def _extract_comments(tree):
"""解析出评论
Args:
tree: ElementTree
Returns:
comments: list
"""
comment_elems = tree.xpath("//div[@id='REVIEWS']/div[@class='reviewSelector']")
comments = []
for comment_elem in comment_elems:
try:
# 第一类型网页
comment_user = flist(comment_elem.xpath(
"div/div[@class='col1of2']"
"//div[@class='username mo']/span/text()"), u"")
comment_score = flist(comment_elem.xpath(
"div/div[@class='col2of2 ']"
"/div[@class='rating reviewItemInline']"
"/span[@class='rate rate_s s50']/img/@content"), u"")
comment_time = flist(comment_elem.xpath(
"div/div[@class='col2of2 ']"
"/div[@class='rating reviewItemInline']"
"/span[@class='ratingDate']/text()"), u"")
content_elems = comment_elem.xpath(
"div/div[@class='col2of2 ']/div[@class='entry']/p")
texts = []
for content_elem in content_elems:
if content_elem.text is not None:
texts.append(content_elem.text)
comment_content = u"".join(texts).strip()
# 第二类型网页
if len(comment_user.strip()) == 0:
comment_user = flist(comment_elem.xpath(
"div/div[@class='col1of2']"
"//div[@class='username']/span/text()"), u"")
if len(comment_time.strip()) == 0:
comment_time = flist(comment_elem.xpath(
"div/div[@class='col2of2']"
"/span[@class='ratingDate']/text()"), u"")
if len(comment_content.strip()) == 0:
content_elems = comment_elem.xpath(
"div/div[@class='col2of2']"
"/div[@class='entry']/p")
texts = []
for content_elem in content_elems:
if content_elem.text is not None:
texts.append(content_elem.text)
comment_content = u"".join(texts).strip()
except Exception, e:
print "extract one comment failed error:%s" % e
print traceback.format_exc()
else:
if len(unicode(comment_content)) != 0:
comment_item = CommentItem(unicode(comment_user).strip(),
unicode(comment_time).strip(),
unicode(comment_score).strip(),
unicode(comment_content).strip())
comments.append(comment_item)
def parse(self, task, input_file):
"""解析函数
Args:
task: HttpTask，任务
input_file: StringIO, 网页文件
Yields:
item
"""
self.logger.debug("attraction parser start to parse")
content = input_file.read()
tree = html.parse(StringIO(content))
try:
zip_code = flist(tree.xpath("//span[@class='postal-code']/text()"), u"")
play_spend, play_spend_unit = _extract_play_spend_and_unit(content)
tel_phone = flist(tree.xpath("//div[@id='HEADING_GROUP']"
"/div[@class='wrap infoBox']"
"/div[@class='odcHotel blDetails']"
"/div/div[@class='fl']/text()"), u"")
open_time = u""
total_score = flist(tree.xpath("//div[@class='rs rating']"
"/span/img/@content"))
ticket_info = u""
preview_relate_path = flist(tree.xpath(
"//div[@class='listing_description']/a/@href"), u"")
lon, lat = _extract_lon_lat(flist(tree.xpath(
"//div[@class='js_mapThumb']"
"/div[@id='bmapContainer']/img[1]/@src"), u""))
comments = _extract_comments(tree)
# 生成景点信息(不包括description)
attraction_item = AttractionItem(task.request.url,
task.kwargs['name'],
unicode(play_spend),
play_spend_unit,
task.kwargs['address'],
unicode(tel_phone), unicode(open_time),
unicode(total_score),
unicode(ticket_info),
task.kwargs['hot'],
lon, lat, task.kwargs['rank'],
comments,
unicode(zip_code))
yield attraction_item
# 生成description任务
if len(preview_relate_path) != 0:
description_request = build_description_request(
task.request.url, preview_relate_path)
description_task = HttpTask(description_request,
callback="DescriptionParser",
max_fail_count=3,
cookie_host=LVYOU_HOST,
kwargs={'url': task.request.url})
yield description_task
else:
yield DescriptionItem(task.request.url, u"")
except Exception, e:
print "error:%s" % e
print traceback.format_exc()
def _extract_address(info_elem):
"""解析地址
Args:
info_elem: Element
Returns:
address: unicode, 地址信息
"""
country = flist(info_elem.xpath("address/span[@class='country-name']/text()"), u"")
city = flist(info_elem.xpath("address/span[@class='locality']/text()"), u"")
streat_script = flist(info_elem.xpath("address/span/script/text()"), u"")
streat = _extract_script_text(streat_script)
return u"%s%s%s" % (country, city, streat)
def _extract_address(info_elem):
"""解析地址
Args:
info_elem: Element
Returns:
address: unicode, 地址信息
"""
country = flist(
info_elem.xpath("address/span[@class='country-name']/text()"), u"")
city = flist(info_elem.xpath("address/span[@class='locality']/text()"),
u"")
streat_script = flist(info_elem.xpath("address/span/script/text()"), u"")
streat = _extract_script_text(streat_script)
return u"%s%s%s" % (country, city, streat)
def parse(self, task, input_file):
"""详情解析器
Args:
task, HttpTask, 任务
input_file: file, 网页文件
Yields:
item: WebItem, 数据
task: HttpTask, 新任务
"""
tree = html.parse(input_file)
pic_url = unicode(
flist(tree.xpath("//div[@class='product-price-left']"
"/p/img/@src"),
default=u""))
desc_elems = tree.xpath("//div[@class='product-detail-alla-cont']")
description = _extract_desc_elems(desc_elems)
date_elems = tree.xpath("//ul[@class='productnew-header-pricea2-ul"
" clearfloat']/li/@d")
telephone = flist(tree.xpath("//div[@class='top-w']//li[@class='tel']"
"/span/text()"),
default=u"")
telephone = telephone.replace(u"-", u"")
if len(telephone) == 0:
telephone = u"4006228228"
price_elems = tree.xpath("//ul[@class='productnew-header-pricec2-ul "
"productnew-"
"header-pricec3-ul productnew-header-"
"pricec2-cq']/li/@title")
price_infos = list()
for price_elem in price_elems:
if unicode(price_elem) not in price_infos:
price_infos.append(unicode(price_elem))
price_info = u"/".join(price_infos)
time_infos = []
for date_elem in date_elems:
time_infos.append(date_elem)
time_info = u";".join(time_infos)
url = task.kwargs.get('url')
cookie_host = task.kwargs.get('cookie_host')
cookie_count = task.kwargs.get('cookie_count')
pictures, pic_task = self._check_and_execute_picture(
pic_url, cookie_host, cookie_count)
# 保存详情信息
yield WebItem(url, telephone, description, pictures, time_info,
price_info, u"")
# 抛出picTask
if pic_task is not None:
yield pic_task
def _extract_play_spend(tree):
"""提取play spend和unit
"""
play_spend = flist(tree.xpath(
"//div[@id='J-aside-info-recommend_visit_time']"
"/span[@class='val recommend_visit_time-value']/text()"), u"")
if play_spend.find(u"天") != -1:
play_spend_unit = u"天"
elif play_spend.find(u"小时") != -1:
play_spend_unit = u"小时"
elif play_spend.find(u"分") != -1 or play_spend.find(u"分钟") != -1:
play_spend_unit = u"分"
elif len(play_spend.strip()) == 0:
play_spend_unit = u""
else:
play_spend_unit = u""
replace_dict = {
u"小时": u"", u"天": u"", u"时": u"", u"小": u"", u"分钟": u"", u"分": u"",
u"钟": u"",
u"至": u"-", u"超过": u">", u"半": u".5", u"一": u"1",
u"二": u"2", u"三": u"3", u"四": u"4", u"五": u"5",
u"六": u"6", u"七": u"7", u"八": u"8", u"九": u"9",
}
for key, value in replace_dict.iteritems():
play_spend = play_spend.replace(key, value)
if play_spend.strip() == u".5":
play_spend = u"0.5"
return play_spend, play_spend_unit
def parse(self, task, input_file):
"""详情解析器
Args:
task, HttpTask, 任务
input_file: file, 网页文件
Yields:
item: WebItem, 数据
task: HttpTask, 新任务
"""
tree = html.parse(input_file)
pic_url = unicode(flist(tree.xpath("//div[@class='product-price-left']"
"/p/img/@src"),
default=u""))
desc_elems = tree.xpath("//div[@class='product-detail-alla-cont']")
description = _extract_desc_elems(desc_elems)
date_elems = tree.xpath("//ul[@class='productnew-header-pricea2-ul"
" clearfloat']/li/@d")
telephone = flist(tree.xpath("//div[@class='top-w']//li[@class='tel']"
"/span/text()"), default=u"")
telephone = telephone.replace(u"-", u"")
if len(telephone) == 0:
telephone = u"4006228228"
price_elems = tree.xpath("//ul[@class='productnew-header-pricec2-ul "
"productnew-"
"header-pricec3-ul productnew-header-"
"pricec2-cq']/li/@title")
price_infos = list()
for price_elem in price_elems:
if unicode(price_elem) not in price_infos:
price_infos.append(unicode(price_elem))
price_info = u"/".join(price_infos)
time_infos = []
for date_elem in date_elems:
time_infos.append(date_elem)
time_info = u";".join(time_infos)
url = task.kwargs.get('url')
cookie_host = task.kwargs.get('cookie_host')
cookie_count = task.kwargs.get('cookie_count')
pictures, pic_task = self._check_and_execute_picture(pic_url,
cookie_host,
cookie_count)
# 保存详情信息
yield WebItem(url, telephone, description, pictures, time_info,
price_info, u"")
# 抛出picTask
if pic_task is not None:
yield pic_task
def parse(self, task, input_file):
"""详情解析器
Args:
task, HttpTask, 任务
input_file: file, 网页文件
Yields:
item: WebItem, 数据
task: HttpTask, 新任务
"""
tree = html.parse(input_file)
name = flist(tree.xpath(u"//div["
u"@clas='product-price-titleul']/h1/text()"))
desc_elems = tree.xpath(u"//div[@class='product-detail-alla-cont']")
description = _extract_desc_elems(desc_elems)
date_elems = tree.xpath(
u"//ul[@class='productnew-header-pricea2-ul clearfloat']/li/@d")
telephone = flist(tree.xpath(
u"//div[@class='top-w']//li[@class='tel']/span/text()"))
telephone = telephone.replace(u"-", u"")
if len(telephone) == 0:
telephone = u"4006228228"
price_elems = tree.xpath(
u"//ul[@class='productnew-header-pricec2-ul productnew-"
u"header-pricec3-ul productnew-header-pricec2-cq']/li/@title")
price_infos = list()
for price_elem in price_elems:
if unicode(price_elem) not in price_infos:
price_infos.append(unicode(price_elem))
price_info = u"/".join(price_infos)
time_infos = []
for date_elem in date_elems:
time_infos.append(date_elem)
time_info = u";".join(time_infos)
url = task.request.url
# 保存详情信息
yield WebItem(url, telephone, description,
time_info, price_info, name)
def _extract_tips(tree):
"""提取出小提示
Args:
tree: ElementTree
Returns:
tips: unicode
"""
elem = flist(tree.xpath("//div[@id='mod-attention']/"
"article[@class='content-article']"), None)
if elem is None:
return u""
else:
texts = []
for text_child in elem.itertext():
texts.append(text_child)
return u"".join(texts)
def _extract_tips(tree):
"""提取出小提示
Args:
tree: ElementTree
Returns:
tips: unicode
"""
elem = flist(
tree.xpath("//div[@id='mod-attention']/"
"article[@class='content-article']"), None)
if elem is None:
return u""
else:
texts = []
for text_child in elem.itertext():
texts.append(text_child)
return u"".join(texts)
def _extract_play_spend(tree):
"""提取play spend和unit
"""
play_spend = flist(
tree.xpath("//div[@id='J-aside-info-recommend_visit_time']"
"/span[@class='val recommend_visit_time-value']/text()"),
u"")
if play_spend.find(u"天") != -1:
play_spend_unit = u"天"
elif play_spend.find(u"小时") != -1:
play_spend_unit = u"小时"
elif play_spend.find(u"分") != -1 or play_spend.find(u"分钟") != -1:
play_spend_unit = u"分"
elif len(play_spend.strip()) == 0:
play_spend_unit = u""
else:
play_spend_unit = u""
replace_dict = {
u"小时": u"",
u"天": u"",
u"时": u"",
u"小": u"",
u"分钟": u"",
u"分": u"",
u"钟": u"",
u"至": u"-",
u"超过": u">",
u"半": u".5",
u"一": u"1",
u"二": u"2",
u"三": u"3",
u"四": u"4",
u"五": u"5",
u"六": u"6",
u"七": u"7",
u"八": u"8",
u"九": u"9",
}
for key, value in replace_dict.iteritems():
play_spend = play_spend.replace(key, value)
if play_spend.strip() == u".5":
play_spend = u"0.5"
return play_spend, play_spend_unit
def _extract_preview(tree):
"""用于提取preview（针对两种类型的网页）
Args:
tree: ElementTree
Returns:
previe: unicode,
"""
# 第一类型网页
preview_elems = tree.xpath("//div[@class='sidebar-mod-inner']/"
"article/div[@class='']/p")
preview_list = []
for preview_elem in preview_elems:
preview_list.append(preview_elem.text)
preview = "".join(preview_list)
# 第二类型网页
if len(preview) == 0:
preview = flist(tree.xpath("//div[@class='view-mod-desc-main']"
"/div[@id='view-mod-abstract']/"
"div[@class='desc-all-holder']"
"/text()"), u"")
return preview
def _extract_preview(tree):
"""用于提取preview（针对两种类型的网页）
Args:
tree: ElementTree
Returns:
previe: unicode,
"""
# 第一类型网页
preview_elems = tree.xpath("//div[@class='sidebar-mod-inner']/"
"article/div[@class='']/p")
preview_list = []
for preview_elem in preview_elems:
preview_list.append(preview_elem.text)
preview = "".join(preview_list)
# 第二类型网页
if len(preview) == 0:
preview = flist(
tree.xpath("//div[@class='view-mod-desc-main']"
"/div[@id='view-mod-abstract']/"
"div[@class='desc-all-holder']"
"/text()"), u"")
return preview
def parse(self, task, input_file):
"""解析函数
Args:
task: HttpTask，任务
input_file: StringIO, 网页文件
Yields:
task: HttpTask, 新任务
"""
tree = html.parse(input_file)
attraction_elems = tree.xpath(
"//div[@id='ATTRACTION_OVERVIEW']"
"/div[@class='attraction-list clearfix']")
for attraction_elem in attraction_elems:
try:
info_elem = flist(
attraction_elem.xpath(
"div[@class='clearfix']/div[@class='info']"), None)
rank_elem = flist(
attraction_elem.xpath(
"div[@class='clearfix']/div[@class='rank']"), None)
relate_path = flist(
info_elem.xpath("div[@class='title']/a/@href"), u"")
name = flist(info_elem.xpath("div[@class='title']/a/text()"),
u"")
address = _extract_address(info_elem)
hot = flist(rank_elem.xpath("a/strong/text()"), u"")
rank = flist(rank_elem.xpath("span[1]/strong/text()"), u"")
# 形成attraction 任务
http_request = build_attraction_request(relate_path)
attraction_task = HttpTask(http_request,
callback="AttractionParser",
max_fail_count=3,
cookie_host=LVYOU_HOST,
kwargs={
"name": unicode(name).strip(),
"address": unicode(address),
"hot": unicode(hot),
"rank": unicode(rank)
})
yield attraction_task
except Exception, e:
self.logger.warn("extract one attraction failed error:%s" % e)
def parse(self, task, input_file):
"""解析函数
Args:
task: HttpTask，任务
input_file: StringIO, 网页文件
Yields:
task: HttpTask, 新任务
"""
tree = html.parse(input_file)
attraction_elems = tree.xpath("//div[@id='ATTRACTION_OVERVIEW']"
"/div[@class='attraction-list clearfix']")
for attraction_elem in attraction_elems:
try:
info_elem = flist(attraction_elem.xpath(
"div[@class='clearfix']/div[@class='info']"), None)
rank_elem = flist(attraction_elem.xpath(
"div[@class='clearfix']/div[@class='rank']"), None)
relate_path = flist(info_elem.xpath(
"div[@class='title']/a/@href"), u"")
name = flist(info_elem.xpath(
"div[@class='title']/a/text()"), u"")
address = _extract_address(info_elem)
hot = flist(rank_elem.xpath("a/strong/text()"), u"")
rank = flist(rank_elem.xpath("span[1]/strong/text()"), u"")
# 形成attraction 任务
http_request = build_attraction_request(relate_path)
attraction_task = HttpTask(
http_request, callback="AttractionParser", max_fail_count=3,
cookie_host=LVYOU_HOST, kwargs={"name": unicode(name).strip(),
"address": unicode(
address),
"hot": unicode(hot),
"rank": unicode(rank)})
yield attraction_task
except Exception, e:
self.logger.warn("extract one attraction failed error:%s" % e)
def parse(self, task, input_file):
"""解析函数
Args:
task: HttpTask，任务
input_file: StringIO, 网页文件
Yields:
item
"""
self.logger.debug("attraction parser start to parse")
content = input_file.read()
tree = html.parse(StringIO(content))
try:
zip_code = flist(tree.xpath("//span[@class='postal-code']/text()"),
u"")
play_spend, play_spend_unit = _extract_play_spend_and_unit(content)
tel_phone = flist(
tree.xpath("//div[@id='HEADING_GROUP']"
"/div[@class='wrap infoBox']"
"/div[@class='odcHotel blDetails']"
"/div/div[@class='fl']/text()"), u"")
open_time = u""
total_score = flist(
tree.xpath("//div[@class='rs rating']"
"/span/img/@content"))
ticket_info = u""
preview_relate_path = flist(
tree.xpath("//div[@class='listing_description']/a/@href"), u"")
lon, lat = _extract_lon_lat(
flist(
tree.xpath("//div[@class='js_mapThumb']"
"/div[@id='bmapContainer']/img[1]/@src"), u""))
comments = _extract_comments(tree)
# 生成景点信息(不包括description)
attraction_item = AttractionItem(task.request.url,
task.kwargs['name'],
unicode(play_spend),
play_spend_unit,
task.kwargs['address'],
unicode(tel_phone),
unicode(open_time),
unicode(total_score),
unicode(ticket_info),
task.kwargs['hot'], lon, lat,
task.kwargs['rank'], comments,
unicode(zip_code))
yield attraction_item
# 生成description任务
if len(preview_relate_path) != 0:
description_request = build_description_request(
task.request.url, preview_relate_path)
description_task = HttpTask(description_request,
callback="DescriptionParser",
max_fail_count=3,
cookie_host=LVYOU_HOST,
kwargs={'url': task.request.url})
yield description_task
else:
yield DescriptionItem(task.request.url, u"")
except Exception, e:
print "error:%s" % e
print traceback.format_exc()
class AttractionListParser(BaseParser):
"""用于解析Attraction list的解析器
"""
def __init__(self, namespace):
BaseParser.__init__(self, namespace)
self.logger.info("init attraction list parser finish")
def parse(self, task, input_file):
"""解析函数
Args:
task: HttpTask，任务
input_file: StringIO, 网页文件
Yields:
task: HttpTask, 新任务
"""
tree = html.parse(input_file)
attraction_elems = tree.xpath(
"//div[@id='ATTRACTION_OVERVIEW']"
"/div[@class='attraction-list clearfix']")
for attraction_elem in attraction_elems:
try:
info_elem = flist(
attraction_elem.xpath(
"div[@class='clearfix']/div[@class='info']"), None)
rank_elem = flist(
attraction_elem.xpath(
"div[@class='clearfix']/div[@class='rank']"), None)
relate_path = flist(
info_elem.xpath("div[@class='title']/a/@href"), u"")
name = flist(info_elem.xpath("div[@class='title']/a/text()"),
u"")
address = _extract_address(info_elem)
hot = flist(rank_elem.xpath("a/strong/text()"), u"")
rank = flist(rank_elem.xpath("span[1]/strong/text()"), u"")
# 形成attraction 任务
http_request = build_attraction_request(relate_path)
attraction_task = HttpTask(http_request,
callback="AttractionParser",
max_fail_count=3,
cookie_host=LVYOU_HOST,
kwargs={
"name": unicode(name).strip(),
"address": unicode(address),
"hot": unicode(hot),
"rank": unicode(rank)
})
yield attraction_task
except Exception, e:
self.logger.warn("extract one attraction failed error:%s" % e)
# 形成下一页任务
next_page_relate = flist(
tree.xpath(
"//div[@class='pagination']/div"
"/a[@class='next sprite-arrow-right-green ml6 ']/@href"), u"")
if len(next_page_relate) != 0:
next_page_request = build_next_page_request(next_page_relate)
next_page_task = HttpTask(next_page_request,
callback="AttractionListParser",
max_fail_count=5,
cookie_host=LVYOU_HOST)
yield next_page_task
def _extract_comments(tree):
"""解析出评论
Args:
tree: ElementTree
Returns:
comments: list
"""
comment_elems = tree.xpath(
"//div[@id='REVIEWS']/div[@class='reviewSelector']")
comments = []
for comment_elem in comment_elems:
try:
# 第一类型网页
comment_user = flist(
comment_elem.xpath("div/div[@class='col1of2']"
"//div[@class='username mo']/span/text()"),
u"")
comment_score = flist(
comment_elem.xpath(
"div/div[@class='col2of2 ']"
"/div[@class='rating reviewItemInline']"
"/span[@class='rate rate_s s50']/img/@content"), u"")
comment_time = flist(
comment_elem.xpath("div/div[@class='col2of2 ']"
"/div[@class='rating reviewItemInline']"
"/span[@class='ratingDate']/text()"), u"")
content_elems = comment_elem.xpath(
"div/div[@class='col2of2 ']/div[@class='entry']/p")
texts = []
for content_elem in content_elems:
if content_elem.text is not None:
texts.append(content_elem.text)
comment_content = u"".join(texts).strip()
# 第二类型网页
if len(comment_user.strip()) == 0:
comment_user = flist(
comment_elem.xpath("div/div[@class='col1of2']"
"//div[@class='username']/span/text()"),
u"")
if len(comment_time.strip()) == 0:
comment_time = flist(
comment_elem.xpath("div/div[@class='col2of2']"
"/span[@class='ratingDate']/text()"),
u"")
if len(comment_content.strip()) == 0:
content_elems = comment_elem.xpath("div/div[@class='col2of2']"
"/div[@class='entry']/p")
texts = []
for content_elem in content_elems:
if content_elem.text is not None:
texts.append(content_elem.text)
comment_content = u"".join(texts).strip()
except Exception, e:
print "extract one comment failed error:%s" % e
print traceback.format_exc()
else:
if len(unicode(comment_content)) != 0:
comment_item = CommentItem(
unicode(comment_user).strip(),
unicode(comment_time).strip(),
unicode(comment_score).strip(),
unicode(comment_content).strip())
comments.append(comment_item)
def parse(self, task, input_file):
"""parse method
Args:
task: Task, task
input_file: file: file with content
Yields:
item: Item, result of parse
task: Task, new task
"""
self.logger.debug("room parser begin to parse")
try:
try:
soap_tree = etree.fromstring(input_file.read())
except Exception, e:
self.logger.error("not complete xml:%s" % e)
raise ParserError("not complete xml")
hotel_address_dict = task.kwargs.get('address')
soap_elems = xpath_namespace(
soap_tree, "/soap:Envelope/soap:Body/"
"RequestResponse/RequestResult")
xml_str = soap_elems[0].text
tree = etree.fromstring(xml_str)
elems = tree.xpath("/Response/Header")
header = elems[0]
if "ResultCode" not in header.attrib or \
header.attrib['ResultCode'] != "Success":
self.logger.error("not has resultcode or "
"resultcode is not success")
raise ParserError("ResultCode error")
else:
content_elems = xpath_namespace(
tree, "/Response/HotelResponse/"
"OTA_HotelDescriptiveInfoRS/"
"HotelDescriptiveContents/"
"HotelDescriptiveContent")
for content_elem in content_elems:
item_hotel_code = None
item_hotel_city_code = task.kwargs.get('citycode')
try:
item_hotel_code = content_elem.attrib.get('HotelCode')
item_hotel_name = content_elem.attrib.get('HotelName')
item_hotel_brand_id = \
content_elem.attrib.get('BrandCode')
position_elem = flist(
xpath_namespace(content_elem,
"HotelInfo//Position"), None)
item_hotel_latitude = "" if position_elem is None \
or "Latitude" not in position_elem.attrib \
else position_elem.attrib.get('Latitude')
item_hotel_longitude = "" if position_elem is None \
or "Longitude" not in position_elem.attrib \
else position_elem.attrib.get('Longitude')
service_elems = xpath_namespace(
content_elem, "HotelInfo/Services/Service")
item_hotel_service = u"、".join([
flist(
service.xpath(
"*[local-name()='DescriptiveText']/text()")
) for service in service_elems
if "Code" in service.attrib
and service.attrib["Code"] in HOTEL_SERVICE_CODES
])
item_room_service = u"、".join([
flist(
service.xpath(
"*[local-name()='DescriptiveText']/text()")
) for service in service_elems
if "Code" in service.attrib
and service.attrib["Code"] in ROOM_SERVICE_CODES
])
awards_elem = flist(
xpath_namespace(content_elem,
"AffiliationInfo/Awards"), None)
item_hotel_star, item_hotel_rate = ("", "") \
if awards_elem is None else \
(flist(awards_elem.xpath(
"*[local-name()='Award' and "
"@Provider='HotelStarRate']/@Rating")),
flist(awards_elem.xpath(
"*[local-name()='Award' and "
"@Provider='CtripStarRate']/@Rating")))
multimedia_elem = flist(
xpath_namespace(content_elem,
"MultimediaDescriptions"), None)
image_elems = [] if multimedia_elem is None \
else xpath_namespace(
multimedia_elem,
"MultimediaDescription/ImageItems/ImageItem")
item_image_list = []
for image_elem in image_elems:
image_url = flist(
image_elem.xpath(
"*[local-name()='ImageFormat']/"
"*[local-name()='URL']/text()"))
image_type = flist(image_elem.xpath("@Category"))
if not image_url and not image_type:
continue
image_text = flist(
image_elem.xpath(
"*[local-name()='Description']/@Caption"))
item_image_dict = {
"image_url": image_url,
"image_type": image_type,
"image_text": image_text.encode('utf-8')
}
item_image_list.append(item_image_dict)
if item_hotel_code and image_url:
image_item = ImageItem(item_hotel_code,
str(image_type),
unicode(image_text),
str(image_url))
yield image_item
text_items_elem = flist(
xpath_namespace(multimedia_elem,
"MultimediaDescription/TextItems"),
None)
item_hotel_preview = "" if text_items_elem is None \
else flist(text_items_elem.xpath(
"*[local-name()='TextItem' and @Category='5']/"
"*[local-name()='Description']/text()"))
room_elems = xpath_namespace(
content_elem, "FacilityInfo/"
"GuestRooms/GuestRoom")
item_room_list = []
for room_elem in room_elems:
room_info_id = flist(
room_elem.xpath(
"*[local-name()='TypeRoom']/@RoomTypeCode")
)
room_info_name = flist(
room_elem.xpath("@RoomTypeName"))
room_bed_type = flist(
room_elem.xpath(
"*[local-name()='TypeRoom']/@BedTypeCode"))
room_net_service, room_net_service_fee = \
_extract_net_service(room_elem)
room_info_rate_price = ""
room_hot = ""
room_floor = flist(
room_elem.xpath(
"*[local-name()='TypeRoom']/@Floor"))
room_breakfast = ""
room_area = ""
room_info_dict = {
'roomInfo_id': room_info_id,
'roomInfo_ratePrice': room_info_rate_price,
'hot': room_hot
}
if room_info_id and room_info_name and \
item_hotel_code:
item_room_list.append(room_info_dict)
room_item = RoomInfoItem(
item_hotel_code, str(room_info_id),
room_info_name, room_floor,
room_net_service, room_net_service_fee,
room_bed_type, room_breakfast, room_area)
yield room_item
item_hotel_address = "" if item_hotel_code not in \
hotel_address_dict \
else hotel_address_dict.get(item_hotel_code)
hotel_item = HotelInfoItem(
item_hotel_code, item_hotel_city_code,
item_hotel_name, item_hotel_brand_id,
item_hotel_latitude, item_hotel_longitude,
item_hotel_service, item_room_service,
item_hotel_star, item_hotel_rate, item_image_list,
item_hotel_preview, item_room_list,
item_hotel_address)
yield hotel_item
except Exception, e:
self.logger.warn("one hotel extract error:%s" % e)
if item_hotel_code is None:
self.logger.error("i am sorry, i can do noting")
else:
chinese_name = task.kwargs.get('chinesename')
yield build_rooms_task_for_hotel(
[item_hotel_code], item_hotel_city_code,
chinese_name, hotel_address_dict)
def parse(self, task, input_file):
"""parse response result
Args:
task: FileTask or HttpTask
input_file: file or StringIO
Yields:
item: Item, result of parse
task: Task, new task
"""
self.logger.info("hotel parser begin to parse")
try:
try:
soap_tree = etree.fromstring(input_file.read())
except Exception, e:
self.logger.error("not complete xml:%s" % e)
raise ParserError("not complete xml")
soap_elems = xpath_namespace(soap_tree,
"/soap:Envelope/soap:Body/"
"RequestResponse/RequestResult")
xml_str = soap_elems[0].text
tree = etree.fromstring(xml_str)
elems = tree.xpath("/Response/Header")
header = elems[0]
if "ResultCode" not in header.attrib or \
header.attrib['ResultCode'] != "Success":
self.logger.error("not has resultcode or "
"resultcode is not success")
raise ParserError("ResultCode error")
else:
# success
property_elems = xpath_namespace(
tree, "/Response/HotelResponse/OTA_HotelSearchRS/"
"Properties/Property")
city_code = task.kwargs.get('citycode')
chinese_name = task.kwargs.get('chinesename')
hotel_requests = list()
hotel_addresses = dict()
for property_elem in property_elems:
hotel_code = str(property_elem.attrib['HotelCode']) \
if "HotelCode" in property_elem.attrib \
else None
hotel_ctrip_city_code = str(
property_elem.attrib['HotelCityCode']) \
if "HotelCityCode" in property_elem.attrib else None
hotel_address = flist(property_elem.xpath(
"*[local-name()='Address']/"
"*[local-name()='AddressLine']/text()"))
if isinstance(hotel_address, unicode):
hotel_address = hotel_address.encode("utf-8")
hotel_address = str(hotel_address)
if hotel_code and hotel_ctrip_city_code:
hotel_url = build_hotel_url(hotel_code)
yield HotelCodeItem(hotel_code, city_code, hotel_url)
hotel_requests.append(hotel_code)
hotel_addresses[hotel_code] = hotel_address
if len(hotel_requests) >= self.batch_count:
yield build_rooms_task_for_hotel(hotel_requests,
city_code,
chinese_name,
hotel_addresses)
hotel_addresses.clear()
del hotel_requests[:]
# send left requests
if len(hotel_requests) > 0:
yield build_rooms_task_for_hotel(hotel_requests, city_code,
chinese_name,
hotel_addresses)
hotel_addresses.clear()
del hotel_requests[:]
def parse(self, task, input_file):
"""解析函数
Args:
task: HTTPTask, 任务
input_file: StringIO, 网页信息
Yields:
task: HTTPTask, 任务
item: Item, 解析的结果
"""
self.logger.debug("attraction parser start to parse")
parser = html.HTMLParser(encoding='utf-8')
tree = html.parse(input_file, parser)
try:
name = flist(tree.xpath("//header[@class='title-head']/a/p/text()"), u"")
play_spend, play_spend_unit = _extract_play_spend(tree)
address = flist(tree.xpath("//div[@id='J-aside-info-address']"
"/span[@class='val address-value']"
"/text()"), u"")
tel_phone = flist(tree.xpath("//div[@id='J-aside-info-phone']"
"/span[@class='val phone-value']"
"/text()"), u"")
time_elems = tree.xpath("//div[@id='J-aside-info-opening_hours']"
"/div[@class='val opening_hours-value']/p")
time_list = []
for time_elem in time_elems:
time_list.append(time_elem.text)
open_time = "".join(time_list)
total_score = flist(tree.xpath("//div[@class='scene-rating']"
"/div/@content"), u"")
ticket_info = flist(tree.xpath("//div[@id='J-aside-info-price']"
"/div[@class='val price-value']"
"/p/text()"), u"")
preview = _extract_preview(tree)
traffic = _extract_traffic(tree)
tips = _extract_tips(tree)
hot = flist(tree.xpath("//section[@id='remark-container']"
"/div[@class='remark-overall-rating']"
"/span[@class='remark-all-counts']"
"/text()"), u"")
lon_lat = task.kwargs['map_info'].split(",")
if len(lon_lat) <= 1:
lon, lat = u"", u""
else:
lon, lat = lon_lat[0], lon_lat[1]
seq_sort = task.kwargs['seq_sort']
sid = task.kwargs['sid']
attraction_item = AttractionItem(unicode(sid), unicode(name),
unicode(play_spend),
unicode(play_spend_unit),
unicode(address),
unicode(tel_phone),
unicode(open_time),
unicode(total_score),
unicode(ticket_info),
unicode(preview), unicode(hot),
unicode(lon), unicode(lat),
unicode(seq_sort),
unicode(traffic), unicode(tips))
yield attraction_item
# yield comment list task
comments_request = build_comment_list_request(
sid, task.kwargs['relate_path'])
comments_task = HttpTask(
comments_request, callback="CommentListParser",
max_fail_count=3,
cookie_host=LVYOU_HOST, kwargs={'sid': sid})
yield comments_task
except Exception, e:
self.logger.error("extract Attraction failed error:%s" % e)
self.logger.error("error traceback:%s" % traceback.format_exc())
raise e
def parse(self, task, input_file):
"""parse response result
Args:
task: FileTask or HttpTask
input_file: file or StringIO
Yields:
item: Item, result of parse
task: Task, new task
"""
self.logger.info("hotel parser begin to parse")
try:
try:
soap_tree = etree.fromstring(input_file.read())
except Exception, e:
self.logger.error("not complete xml:%s" % e)
raise ParserError("not complete xml")
soap_elems = xpath_namespace(
soap_tree, "/soap:Envelope/soap:Body/"
"RequestResponse/RequestResult")
xml_str = soap_elems[0].text
tree = etree.fromstring(xml_str)
elems = tree.xpath("/Response/Header")
header = elems[0]
if "ResultCode" not in header.attrib or \
header.attrib['ResultCode'] != "Success":
self.logger.error("not has resultcode or "
"resultcode is not success")
raise ParserError("ResultCode error")
else:
# success
property_elems = xpath_namespace(
tree, "/Response/HotelResponse/OTA_HotelSearchRS/"
"Properties/Property")
city_code = task.kwargs.get('citycode')
chinese_name = task.kwargs.get('chinesename')
hotel_requests = list()
hotel_addresses = dict()
for property_elem in property_elems:
hotel_code = str(property_elem.attrib['HotelCode']) \
if "HotelCode" in property_elem.attrib \
else None
hotel_ctrip_city_code = str(
property_elem.attrib['HotelCityCode']) \
if "HotelCityCode" in property_elem.attrib else None
hotel_address = flist(
property_elem.xpath(
"*[local-name()='Address']/"
"*[local-name()='AddressLine']/text()"))
if isinstance(hotel_address, unicode):
hotel_address = hotel_address.encode("utf-8")
hotel_address = str(hotel_address)
if hotel_code and hotel_ctrip_city_code:
hotel_url = build_hotel_url(hotel_code)
yield HotelCodeItem(hotel_code, city_code, hotel_url)
hotel_requests.append(hotel_code)
hotel_addresses[hotel_code] = hotel_address
if len(hotel_requests) >= self.batch_count:
yield build_rooms_task_for_hotel(
hotel_requests, city_code, chinese_name,
hotel_addresses)
hotel_addresses.clear()
del hotel_requests[:]
# send left requests
if len(hotel_requests) > 0:
yield build_rooms_task_for_hotel(hotel_requests, city_code,
chinese_name,
hotel_addresses)
hotel_addresses.clear()
del hotel_requests[:]
def parse(self, task, input_file):
"""解析函数
Args:
task: HTTPTask, 任务
input_file: StringIO, 网页信息
Yields:
task: HTTPTask, 任务
item: Item, 解析的结果
"""
self.logger.debug("attraction parser start to parse")
parser = html.HTMLParser(encoding='utf-8')
tree = html.parse(input_file, parser)
try:
name = flist(
tree.xpath("//header[@class='title-head']/a/p/text()"), u"")
play_spend, play_spend_unit = _extract_play_spend(tree)
address = flist(
tree.xpath("//div[@id='J-aside-info-address']"
"/span[@class='val address-value']"
"/text()"), u"")
tel_phone = flist(
tree.xpath("//div[@id='J-aside-info-phone']"
"/span[@class='val phone-value']"
"/text()"), u"")
time_elems = tree.xpath("//div[@id='J-aside-info-opening_hours']"
"/div[@class='val opening_hours-value']/p")
time_list = []
for time_elem in time_elems:
time_list.append(time_elem.text)
open_time = "".join(time_list)
total_score = flist(
tree.xpath("//div[@class='scene-rating']"
"/div/@content"), u"")
ticket_info = flist(
tree.xpath("//div[@id='J-aside-info-price']"
"/div[@class='val price-value']"
"/p/text()"), u"")
preview = _extract_preview(tree)
traffic = _extract_traffic(tree)
tips = _extract_tips(tree)
hot = flist(
tree.xpath("//section[@id='remark-container']"
"/div[@class='remark-overall-rating']"
"/span[@class='remark-all-counts']"
"/text()"), u"")
lon_lat = task.kwargs['map_info'].split(",")
if len(lon_lat) <= 1:
lon, lat = u"", u""
else:
lon, lat = lon_lat[0], lon_lat[1]
seq_sort = task.kwargs['seq_sort']
sid = task.kwargs['sid']
attraction_item = AttractionItem(unicode(sid), unicode(name),
unicode(play_spend),
unicode(play_spend_unit),
unicode(address),
unicode(tel_phone),
unicode(open_time),
unicode(total_score),
unicode(ticket_info),
unicode(preview), unicode(hot),
unicode(lon), unicode(lat),
unicode(seq_sort),
unicode(traffic), unicode(tips))
yield attraction_item
# yield comment list task
comments_request = build_comment_list_request(
sid, task.kwargs['relate_path'])
comments_task = HttpTask(comments_request,
callback="CommentListParser",
max_fail_count=3,
cookie_host=LVYOU_HOST,
kwargs={'sid': sid})
yield comments_task
except Exception, e:
self.logger.error("extract Attraction failed error:%s" % e)
self.logger.error("error traceback:%s" % traceback.format_exc())
raise e
def parse(self, task, input_file):
"""parse method
Args:
task: Task, task
input_file: file: file with content
Yields:
item: Item, result of parse
task: Task, new task
"""
self.logger.debug("room parser begin to parse")
try:
try:
soap_tree = etree.fromstring(input_file.read())
except Exception, e:
self.logger.error("not complete xml:%s" % e)
raise ParserError("not complete xml")
hotel_address_dict = task.kwargs.get('address')
soap_elems = xpath_namespace(soap_tree,
"/soap:Envelope/soap:Body/"
"RequestResponse/RequestResult")
xml_str = soap_elems[0].text
tree = etree.fromstring(xml_str)
elems = tree.xpath("/Response/Header")
header = elems[0]
if "ResultCode" not in header.attrib or \
header.attrib['ResultCode'] != "Success":
self.logger.error("not has resultcode or "
"resultcode is not success")
raise ParserError("ResultCode error")
else:
content_elems = xpath_namespace(tree,
"/Response/HotelResponse/"
"OTA_HotelDescriptiveInfoRS/"
"HotelDescriptiveContents/"
"HotelDescriptiveContent")
for content_elem in content_elems:
item_hotel_code = None
item_hotel_city_code = task.kwargs.get('citycode')
try:
item_hotel_code = content_elem.attrib.get('HotelCode')
item_hotel_name = content_elem.attrib.get('HotelName')
item_hotel_brand_id = \
content_elem.attrib.get('BrandCode')
position_elem = flist(
xpath_namespace(content_elem,
"HotelInfo//Position"), None)
item_hotel_latitude = "" if position_elem is None \
or "Latitude" not in position_elem.attrib \
else position_elem.attrib.get('Latitude')
item_hotel_longitude = "" if position_elem is None \
or "Longitude" not in position_elem.attrib \
else position_elem.attrib.get('Longitude')
service_elems = xpath_namespace(
content_elem, "HotelInfo/Services/Service")
item_hotel_service = u"、".join(
[flist(service.xpath(
"*[local-name()='DescriptiveText']/text()"))
for service in service_elems
if "Code" in service.attrib and
service.attrib["Code"] in HOTEL_SERVICE_CODES])
item_room_service = u"、".join(
[flist(service.xpath(
"*[local-name()='DescriptiveText']/text()"))
for service in service_elems
if "Code" in service.attrib and
service.attrib["Code"] in ROOM_SERVICE_CODES])
awards_elem = flist(
xpath_namespace(content_elem,
"AffiliationInfo/Awards"),
None)
item_hotel_star, item_hotel_rate = ("", "") \
if awards_elem is None else \
(flist(awards_elem.xpath(
"*[local-name()='Award' and "
"@Provider='HotelStarRate']/@Rating")),
flist(awards_elem.xpath(
"*[local-name()='Award' and "
"@Provider='CtripStarRate']/@Rating")))
multimedia_elem = flist(
xpath_namespace(content_elem,
"MultimediaDescriptions"), None)
image_elems = [] if multimedia_elem is None \
else xpath_namespace(
multimedia_elem,
"MultimediaDescription/ImageItems/ImageItem")
item_image_list = []
for image_elem in image_elems:
image_url = flist(image_elem.xpath(
"*[local-name()='ImageFormat']/"
"*[local-name()='URL']/text()"))
image_type = flist(image_elem.xpath("@Category"))
if not image_url and not image_type:
continue
image_text = flist(image_elem.xpath(
"*[local-name()='Description']/@Caption"))
item_image_dict = {"image_url": image_url,
"image_type": image_type,
"image_text":
image_text.encode('utf-8')}
item_image_list.append(item_image_dict)
if item_hotel_code and image_url:
image_item = ImageItem(item_hotel_code,
str(image_type),
unicode(image_text),
str(image_url))
yield image_item
text_items_elem = flist(xpath_namespace(
multimedia_elem, "MultimediaDescription/TextItems"),
None)
item_hotel_preview = "" if text_items_elem is None \
else flist(text_items_elem.xpath(
"*[local-name()='TextItem' and @Category='5']/"
"*[local-name()='Description']/text()"))
room_elems = xpath_namespace(content_elem,
"FacilityInfo/"
"GuestRooms/GuestRoom")
item_room_list = []
for room_elem in room_elems:
room_info_id = flist(room_elem.xpath(
"*[local-name()='TypeRoom']/@RoomTypeCode"))
room_info_name = flist(room_elem.xpath(
"@RoomTypeName"))
room_bed_type = flist(room_elem.xpath(
"*[local-name()='TypeRoom']/@BedTypeCode"))
room_net_service, room_net_service_fee = \
_extract_net_service(room_elem)
room_info_rate_price = ""
room_hot = ""
room_floor = flist(room_elem.xpath(
"*[local-name()='TypeRoom']/@Floor"))
room_breakfast = ""
room_area = ""
room_info_dict = {'roomInfo_id': room_info_id,
'roomInfo_ratePrice':
room_info_rate_price,
'hot': room_hot}
if room_info_id and room_info_name and \
item_hotel_code:
item_room_list.append(room_info_dict)
room_item = RoomInfoItem(item_hotel_code,
str(room_info_id),
room_info_name,
room_floor,
room_net_service,
room_net_service_fee,
room_bed_type,
room_breakfast,
room_area)
yield room_item
item_hotel_address = "" if item_hotel_code not in \
hotel_address_dict \
else hotel_address_dict.get(item_hotel_code)
hotel_item = HotelInfoItem(item_hotel_code,
item_hotel_city_code,
item_hotel_name,
item_hotel_brand_id,
item_hotel_latitude,
item_hotel_longitude,
item_hotel_service,
item_room_service,
item_hotel_star,
item_hotel_rate,
item_image_list,
item_hotel_preview,
item_room_list,
item_hotel_address)
yield hotel_item
except Exception, e:
self.logger.warn("one hotel extract error:%s" % e)
if item_hotel_code is None:
self.logger.error("i am sorry, i can do noting")
else:
chinese_name = task.kwargs.get('chinesename')
yield build_rooms_task_for_hotel(
[item_hotel_code], item_hotel_city_code,
chinese_name, hotel_address_dict)
