].startswith('https'):
                print(tweet.split())
                only_url_uni = True
                break
        for tweet in self._df_poly.tweet:
            if len(tweet.split()) == 1 and tweet.split()[0].startswith('https'):
                print(tweet.split())
                only_url_poly = True
                break
        for tweet in self._df_by_string.tweet:
            if len(tweet.split()) == 1 and tweet.split()[0].startswith('https'):
                print(tweet.split())
                only_url_by_string = True
                break
        self.assertFalse(only_url_uni)
        self.assertFalse(only_url_poly)
        self.assertFalse(only_url_by_string)

    def test_many_users(self):
        correct_users = True
        for user in self._df_uni.username:
            if user not in self.user_list:
                correct_users = False
                break
        self.assertTrue(correct_users)

    def test_many_user_ids(self):
        correct_ids = True
        user_dict = {}
        for i in range(self._df_uni.shape[0]):
            if self._df_uni['username'][i] not in user_dict:
                user_dict[self._df_uni['username'][i]] = self._df_uni['user_id'][i]
            if self._df_uni['user_id'][i] != user_dict[self._df_uni['username'][i]]:
                correct_ids = False
                break
        self.assertTrue(correct_ids)

    def test_string_search(self):
        correct_search = True
        search = re.sub('ö', 'ø', self.search_string)
        search = re.sub('ä', 'æ', search)
        search_strings = [self.search_string, search]
        for tweet in self._df_by_string.tweet:
            if all(search not in tweet.lower() for search in search_strings):
                correct_search = False
        self.assertTrue(correct_search)



if __name__ == '__main__':
    unittest.main()