def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = ''
t = ''
t_stamp = 0
category = ''
author = ''
content = ''
if instant_pattern.match(task.url):
title = util.get_filtered_title(doc, {'title'}, ur' - 信報網站 hkej.com')
t = util.get_time_string_from_selectors(doc, {'span.date'})
time_part = min_sec_pattern.findall(t)[0]
t_stamp = util.get_timestamp_from_string(time_part) + time.localtime().tm_sec
category = doc('span.cate').text()
content = util.get_paragraphs_from_selector(doc, '#article-content p')
elif daily_pattern.match(task.url) or headline_article_pattern.match(task.url):
title = util.get_filtered_title(doc, {'title'}, ur' - .+')
t = util.get_time_string_from_selectors(doc, {'#date'})
t_stamp = util.get_timestamp_from_string(t) + time.localtime().tm_hour*3600 + time.localtime().tm_min*60 + time.localtime().tm_sec
category = doc('#hkej_navSubMenu_2014 .on').text()
content = util.get_paragraphs_from_selector(doc, '#article-content p')
if content == '':
content = util.get_paragraphs_from_selector(doc, '#article-detail-wrapper')
content = re.sub(ur'（節錄）(.|\n|\t|\r)*', u'', content, re.M | re.I | re.U)
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'HKEJ'
item.task_no = self.BATCH_NUMBER
for img in doc('#article-detail-wrapper p img, #article-detail-wrapper .hkej_detail_thumb_2014 img').items():
if img.parent('a').attr('href') != '':
des = ''
if img.parent('a') and img.parent('a').attr('title'):
des = img.parent('a').attr('title')
media = self.NewsItem.MediaItem(media_url=img.parent('a').attr('href'), type='image', description=des,
created_at=item.fetched_at)
item.media_list.append(media)
for a in doc('iframe').items():
if a.attr('src') and re.match(r'.*youtube\.com.*', a.attr('src')):
media_u = a.attr('src')
if re.match(r'//.+', media_u):
media_u = 'http:' + media_u
media = self.NewsItem.MediaItem(media_url=media_u, type='youtube', description='youtube',
created_at=item.fetched_at)
item.media_list.append(media)
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'h1.entry-title'})
t = doc('meta[property="article:published_time"]').attr('content')
t_stamp = 0
if t:
t_stamp = util.get_timestamp_from_string(t)
category = ''
cat_find_res = cat_pattern.findall(task.url)
if cat_find_res:
category = cat_find_res[0]
category = u'科技/' + category
author = ''
content = util.get_paragraphs_from_selector(doc, '#content div.entry p:not(.meta)')
content = re.sub(ur'(來源：|來源:|Tags:).+', u'', content)
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'Unwire'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = doc('h1').text()
meta_txt = doc('.metaStuff').text()
t = re.findall(ur'[^\s]+月.+', meta_txt)[0]
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_hour * 3600 + time.localtime(
).tm_min * 60 + time.localtime().tm_sec
category = doc('#crumbs a').text().split(' ')[-1]
author = meta_txt.split(' ')[0]
content = util.get_paragraphs_from_selector(doc, '.entry p')
if content == '':
content = util.get_paragraphs_from_selector(doc, '.entry div')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'Savantas'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'h1 a'})
t = util.get_time_string_from_selectors(doc, {'span.postdate'})
t_stamp = util.get_timestamp_from_string(t) + int(
time.localtime().tm_sec)
category = doc('span.postcat a').text()
author = doc('span.postauthor a').text()
content = util.get_paragraphs_from_selector(doc, 'div p')
content = re.sub(ur'投稿:[.\n\r\t]*.*', u'', content, re.M | re.U | re.I)
content = re.sub(ur'則留言[.\n\r\t]*', u'', content, re.M | re.U | re.I)
content = re.sub(ur'大道之行也，天下為公，選賢與能，講信修睦。－－－《禮運．大同》[.\n\r\t]*', u'',
content, re.M | re.U | re.I)
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'VJMedia'
item.task_no = self.BATCH_NUMBER
for img in doc(
'#container img.size-full, #container img.size-large').items():
if img.attr('src') != '':
des = ''
if img.attr('alt'):
des = img.attr('alt')
elif img.siblings('p'):
des = img.siblings('p').text()
media = self.NewsItem.MediaItem(media_url=img.attr('src'),
type='image',
description=des,
created_at=item.fetched_at)
item.media_list.append(media)
for iframe in doc('iframe').items():
if iframe.attr('src') and re.match(r'.*youtube\.com.+',
iframe.attr('src')):
media = self.NewsItem.MediaItem(media_url=iframe.attr('src'),
type='youtube',
description='youtube',
created_at=item.fetched_at)
item.media_list.append(media)
if util.within_active_interval(6, 1200):
_comments = util.get_filtered_facebook_comments_data(
'214585295294555',
doc('div.fb-comments').attr('data-href'), task.url)
if _comments:
for _comment in _comments:
item.media_list.append(
self.NewsItem.MediaItem(
media_url=_comment['json_string'],
type='comments',
description='comments',
created_at=item.fetched_at))
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'title'}, ur' – 都市日報')
author = ''
category = doc('.mobile-page-name span').text()
# tags = doc('meta[name=keywords]').attr('content')
content = util.get_paragraphs_from_selector(
doc, '.main-content .content p')
t = doc('.main-content .date p').text()
t = re.sub(ur'\(.+?\)', '', t)
t = re.sub(ur'上午', 'AM', t)
t = re.sub(ur'下午', 'PM', t)
t_stamp = util.get_timestamp_from_string(t)
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'MetroHK'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'.PlaylistRow kanhanpass'})
t = util.get_time_string_from_selectors(doc, {'.PlaylistRow td'},
date_patterns={date_pattern})
t_stamp = util.get_timestamp_from_string(t, '%Y/%m/%d %I:%M %p') + int(
time.localtime().tm_sec)
category = doc('#topMenu a.on').text()
author = ''
c_id = re.findall(url_id_pattern, task.url)[0]
r = requests.get(
'http://webcast.legco.gov.hk/Public_uat_embedded/Service1.asmx/GetTimeMarker?meetingID='
+ c_id + '&lang=tc')
j_obj = json.loads(r.text)
content = ''
for agenda in j_obj['TimeMarkerItems']:
a_time = agenda['AgendaTime']
a_item = agenda['AgendaItem']
content = content + a_time + u' - ' + a_item + u'\n'
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'LegislativeCouncil'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'h1'})
t = doc('meta[property="article:published_time"]').attr('content')
t_stamp = 0
if t:
t_stamp = util.get_timestamp_from_string(t)
category = ''
author = ''
content = util.get_paragraphs_from_selector(doc, '#main-content .entry-content p')
content = re.sub(ur'繼續閱讀[\n\s\S.]*', '', content)
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'UBeat'
item.task_no = self.BATCH_NUMBER
for img in doc('figure.entry-thumbnail img').items():
if img.attr('src') != '':
media_u = img.attr('src')
des = ''
if img.attr('alt'):
des = img.attr('alt')
media = self.NewsItem.MediaItem(media_url=media_u, type='image', description=des,
created_at=item.fetched_at)
item.media_list.append(media)
def normal_item_solver(self, item, task, response):
response.encoding = 'utf-8'
doc = self.get_doc(response)
title = doc('#page-h1').text()
t = util.get_time_string_from_selectors(doc, {'h5 small'},
{date_pattern})
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_hour * 3600 + time.localtime(
).tm_min * 60 + time.localtime().tm_sec
category = doc('h5 small a').text()
author = ''
content = util.get_paragraphs_from_selector(doc, '.content-show p')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'Bauhinia'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
response.encoding = 'big5'
doc = self.get_doc(response)
title = doc('p.bigheading').text().split(' ')[-1]
with self.url_time_dict_lock:
t = self.url_time_dict[task.url]
t_stamp = util.get_timestamp_from_string(t) + time.localtime(
).tm_hour * 3600 + time.localtime().tm_min * 60
category = ''
author = ''
content = util.get_paragraphs_from_selector(doc, 'p:not(.bigheading)')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'Y28'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'title'}, u'881903.com 商業電台 - ')
t = util.get_time_string_from_selectors(
doc, {'#divnewsTextDate', '#part6808_ctl00_lblDetailDate'})
t_stamp = util.get_timestamp_from_string(t, '%d.%m.%Y %H:%M') + int(
time.localtime().tm_sec)
category = doc('#part8425_ctl00_divtitle').text()
author = ''
content = util.get_paragraphs_from_selector(doc,
'#divnewsTextContent p')
if content == '':
content = util.get_paragraphs_from_selector(doc, '#tdContent p')
if content == '':
content = util.get_paragraphs_from_selector(
doc, '.newsTextContent2')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'CommercialRadio'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'h2'})
t = util.get_time_string_from_selectors(doc, {'div.dateforarticle'})
t_stamp = 0
if relative_time_pattern.match(t):
t_stamp = self._get_timestamp_from_relative_time_str(t)
elif absolute_time_pattern.match(t):
t_stamp = util.get_timestamp_from_string(t) + int(
time.localtime().tm_sec)
category = '新聞'
author = ''
content = util.get_paragraphs_from_selector(doc, '#mymain')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'AM730'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'h1.conttit'})
t = util.get_time_string_from_selectors(doc, {'div.pubtime'})
t_stamp = util.get_timestamp_from_string(t)
category = 'hk'
author = ''
content = util.get_paragraphs_from_selector(doc, 'div.contentbox p')
content = re.sub(r'READMORE\: .+\n', '', content)
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'ChinaDaily'
item.task_no = self.BATCH_NUMBER
for img in doc('div.contentbox img').items():
if img.attr('src') != '':
media_u = 'http://www.chinadailyasia.com/' + re.sub(r'.+(?=attachement)', '', img.attr('src'))
des = ''
if img.attr('alt'):
des = img.attr('alt')
elif img.siblings('p'):
des = img.siblings('p').text()
media = self.NewsItem.MediaItem(media_url=media_u, type='image', description=des,
created_at=item.fetched_at)
item.media_list.append(media)
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = doc('h1').text()
t = ''
category = ''
with self.time_cat_dict_lock:
if task.url in self.time_cat_dict:
t = self.time_cat_dict[task.url][0]
category = self.time_cat_dict[task.url][1]
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_hour * 3600 + time.localtime(
).tm_min * 60 + time.localtime().tm_sec
author = ''
content = util.get_paragraphs_from_selector(doc, '#content p')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'HKEDB'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'.article-title'})
t = ''
with self.url_time_dict_lock:
t = self.url_time_dict[task.url]
print t
t_stamp = util.get_timestamp_from_string(t) + time.localtime().tm_hour*3600 + time.localtime().tm_min*60 + time.localtime().tm_sec
category = '新聞發佈'
author = ''
content = util.get_paragraphs_from_selector(doc, '.article-content')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'NewCenturyForum'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = doc('h4').text()
t = doc('#ContentPlaceHolder1_IndividualNewsList_lblTime_0').text()
t_stamp = util.get_timestamp_from_string(t,
time_format='%d/%m/%Y %H:%M')
category = ''
cat = doc('title').text()
if cat_pattern.findall(cat):
category = cat_pattern.findall(cat)[0]
author = ''
content = util.get_paragraphs_from_selector(
doc, '#ContentPlaceHolder1_IndividualNewsList_lblContent_0')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'MetroFinance'
item.task_no = self.BATCH_NUMBER
item.id = ''
if news_id_pattern.findall(task.url):
item.id = (news_id_pattern.findall(task.url)[0])[2:]
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'h1'})
t_divs = doc('h1').siblings('div').items()
t = ''
t_stamp = 0
for _div in t_divs:
if _div.css('color') == 'rgb(128, 128, 128)':
t = _div.text()
t_stamp = util.get_timestamp_from_string(t) + int(
time.localtime().tm_sec)
break
category = u'电子'
author = ''
content = util.get_paragraphs_from_selector(doc, '.href_txt_blog2')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'HKGolden'
item.task_no = self.BATCH_NUMBER
def task_filter(self, doc, url, doc_url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
if year_date_pattern.findall(url):
t = year_date_pattern.findall(url)[0]
t_stamp = util.get_timestamp_from_string(t)
else:
if doc('time') or doc('span.date'):
t = util.get_time_string_from_selectors(doc, {'time', 'span.date'})
t_stamp = util.get_timestamp_from_string(t)
else:
return True
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
return False
def normal_item_solver(self, item, task, response):
response.encoding = 'utf-8'
doc = self.get_doc(response)
title = re.sub(u'香港金融管理局 - ', '',
re.findall(title_pattern, response.text)[0])
t = re.sub(u'修訂日期: ', '', doc('#lastUpdate').text())
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_hour * 3600 + time.localtime(
).tm_min * 60 + time.localtime().tm_sec
category = doc('#content').text().split(' ')[0]
author = ''
content = doc('.item').text()
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'HKMA'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = doc('#News_Body_Title').text()
t = ''
if doc('#News_Body_Time'):
t = date_pattern.findall(str(doc('#News_Body_Time')))[0]
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_hour * 3600 + time.localtime(
).tm_min * 60 + time.localtime().tm_sec
category = doc('.Top_Index_A a:last-child').text()
author = ''
content = util.get_paragraphs_from_selector(doc, '#News_Body_Txt_A p')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'FMCOPRC'
item.task_no = self.BATCH_NUMBER
def normal_item_solver(self, item, task, response):
response.encoding = 'utf-8'
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'.article-header h1'})
t = doc('.date').text()
if date_pattern.findall(t):
t = date_pattern.findall(t)[0]
t_stamp = util.get_timestamp_from_string(t)
category = doc('.now-here').text()
author = ''
content = util.get_paragraphs_from_selector(doc, '.article p')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'JD Online'
item.task_no = self.BATCH_NUMBER
def task_filter(self, doc, url, doc_url):
wanted = False
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
t = full_article_pattern.findall(url)[0]
if util.get_timestamp_from_string(t, '%m-%d-%Y') >= util.get_day_stamp(offset=self.OFFSET):
wanted = True
return wanted
def page_filter(self, doc, url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
t = util.get_time_string_from_selectors(doc, {'span.time'})
t_stamp = util.get_timestamp_from_string(t)
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
return False
def page_filter(self, doc, url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
t = doc('div[itemprop="dateCreated"]').attr('datetime')
t_stamp = util.get_timestamp_from_string(t) + 8 * 3600
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
return False
def page_filter(self, doc, url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
t = doc('.updated').text()
t_stamp = util.get_timestamp_from_string(t)
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
return False
def page_filter(self, doc, url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
meta_txt = doc('.metaStuff').text()
t = re.findall(ur'[^\s]+月.+', meta_txt)[0]
t_stamp = util.get_timestamp_from_string(t)
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
return False
def page_filter(self, doc, url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
t = util.get_time_string_from_selectors(
doc, {'#divnewsTextDate', '#part6808_ctl00_lblDetailDate'})
t_stamp = util.get_timestamp_from_string(t, '%d.%m.%Y %H:%M')
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
return False
def page_filter(self, doc, url):
wanted = False
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
if doc('div.post_time'):
t = util.get_time_string_from_selectors(doc, {'div.post_time'})
t_stamp = util.get_timestamp_from_string(t)
if t_stamp >= util.get_day_stamp():
return True
return wanted
def normal_item_solver(self, item, task, response):
doc = self.get_doc(response)
title = util.get_filtered_title(doc, {'title'},
u' - BBC 中文网| BBC Zhongwen')
t = ''
t_stamp = 0
if doc('.story-body .mini-info-list .date').attr('data-datetime'):
t = doc('.story-body .mini-info-list .date').attr('data-datetime')
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_sec
elif doc('.timeline-status h3') and date_pattern.findall(
doc('.timeline-status h3').text()):
t = date_pattern.findall(doc('.timeline-status h3').text())[0]
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_sec
elif doc('.story-body .date strong') and date_pattern.findall(
doc('.story-body .date strong').text()):
t = date_pattern.findall(doc('.story-body .date strong').text())[0]
t_stamp = util.get_timestamp_from_string(
t) + time.localtime().tm_sec
category = doc('meta[property="article:section"]').attr('content')
author = doc('span.byline__name').text()
content = util.get_paragraphs_from_selector(
doc, 'div[property="articleBody"] p')
if content == '':
content = util.get_paragraphs_from_selector(
doc, '.article-wrapper p')
if content == '':
content = util.get_paragraphs_from_selector(doc, '.map-body p')
item.raw = doc.text()
item.title = title
item.t = t
item.t_stamp = t_stamp
item.fetched_at = task.fetched_at
item.category = category
item.author = author
item.content = content
item.url = task.url
item.source = 'BBC Chinese'
item.task_no = self.BATCH_NUMBER
def task_filter(self, doc, url, doc_url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
if not reg_pattern.match(doc_url):
return True
t = util.get_time_string_from_selectors(doc, {'#article_date'})
t_stamp = util.get_timestamp_from_string(t)
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
def page_filter(self, doc, url):
for reg_pattern in self.reg_patterns:
if reg_pattern.match(url):
t = doc('meta[property="article:published_time"]').attr(
'content')
t_stamp = util.get_timestamp_from_string(t)
if t_stamp >= util.get_day_stamp(self.OFFSET):
return True
return False
return False