def web_scraper(input):
key_words = input.split(',')
keys = []
#dates = []
time = str(datetime.datetime.now().date())
for key_word in key_words:
try:
while (key_word[0] == ' '):
key_word = key_word[1:]
while (key_word[len(key_word) - 1] == ' '):
key_word = key_word[:-1]
except IndexError:
continue
if (len(key_word) == 0):
continue
inputs = key_word.split(' ')
key = ''
for i in inputs:
key += i
key += '%20'
key = key[:-3].lower()
key = re.sub(r'[ ]+', ' ', key)
keys.append(key)
#dates.append(check.check_date(data_base, kkeys[i]))
for i in range(0, len(keys)):
previous_len = 0
try:
for lines in open("./documents/web/Articles.bank",
'r',
encoding='utf-8'):
previous_len += 1
except FileNotFoundError:
pass
stored_keys = check.load_keywords_info()
data_base = check.load_url_info()
data_print = {}
kkey = fmt.file_name(keys[i], '_')
for kk in stored_keys:
if abs(len(kkey) - len(kk)) < 2 and check.MinEditDist(kkey,
kk) == 1:
kkey = kk
if kkey in stored_keys:
date = int(re.sub(r'-', '', stored_keys[kkey]))
else:
date = 0
stored_keys[kkey] = time
if kkey not in data_base:
data_base[kkey] = []
cnn.cnn(data_base, data_print, keys[i], date, previous_len)
foxnews.foxnews(data_base, data_print, keys[i], date, previous_len)
cbs.cbs(data_base, data_print, keys[i], date, previous_len)
politico.politico(data_base, data_print, keys[i], date, previous_len)
#washington_post.washington_post(data_base,data_print,key,date)
if len(data_print) > 1:
check.save_keywords_info(stored_keys)
check.save_url_info(data_base)
output_file.output(data_print, data_base, previous_len)
print("Update date: " + time)
def washington_post(data_base,data_print,key,date):
hp.login("*****@*****.**", "qazxdr12")
kkey = fmt.file_name(key,'_')
kkkey = fmt.file_name(key,'+')
print("----- "+"washington_post."+kkey+" -----")
print("Start loading Urls...")
#case for exact keyword search
url1='https://sitesearchapp.washingtonpost.com/sitesearch-api/v2/search.json?count=20&datefilter=displaydatetime:%5B*+TO+NOW%2FDAY%2B1DAY%5D&facets.fields=%7B!ex%3Dinclude%7Dcontenttype,%7B!ex%3Dinclude%7Dname&filter=%7B!tag%3Dinclude%7Dcontenttype:("Article"+OR+(contenttype:"Blog"+AND+name:("Opinions")))&highlight.fields=headline,body&highlight.on=true&highlight.snippets=1&query="'
url2='"&sort=&startat='
url3='&callback=angular.callbacks._0'
baseurl = url1+kkkey+url2+'0'+url3
try:
page = hp.getHtml(baseurl)
except urllib.error.URLError:
pass
article_number = regex.get_data('"total"\S(\S+?),"documents',page)[0]
if article_number == 0:
url1='https://sitesearchapp.washingtonpost.com/sitesearch-api/v2/search.json?count=20&datefilter=displaydatetime:%5B*+TO+NOW%2FDAY%2B1DAY%5D&facets.fields=%7B!ex%3Dinclude%7Dcontenttype,%7B!ex%3Dinclude%7Dname&filter=%7B!tag%3Dinclude%7Dcontenttype:("Article"+OR+(contenttype:"Blog"+AND+name:("Opinions")))&highlight.fields=headline,body&highlight.on=true&highlight.snippets=1&query='
url2='&sort=&startat='
url3='&callback=angular.callbacks._0'
baseurl = url1+kkkey+url2+'0'+url3
try:
page = hp.getHtml(baseurl)
except urllib.error.URLError:
print("Washington Post website is not correct, please check the code!")
return -1
article_number = regex.get_data('"total"\S(\S+?),"documents',page)[0]
if article_number == 0:
print("No Washington Post article was found by this key word")
return -1
#get all urls
count = 0
index = 0
urls = []
page_total = int(article_number) / 20 + 1
while(count < page_total):
currenturl = url1+key+url2+str(index)+url3
try:
page = hp.getHtml(currenturl)
except urllib.error.URLError:
continue
url = regex.get_data('"contenturl"\S"(https:\/\/www.washingtonpost.com\/opinions/\S+?)"\S"',page)
if date != 0:
a_num = check.check_last_update(url,date)
if a_num != -1:
urls += url[:-(len(url)-a_num )]
break
urls += url
index += 20
count += 1
print(str(len(urls))+" Urls loaded...")
print("There are "+str(len(data_base)+len(data_print))+" loaded file...",)
print("Now starting updating...",)
count = 0
#count2 = 0
for url in urls:
if url in data_base and kkey in data_base[url]:
#if check.update_key(data_base, url, kkey):
# count2 += 1
continue
try:
html = hp.getHtml(url)
except urllib.error.URLError:
continue
title = regex.get_data('"headline":"(.*?)",',html)
#<meta content="Julian Zelizer, CNN Political Analyst" name="author">
author = regex.get_data('this.props.author="(.*?)";',html)
#<meta content="2018-02-17T00:19:47Z" name="pubdate">
date = regex.get_data('"datePublished":"(\S+?)T',html)
text2 = regex.get_data('<article.*?>(.*?)<\/p>\s<\/article>',html)
if text2 != []:
text = regex.get_data('<p.*?>(.*?)<\/p>',text2[0])
else:
text = regex.get_data('<p.*?>(.*?)<\/p>',html)
if text == [] or title == []:
continue
data_print[url] = defaultdict(str)
# line 1
data_print[url]['ID'] = fmt.formatted_id(len(data_base)+len(data_print)-1)
data_print[url]['key'] = fmt.formatted_key(kkey)
# line 2
data_print[url]['title'] = fmt.formatted_title(title)
# line 3
data_print[url]['source'] = fmt.formatted_source("Washington Post")
# line 4
data_print[url]['url'] = fmt.formatted_url(url)
# line 5
data_print[url]['date'] = fmt.formatted_date(date)
# line 6
data_print[url]['author'] = fmt.formatted_author(author,';')
# line 7
data_print[url]['content1'] = fmt.formatted_content_with_symbol(text)
# line 8
data_print[url]['content2'] = fmt.formatted_content(text)
count += 1
print("Updated "+str(count)+" articles...")
#if count2 > 0:
# print("Updated "+str(count2)+" keys...")
print("There are "+str(len(data_base)+len(data_print))+" articles...")
def cbs(data_base, data_print, key, date_, previous_len):
kkey = fmt.file_name(key, '_')
kkkey = fmt.file_name(key, '+')
print("----- " + "cbs." + kkey + " -----")
print("Start loading Urls...")
kkkey = re.sub(r'/', '%2F', kkkey)
#kkkey = re.sub(r'+', '%2B', kkkey)
kkkey = re.sub(r'%', '%25', kkkey)
#case for exact keyword search
url1 = 'https://www.cbsnews.com/search/?q='
url2 = '&o=1&p='
url3 = '&t=opinion'
baseurl = url1 + kkkey + url2 + '1' + url3
article_number = '0'
try:
page = hp.getHtml(baseurl)
except urllib.error.URLError:
print("CBS website is not correct, please check the code!")
return -1
try:
article_number = regex.get_text('<h1\sclass="result-title">(\d+)\s',
page)[0]
except IndexError:
article_number = '0'
if int(article_number) == 0:
print("No CBS article was found by this key word")
return -1
#get all urls
count = 0
index = 0
page_num = 1
urls = defaultdict(str)
page_total = int(int(article_number) / 10 + 1)
reach_updated = False
print("There are " + article_number + " articles...")
print("Start loading and Updating...")
while (count < page_total):
currenturl = url1 + key + url2 + str(page_num) + url3
try:
page = hp.getHtml(currenturl)
except urllib.error.URLError:
continue
url = regex.get_text('<a\shref="(\S+?)"><h3\sclass="title"', page)
date = regex.get_text(
'<span\sclass="date">(\S+?\s\d+,\s\d+?)\s\S+\s\S+?\s\S+<\/span>',
page)
for cnt in range(0, len(date)):
date[cnt] = fmt.convert_date(date[cnt])
if date_ > date[cnt]:
reach_updated = True
break
for i in range(0, cnt + 1):
try:
urls['https://www.cbsnews.com' +
url[i]] = str(date[i])[0:4] + '-' + str(
date[i])[4:6] + '-' + str(date[i])[6:8]
except IndexError:
break
if reach_updated:
break
index += 10
page_num += 1
count += 1
print(str(len(urls)) + " URLs loaded...")
print("Updating database...")
for url in urls:
if url in data_base[kkey]:
continue
try:
html = hp.getHtml(url)
except urllib.error.URLError:
continue
title = regex.get_data('<title>([^<]+?)\s-\s[^<]+?<\/title>', html)
if title == 'Noun':
title = regex.get_data('<title>([^<]+?)<\/title>', html)
author = regex.get_data(
'"author":{".type":"[^"]+?","name":"([^"]+?)"}', html)
text1 = []
text1.append(
regex.get_data('<div\sdata-page=[^>]+?><[^>]*?>\n?([^\n]+?)<.?p>',
html))
text2 = regex.get_text('<p>([^\n]+?)<\/p>', html)
text = text1 + text2
if text == [] or title == "Noun":
continue
data_base[kkey].append(url)
data_print[url] = defaultdict(str)
# line 1
data_print[url]['ID'] = fmt.formatted_id(
len(data_base[kkey]) - 1 + previous_len)
# line 2
data_print[url]['key'] = fmt.formatted_key(kkey)
# line 3
data_print[url]['title'] = fmt.formatted_title(title)
# line 4
data_print[url]['source'] = fmt.formatted_source("CBS")
# line 5
data_print[url]['url'] = fmt.formatted_url(url)
# line 6
data_print[url]['date'] = fmt.formatted_date(urls[url])
# line 7
if len(author) != 0:
aa = author.split(',')
if len(aa) > 1:
author = ','.join(aa[:-1])
elif len(aa) == 0:
author = 'Noun Noun'
else:
author = 'Noun Noun'
data_print[url]['author'] = fmt.formatted_author(author, ',')
# line 8
data_print[url]['content'] = fmt.formatted_content(text)
# line 9
#data_print[url[i]]['content2'] = fmt.formatted_content(text)
print('■', end='', flush=True)
print("\nThere are " + str(len(data_print) + previous_len) +
" articles...")
print("Updated " + str(len(data_print)) + " articles...")
def cnn(data_base, data_print, key, date_, previous_len):
kkey = fmt.file_name(key, '_')
kkkey = fmt.file_name(key, '+')
print("----- " + "cnn." + kkey + " -----")
#case for exact keyword search
url1 = 'https://search.api.cnn.io/content?size=10&q=%22'
url2 = '%22&category=opinion'
baseurl = url1 + key + url2
article_number = '0'
try:
page = hp.getHtml(baseurl)
except urllib.error.URLError:
print("CNN website is not correct, please update the scraper!")
return -1
article_number = regex.get_text('"meta":{\S+?"of":(\d+?),"maxScore',
page)[0]
if int(article_number) == 0:
print("No CNN article was found by this key word")
return -1
#get all urls
count = 0
index = 0
page_num = 1
urls = defaultdict(str)
page_total = int(int(article_number) / 10 + 1)
reach_updated = False
print("There are " + article_number + " articles...")
print("Start loading URLs...")
while (count < page_total):
currenturl = url1 + key + url2 + '&from=' + str(
index) + '&page=' + str(page_num)
try:
page = hp.getHtml(currenturl)
except urllib.error.URLError:
continue
url = regex.get_text('"url":"([^,]+?.html)"\S"', page)
#title = regex.get_text('"headline":"([^{]*?)"',page)
#author = regex.get_text('"byLine":(.*?),',page)
for i in range(0, len(url)):
try:
d = regex.get_data('\/(\d+?\/\d+?\/\d+?)\/', url[i])
except IndexError:
break
d_int = int(re.sub(r'/', '', d))
if date_ > d_int:
reach_updated = True
break
urls[url[i]] = re.sub(r'/', '-', d)
if reach_updated:
break
index += 10
page_num += 1
count += 1
print(str(len(urls)) + " URLs loaded...")
print("Updating database...")
for url in urls:
if url in data_base[kkey]:
continue
try:
html = hp.getHtml(url)
except urllib.error.URLError:
continue
title = regex.get_data('<title>([^<]+?)\s-\s\w+?<\/title>', html)
if title == 'Noun':
title = regex.get_data('<title>([^<]+?)<\/title>', html)
author = regex.get_data('<meta\scontent\S"([^"]+?)"\sname="author">',
html)
text2 = []
text2.append(
regex.get_data(
'<cite\sclass="el-editorial-source">\s\S\S\S\S\S</cite>([^=]*?)<\/p><\/div>',
html))
text1 = regex.get_text(
'<div\sclass="zn-body__paragraph\s*?\w*?">([^=]+?)</div>?', html)
text = text2 + text1
if text == [] or title == "Noun":
continue
data_base[kkey].append(url)
data_print[url] = defaultdict(str)
# line 1
data_print[url]['ID'] = fmt.formatted_id(
len(data_base[kkey]) - 1 + previous_len)
# line 2
data_print[url]['key'] = fmt.formatted_key(kkey)
# line 3
data_print[url]['title'] = fmt.formatted_title(title)
# line 4
data_print[url]['source'] = fmt.formatted_source("CNN")
# line 5
data_print[url]['url'] = fmt.formatted_url(url)
# line 6
data_print[url]['date'] = fmt.formatted_date(urls[url])
# line 7
if len(author) > 5:
if author[0:3] == "By ":
author = author[3:]
aa = author.split(',')
if len(aa) > 1:
author = ','.join(aa[:-1])
else:
author = 'Noun Noun'
data_print[url]['author'] = fmt.formatted_author(author, ',')
# line 8
data_print[url]['content'] = fmt.formatted_content(text)
# line 9
#data_print[url[i]]['content2'] = fmt.formatted_content(text)
print('■', end='', flush=True)
print("\nThere are " + str(len(data_print) + previous_len) +
" articles...")
print("Updated " + str(len(data_print)) + " articles...")
def web_scraper(input):
key_words = input.split(',')
keys = []
#dates = []
symbols = ['(',')',';','[',']','{','}','£','!','?','$','#','&',\
'@','^','*','+','\\','<','>','“','”','~']
time = str(datetime.datetime.now().date())
for key_word in key_words:
try:
while (key_word[0] == ' '):
key_word = key_word[1:]
while (key_word[len(key_word) - 1] == ' '):
key_word = key_word[:-1]
except IndexError:
continue
if (len(key_word) == 0):
continue
inputs = key_word.split(' ')
key = ''
for i in inputs:
key += i
key += '%20'
key = key[:-3].lower()
key = re.sub(r'[ ]+', ' ', key)
keys.append(key)
#dates.append(check.check_date(data_base, kkeys[i]))
for i in range(0, len(keys)):
s_key = fmt.file_name(keys[i], ' ')
l_key = ''
for letter in s_key:
if letter.isdigit() or letter.isalpha():
l_key += letter
else:
l_key += ' '
keys[i] = re.sub(r'[ ]+', '%20', l_key)
for i in range(0, len(keys)):
previous_len = 0
try:
for lines in open("Articles.bank", 'r', encoding='utf-8'):
previous_len += 1
except FileNotFoundError:
pass
stored_keys = check.load_keywords_info()
data_base = check.load_url_info()
data_print = {}
kkey = fmt.file_name(keys[i], '_')
for kk in stored_keys:
if abs(len(kkey) - len(kk)) <= 2 and check.MinEditDist(kkey,
kk) <= 1:
kkey = kk
keys[i] = re.sub(r'_', '%20', kkey)
if kkey in stored_keys:
date = int(re.sub(r'-', '', stored_keys[kkey]))
else:
date = 0
stored_keys[kkey] = time
if kkey not in data_base:
data_base[kkey] = []
cnn.cnn(data_base, data_print, keys[i], date, previous_len)
foxnews.foxnews(data_base, data_print, keys[i], date, previous_len)
cbs.cbs(data_base, data_print, keys[i], date, previous_len)
politico.politico(data_base, data_print, keys[i], date, previous_len)
#washington_post.washington_post(data_base,data_print,key,date)
if len(data_print) > 0:
check.save_keywords_info(stored_keys)
check.save_url_info(data_base)
output_file.output(data_print, data_base, previous_len)
print("Update date: " + time)
def politico(data_base, data_print, key, date_, previous_len):
kkey = fmt.file_name(key, '_')
kkkey = fmt.file_name(key, '+')
print("----- " + "politico." + kkey + " -----")
# https://www.politico.com/search/2?s=newest&q=tax%20reform&adv=true&c=0000014b-324d-d4f3-a3cb-f3ff415e0035&pv=0000014e-a307-d012-a3fe-bb8793910000
url1 = 'https://www.politico.com/search/'
url2 = '?s=newest&q="'
url3 = '"&adv=true&c=0000014b-324d-d4f3-a3cb-f3ff415e0035&pv=0000014e-a307-d012-a3fe-bb8793910000'
baseurl = url1 + '1' + url2 + key + url3
article_number = '0'
try:
page = hp.getHtml(baseurl)
except urllib.error.URLError:
print("Politico website is not correct, please update the scraper!")
return -1
article_number = regex.get_text(
'<h1>Results[^<]+?<\/h1>[^<]+?<p>(\d+?)\sSearch\sResults<\/p>',
page)[0]
if int(article_number) == 0:
print("No Policito article was found by this key word")
return -1
#get all urls
count = 0
page_num = 1
urls = defaultdict(str)
page_total = int(int(article_number) / 20 + 1)
reach_updated = False
print("There are " + article_number + " articles...")
print("Start loading URLs...")
while (count < page_total):
currenturl = url1 + str(page_num) + url2 + key + url3
try:
page = hp.getHtml(currenturl)
except urllib.error.URLError:
continue
url = regex.get_text('<a\shref="([^"]+?)"\s[^<]+?<\/a><\/h3>', page)
date = regex.get_text(
'<time datetime=.(\d+?\-\d+?\-\d+?)T\S+.>[^<]+?<\/time><\/p>',
page)
#title = regex.get_data('"title":"([^{]*?)",',page)
for cnt in range(0, len(date)):
date[cnt] = int(re.sub(r'-', '', date[cnt]))
if date_ > date[cnt]:
reach_updated = True
break
for i in range(0, cnt + 1):
try:
urls[url[i]] = str(date[i])[0:4] + '-' + str(
date[i])[4:6] + '-' + str(date[i])[6:8]
except IndexError:
break
if reach_updated:
break
page_num += 1
count += 1
print(str(len(urls)) + " URLs loaded...")
print("Updating database...")
for url in urls:
if url in data_base[kkey]:
continue
try:
html = hp.getHtml(url)
except urllib.error.URLError:
continue
title = regex.get_data('<title>([^-]+?)\s-\s[^<]+?<\/title>', html)
if title == 'None':
title = regex.get_data('<title>([^<]+?)<\/title>', html)
author = regex.get_data(
'<div\sitemprop="author"[^>]+?>[^<]+?<meta\s[^<]+?\s[^"]+?="([^>]+?)"\/>',
html)
text = regex.get_text('<p>([^\n]*?)</p>', html)
if text != []:
text = text[:-1]
if text == [] or title == "Noun":
continue
data_base[kkey].append(url)
data_print[url] = defaultdict(str)
# line 1
data_print[url]['ID'] = fmt.formatted_id(
len(data_base[kkey]) - 1 + previous_len)
# line 2
data_print[url]['key'] = fmt.formatted_key(kkey)
# line 3
data_print[url]['title'] = fmt.formatted_title(title)
# line 4
data_print[url]['source'] = fmt.formatted_source("Politico")
# line 5
data_print[url]['url'] = fmt.formatted_url(url)
# line 6
data_print[url]['date'] = fmt.formatted_date(urls[url])
# line 7
data_print[url]['author'] = fmt.formatted_author(author, ',')
# line 8
data_print[url]['content'] = fmt.formatted_content(text)
# line 9
#data_print[url[i]]['content2'] = fmt.formatted_content(text)
print('■', end='', flush=True)
print("\nThere are " + str(len(data_print) + previous_len) +
" articles...")
print("Updated " + str(len(data_print)) + " articles...")
def foxnews(data_base, data_print, key, date_, previous_len):
kkey = fmt.file_name(key, '_')
print("----- " + "foxnews." + kkey + " -----")
print("Start loading Urls...")
#case for exact keyword search
url1 = 'http://api.foxnews.com/v1/content/search?q="'
url2 = '"&fields=date,description,title,url,image,type,taxonomy&sort=latest§ion.path=fnc/opinion&type=article&start='
url3 = '&callback=angular.callbacks._0'
baseurl = url1 + key + url2 + '0' + url3
article_number = '0'
try:
page = hp.getHtml(baseurl)
except urllib.error.URLError:
print("Foxnews website is not correct, please update the scraper!")
return -1
article_number = regex.get_text('"response"\S\S"numFound":(\S+),"docs":\S',
page)[0]
if int(article_number) == 0:
print("No Foxnews article was found by this key word")
return -1
#get all urls
count = 0
index = 0
urls = defaultdict(str)
page_total = int(int(article_number) / 10 + 1)
reach_updated = False
print("There are " + article_number + " articles...")
print("Start loading and Updating...")
while (count < page_total):
currenturl = url1 + key + url2 + str(index) + url3
try:
page = hp.getHtml(currenturl)
except urllib.error.URLError:
continue
url = regex.get_text('url":\S"(\S+?)"\S', page)
#title = regex.get_data('"title":"([^{]*?)",',page)
for i in range(0, len(url)):
try:
d = regex.get_data('\/(\d+\/\d+\/\d+)', url[i])
except IndexError:
break
d_int = int(re.sub(r'/', '', d))
if date_ > d_int:
reach_updated = True
break
urls[url[i]] = re.sub(r'/', '-', d)
if reach_updated:
break
index += 10
count += 1
print(str(len(urls)) + " URLs loaded...")
print("Updating database...")
for url in urls:
if url in data_base[kkey]:
continue
try:
html = hp.getHtml(url)
except urllib.error.URLError:
continue
title = regex.get_data('<meta\sname="dc.title"\scontent="([^=]+?)">',
html)
author = regex.get_data(
'<meta\sname="dc.creator"\scontent="([^"]+?)">', html)
text = regex.get_text('<p[^>]*?>([^\n]*?)</p>[^<]*?<[^/]', html)
if text != []:
text = text[:-1]
if text == [] or title == "Noun":
continue
data_base[kkey].append(url)
data_print[url] = defaultdict(str)
# line 1
data_print[url]['ID'] = fmt.formatted_id(
len(data_base[kkey]) - 1 + previous_len)
# line 2
data_print[url]['key'] = fmt.formatted_key(kkey)
# line 3
data_print[url]['title'] = fmt.formatted_title(title)
# line 4
data_print[url]['source'] = fmt.formatted_source("Foxnews")
# line 5
data_print[url]['url'] = fmt.formatted_url(url)
# line 6
data_print[url]['date'] = fmt.formatted_date(urls[url])
# line 7
data_print[url]['author'] = fmt.formatted_author(author, ',')
# line 8
data_print[url]['content'] = fmt.formatted_content(text)
# line 9
#data_print[url[i]]['content2'] = fmt.formatted_content(text)
print('■', end='', flush=True)
print("\nThere are " + str(len(data_print) + previous_len) +
" articles...")
print("Updated " + str(len(data_print)) + " articles...")
