def parse_authors_date(data):
if 'page' in data:
if 'authors' in data['page'] and 'date' in data['page']:
if data['page']['authors'] and data['page']['date']:
try:
date = parse_datetime(data['page']['date']['pub'])
except:
date = None
author = ' and '.join(data['page']['authors'])
if 'by' in author.lower() or 'special to' in author.lower():
author = ''.join(
author.lower().split("special to")).replace(
'by', '').strip().title()
return author_title(author), date
else:
if 'author' in data and 'datePublished' in data:
try:
date = parse_datetime(data['datePublished'])
except:
date = None
author = data['author']['name']
if 'by' in author.lower() or 'special to' in author.lower():
author = ''.join(author.lower().split("special to")).replace(
'by', '').strip().title()
return author_title(author), date
def parse_blog(self, response):
# Post
blog = Posts()
blog['domain'] = self.domain
blog['url'] = response.url
blog['title'] = response.css('.entry-title::text').extract_first()
blog['author'] = response.xpath(
'//*[contains(@id,"post-")]/div[1]/div/div/header/div/div/a//text()'
).extract_first()
date = response.css('.updated::text').extract_first()
blog['published_date'] = parse_datetime(date) if date else None
#-Cleaning Post
posts = "".join(
response.xpath("//div[contains(@class, 'td-post-content')]//text()"
).extract()).strip().replace('\n', ' ').replace(
'\t', ' ')
posts = str(re.sub(' +', ' ', posts))
blog['content'] = posts.replace(
'Facebook Twitter reddit LinkedIn', ''
).replace(
'We’re asking readers, like you, to make a contribution in support of True North’s fact-based, independent journalism.\r Unlike the mainstream media, True North isn’t getting a government bailout. Instead, we depend on the generosity of Canadians like you.\r How can a media outlet be trusted to remain neutral and fair if they’re beneficiaries of a government handout? We don’t think they can.\r This is why independent media in Canada is more important than ever. If you’re able, please make a tax-deductible donation to True North today. Thank you so much.',
'').replace('\r', ' ')
blog['content_html'] = "".join(
response.xpath(
"//div[contains(@class, 'td-post-content')]").extract())
blog['links'] = get_links("".join(
response.xpath(
"//div[contains(@class, 'td-post-content')]").extract()))
blog['tags'] = None
yield blog
def parse_blog(self, response):
# Posts
blog = Posts()
blog['domain'] = self.domain
blog['url'] = response.url
blog['title'] = parse_title(response)
date, author = parse_author_date(response)
author = parse_author(response)
blog['author'] = str(author)[0:99] if author else None
blog['published_date'] = parse_datetime(date) if date else None
content, content_html = parse_content(response)
if content_html:
blog['content'] = content
blog['content_html'] = content_html
blog['links'] = get_links(content_html)
tags = response.xpath('//*[@id="article-tags"]/div//text()').extract()
blog['tags'] = tags_to_json(list(filter(lambda x: '\n' not in x and '\t' not in x, tags))) if tags else None
yield blog
# Comments
comment_data = facebook_comments(response.url, 318812448281278)
comments = comment_data['comments']
authors = comment_data['authors']
reply_dic = comment_data['reply_dic']
if comments: #Catches no comments
for c in comments:
parsed_comment = Comments()
parsed_comment['domain'] = self.domain
parsed_comment['url'] = response.url
parsed_comment['comment_id'] = c['id']
parsed_comment['username'] = [x['name'] for x in authors if c['authorID'] == x['id']][0]
parsed_comment['user_id'] = c['authorID']
parsed_comment['comment'] = c['body']['text']
parsed_comment['comment_original'] = None
parsed_comment['links'] = get_links(c['body']['text'])
parsed_comment['upvotes'] = c['likeCount']
parsed_comment['downvotes'] = None
parsed_comment['published_date'] = parse_datetime(c['timestamp']['text'])
if 'public_replies' in c:
parsed_comment['reply_count'] = len([x for x in comments if 'targetID' in x and x['targetID'] == c['id']])
else:
parsed_comment['reply_count'] = 0
if c['id'] in reply_dic:
parsed_comment['reply_to'] = reply_dic[c['id']]
else:
parsed_comment['reply_to'] = None
yield parsed_comment
# #Stats
stat = Stats()
stat['domain'] = self.domain
stat['url'] = response.url
stat['views'] = None
stat['likes'] = None
if comments is None:
stat['comments'] = 0
else:
stat['comments'] = len(comments)
yield stat
def parse_blog(self, response):
# Posts
script = response.xpath(
"//script[contains(., 'identity')]/text()").extract_first()
try:
data = json.loads(script)
except Exception as e:
print(str(e) + f"\n{str(response.url)}")
data = {}
if data:
blog = Posts()
blog['domain'] = get_domain(response.url)
blog['url'] = response.url
blog['title'] = response.css(
'.article-title::text').extract_first()
author, date = parse_authors_date(data) if data else None
blog['author'] = author.replace(" ",
"").strip() if author else None
blog['published_date'] = date if date else None
blog['tags'] = tags_to_json(
data['page']['tags']) if 'page' in data else None
blog['content'] = get_content(response)
blog['content_html'] = " ".join(
response.xpath('//*[@class="article-content"]').extract())
blog['links'] = get_links(" ".join(
response.xpath('//*[@class="article-content"]').extract()))
yield blog
else:
print('here')
pass
# Comments
article_id = data['page']['articleId'] if 'page' in data else None
comments = get_torontosun_comments(article_id)
if comments: #Catches no comments
for c in comments:
if 'content' in c and c['content']: #Skipping empty comments
parsed_comment = Comments()
parsed_comment['domain'] = self.domain
parsed_comment['url'] = response.url
parsed_comment['comment_id'] = c['content_uuid']
parsed_comment[
'username'] = None #Could not find API to get this
parsed_comment['user_id'] = c['actor_uuid']
parsed_comment[
'comment'] = c['content'] if 'content' in c else None
parsed_comment['comment_original'] = None
parsed_comment['links'] = get_links(
c['content']) if 'content' in c else None
parsed_comment['upvotes'] = c['total_likes']
parsed_comment['downvotes'] = c['total_dislikes']
parsed_comment['published_date'] = parse_datetime(
time.strftime('%m/%d/%Y %H:%M:%S',
time.gmtime(c['date_created'] / 1000.)))
if c['total_replies'] > 0:
parsed_comment['reply_count'] = c['total_replies']
else:
parsed_comment['reply_count'] = 0
if c['content_container_uuid'] != c['thread_uuid'] and c[
'content_container_uuid'] != c['parent_uuid']:
parsed_comment['reply_to'] = c['thread_uuid']
else:
parsed_comment['reply_to'] = None
yield parsed_comment
#Stats
stat = Stats()
stat['domain'] = self.domain
stat['url'] = response.url
stat['views'] = None
stat['likes'] = None
if comments is None:
stat['comments'] = 0
else:
stat['comments'] = len(comments)
yield stat
def parse_article(self, response):
blog = Posts()
blog['domain'] = self.domain
blog['url'] = response.url
blog['title'] = response.xpath(
'//*[@class="news-article-header__title"]/text()').get()
blog['author'] = response.xpath(
'//*[@class="news-byline-full__info-wrapper"]/span/text()').get()
blog['published_date'] = get_date(
response.xpath(
'//*[@class="news-article-header__timestamps-posted"]/text()').
get())
blog['content'] = " ".join(
response.xpath(
'//*[@id="js-post-container"]/div/div[1]/div[1]/div/p/text()').
getall()).strip()
blog['content_html'] = " ".join(
response.xpath(
'//*[@id="js-post-container"]/div/div[1]/div[1]/div').getall())
blog['links'] = get_links(blog['content_html'])
blog['tags'] = tags_to_json(parse_tags(response))
yield blog
#Comments requests
article_url = format_comment_url(response.url)
comment_data = facebook_comments(article_url, '162111247988300')
comments = comment_data['comments']
authors = comment_data['authors']
reply_dic = comment_data['reply_dic']
if comments: #Catches no comments
for c in comments:
parsed_comment = Comments()
parsed_comment['domain'] = self.domain
parsed_comment['url'] = response.url
parsed_comment['comment_id'] = c['id']
parsed_comment['username'] = [
x['name'] for x in authors if c['authorID'] == x['id']
][0]
parsed_comment['user_id'] = c['authorID']
parsed_comment['comment'] = c['body']['text']
parsed_comment['comment_original'] = None
parsed_comment['links'] = get_links(c['body']['text'])
parsed_comment['upvotes'] = c['likeCount']
parsed_comment['downvotes'] = None
parsed_comment['published_date'] = parse_datetime(
c['timestamp']['text'])
if 'public_replies' in c:
parsed_comment['reply_count'] = len([
x for x in comments
if 'targetID' in x and x['targetID'] == c['id']
])
else:
parsed_comment['reply_count'] = 0
if c['id'] in reply_dic:
parsed_comment['reply_to'] = reply_dic[c['id']]
else:
parsed_comment['reply_to'] = None
yield parsed_comment
#Stats
stat = Stats()
stat['domain'] = self.domain
stat['url'] = response.url
stat['views'] = None
stat['likes'] = None
if comments is None:
stat['comments'] = 0
else:
stat['comments'] = len(comments)
yield stat
def get_date(date_string):
date_string = date_string.strip()
date_string = date_string.replace('Posted on ', '')
date_string = date_string.replace('Last updated on ', '')
return parse_datetime(date_string)