scrap twitter wall given the search page address or simulating a browser via mechanize

twitter_scrapper.py

'''
    Written by:
        Eren Golge - [email protected]
'''

import json
import pdb
import urllib2
import mechanize
import cookielib
import re
from BeautifulSoup import BeautifulSoup
import csv
import os

def set_mechanize():
    '''
        Set mechanize configs
    '''

    br = mechanize.Browser()

    # set cookies
    cookies = cookielib.LWPCookieJar()
    br.set_cookiejar(cookies)
    
    # browser settings (used to emulate a browser)
    br.set_handle_equiv(True)
    br.set_handle_redirect(True)
    br.set_handle_referer(True)
    br.set_handle_robots(False)
    br.set_debug_http(False)
    br.set_debug_responses(False)
    br.set_debug_redirects(False)
    br.set_handle_refresh(mechanize._http.HTTPRefreshProcessor(), max_time = 1)
    br.addheaders = [('User-agent', 'Mozilla/5.0 (X11; U; Linux i686; en-US; rv:1.9.0.1) Gecko/2008071615 Fedora/3.0.1-1.fc9 Firefox/3.0.1')]
    return br


def login_twitter(br):
    '''
    Login to twitter and return the wall content

    Inputs:
        br - mechanize instance configurated initially
    
    Outputs:
        wall - the raw html content of the home twitter wall
    '''
    name = 'erogol'
    password = 'ern88gol'
    
    theurl = 'http://twitter.com/login'
    br.open(theurl)
    #test_write(br)
    br.select_form(nr=0)
    br["session[username_or_email]"] = name
    br['password'] = password
    wall = br.submit().read()
    return wall

def load_page(url_addr):
    return urllib2.urlopen(url_addr).read()


def scrap_wall(wall):
    '''
    Given the twitter wall content, it scraps the images
    and the twitter texts

    Inputs:
        wall - raw html of the twitter wall

    Outputs:
        data_list - A list of dictionaries where each dict 
            includes meta information for each image twit
    '''

    data_list = []

    total_img_num = 0

    wall_soup = BeautifulSoup(wall)
    divs = wall_soup.findAll('div', {'class':'content'})
    for div in divs:
        #imgs = div.findAll('img', {'alt':'Embedded image permalink'})
        imgs = div.findAll('a', {'data-resolved-url-large': re.compile("jpg:large$") })
        if len(imgs) > 0:
            total_img_num = total_img_num + len(imgs)
            #img_links = div.findAll('a', {'class':'media media-thumbnail twitter-timeline-link media-forward is-preview'})
            img_links = div.findAll('a', {'data-resolved-url-large': re.compile("jpg:large$")})
            
            if len(img_links) == 0:
                print 'ERROR: img links are changed on twitter wall!!!'
            img_link = img_links[0].get('data-resolved-url-large')
            
            twit_texts = div.findAll('p', {'class' : 'js-tweet-text tweet-text'})
            if len(twit_texts) < 1:
                print 'ERROR: twitter text container is changed !!!'
            twit_text = twit_texts[0].getText()
            print twit_text

            details = div.findAll('a', {'class' : 'permalink-link js-permalink js-nav'})
            if len(details) < 1 :
                print 'ERROR: "details" container is changed !!!'
            twit_url = details[0]['href'] 

            data_list.append({'imgs' : img_link.encode('utf8'), 'twits' : twit_text.encode('utf8'), 'links' : twit_url.encode('utf8')})

    print 'Total number of images --- ', total_img_num
    print data_list
    return data_list

def write_csv(data_list, file_name, folder_name = ''):
    '''
        Save the given list of dicts
    '''
    if  len(data_list) == 0:
        print "NO DATA TO WRITE!!!"
        return 
    full_path = folder_name + '/' + file_name
    if len(folder_name) > 0:
        if not os.path.exists(folder_name):
            print 'New foder created as ' + folder_name
            os.makedirs(folder_name)

    keys = data_list[0].keys()
    dict_writer = csv.DictWriter(open(full_path,'wb'), keys)
    dict_writer.writer.writerow(keys)
    dict_writer.writerows(data_list)
    print 'DATA has been saved to ' + full_path

# auxiliary functions 
def test_read(file_name = 'test.html'):
    f =  open(file_name, 'r')
    page = f.read()
    return page
    
def test_write(result, file_name = 'test.html'):
    f = open(file_name,'wb')
    f.write(result)
    f.close()

if __name__ == '__main__':
    #test_write(load_page('https://twitter.com/search?q=%22Fenerbah%C3%A7eSu%C3%A7suzdur%20DavaYenidenG%C3%B6r%C3%BCls%C3%BCn%22&src=tren'),'test2.html')
    write_csv(scrap_wall( test_read('tests.html') ), 'test.csv', 'SAVED_DATA')