def deal_comment(self, req_url, text):
"""
解析评论
:param req_url:
:param text:
:return:
"""
data = tools.get_json(text)
__biz = tools.get_param(req_url, "__biz")
comment_id = tools.get_param(req_url, "comment_id") # 与文章关联
elected_comment = data.get("elected_comment", [])
comment_datas = [
dict(
__biz=__biz,
comment_id=comment_id,
nick_name=comment.get("nick_name"),
logo_url=comment.get("logo_url"),
content=comment.get("content"),
create_time=tools.timestamp_to_date(
comment.get("create_time")),
content_id=comment.get("content_id"),
like_num=comment.get("like_num"),
is_top=comment.get("is_top"),
spider_time=tools.get_current_date(),
) for comment in elected_comment
]
if comment_datas:
data_pipeline.save_article_commnet(comment_datas)
def deal_comment(self, req_url, text):
data = tools.get_json(text)
__biz = tools.get_param(req_url, '__biz')
comment_id = tools.get_param(req_url, 'comment_id') # 与文章关联
elected_comment = data.get('elected_comment', [])
comment_datas = [
dict(
__biz=__biz,
comment_id=comment_id,
nick_name=comment.get('nick_name'),
logo_url=comment.get('logo_url'),
content=comment.get('content'),
create_time=tools.timestamp_to_date(comment.get('create_time')),
content_id=comment.get('content_id'),
like_num=comment.get('like_num'),
is_top=comment.get('is_top'),
spider_time=tools.get_current_date()
)
for comment in elected_comment
]
if comment_datas:
data_pipeline.save_article_commnet(comment_datas)
def deal_article(self, req_url, text):
"""
解析文章
:param req_url:
:param text:
:return:
"""
sn = tools.get_param(req_url, "sn")
if not text:
self._task_manager.update_article_task_state(sn, -1)
return None
selector = Selector(text)
content = selector.xpath(
'//div[@class="rich_media_content "]|//div[@class="rich_media_content"]|//div[@class="share_media"]'
).extract_first(default="")
title = (selector.xpath('//h2[@class="rich_media_title"]/text()').
extract_first(default="").strip())
account = (selector.xpath('//a[@id="js_name"]/text()').extract_first(
default="").strip())
author = (selector.xpath(
'//span[@class="rich_media_meta rich_media_meta_text"]//text()').
extract_first(default="").strip())
publish_timestamp = selector.re_first('n="(\d{10})"')
publish_timestamp = int(
publish_timestamp) if publish_timestamp else None
publish_time = (tools.timestamp_to_date(publish_timestamp)
if publish_timestamp else None)
biz = tools.get_param(req_url, "__biz")
text = remove_tags(content).strip()
spider_name = 'wechat'
collection_mode = 'spider'
data_source_type = '微信公众号'
article_data = {
"data_type": account,
"title": title,
"data_address": req_url,
"author": author,
"publish_time": publish_time,
"__biz": biz,
"text": text,
"spider_name": spider_name,
"collection_mode": collection_mode,
"data_source_type": data_source_type,
"sn": sn,
"collection_time": tools.get_current_date(),
}
# 入库
if article_data and data_pipeline.save_article(
article_data) is not None:
self._task_manager.update_article_task_state(sn, 1)
return self._task_manager.get_task()
def parse_article_info(article_info, release_time):
if not article_info:
return
# log.debug(tools.dumps_json(article_info))
title = article_info.get('title')
summary = article_info.get('digest')
url = article_info.get('content_url').replace('\\', '').replace(
'amp;', '')
source_url = article_info.get('source_url').replace('\\',
'') # 引用的文章链接
cover = article_info.get('cover').replace('\\', '')
author = article_info.get('author')
if url and url.startswith(
'http://mp.weixin.qq.com/'
): # 被发布者删除的文章 无url和其他信息， 此时取不到mid 且不用入库, 或者商城类的url不入库
mid = tools.get_param(url, 'mid') or tools.get_param(
url, 'appmsgid') # 图文消息id 同一天发布的图文消息 id一样
idx = tools.get_param(url, 'idx') or tools.get_param(
url, 'itemidx') # 第几条图文消息 从1开始
article_id = mid + idx # 用mid和idx 拼接 确定唯一一篇文章 如mid = 2650492260 idx = 1，则article_id = 26504922601
# 判断该文章库中是否已存在
if WechatAction._wechat_service.is_exist(
'wechat_article',
article_id) or (ONLY_TODAY_MSG and release_time <
tools.get_current_date('%Y-%m-%d')):
self._is_need_get_more = False
return # 不往下进行 舍弃之后的文章
__biz = tools.get_param(url, '__biz') # 用于关联公众号
# 缓存文章信息
WechatAction._article_info[article_id] = {
'article_id': int(article_id),
'title': title,
'summary': summary,
'release_time': release_time,
'url': url,
'source_url': source_url,
'cover': cover,
'account': '',
'author': author,
'__biz': __biz,
'read_num': None,
'like_num': None,
'content': '',
'comment': [],
'record_time': tools.get_current_date()
}
# 将文章url添加到待抓取队列
WechatAction._todo_urls.append(url)
def get_read_watched_count(self, data, req_url):
'''
@summary:
---------
@param data:
{
"advertisement_num":0,
"advertisement_info":[
],
"appmsgstat":{
"show":true,
"is_login":true,
"liked":false,
"read_num":38785,
"like_num":99,
"ret":0,
"real_read_num":0
},
"comment_enabled":1,
"reward_head_imgs":[
],
"only_fans_can_comment":false,
"is_ios_reward_open":0,
"base_resp":{
"wxtoken":3465907592
}
}
@param req_url:
---------
@result:
'''
log.debug('获取观看和点赞量')
req_url = req_url.replace('amp;', '')
mid = tools.get_param(req_url, 'mid') # 图文消息id 同一天发布的图文消息 id一样
idx = tools.get_param(req_url, 'idx') # 第几条图文消息 从1开始
article_id = mid + idx # 用mid和idx 拼接 确定唯一一篇文章 如mid = 2650492260 idx = 1，则article_id = 26504922601
data = tools.get_json(data)
read_num = data.get('appmsgstat', {}).get('read_num')
like_num = data.get('appmsgstat', {}).get('like_num')
# 缓存文章阅读量点赞量
WechatAction._article_info[article_id]['read_num'] = read_num
WechatAction._article_info[article_id]['like_num'] = like_num
if not data.get('comment_enabled'): # 无评论区，不请求get_comment 函数，此时直接入库
self._wechat_service.add_article_info(
WechatAction._article_info.pop(article_id))
def deal_article(self, req_url, text):
sn = tools.get_param(req_url, 'sn')
if not text:
self._task_manager.update_article_task_state(sn, -1)
return None
selector = Selector(text)
content = selector.xpath('//div[@class="rich_media_content "]|//div[@class="rich_media_content"]|//div[@class="share_media"]')
title = selector.xpath('//h2[@class="rich_media_title"]/text()').extract_first(default='').strip()
account = selector.xpath('//a[@id="js_name"]/text()').extract_first(default='').strip()
author = selector.xpath('//span[@class="rich_media_meta rich_media_meta_text"]//text()').extract_first(default='').strip()
publish_timestamp = selector.re_first('n="(\d{10})"')
publish_timestamp = int(publish_timestamp) if publish_timestamp else None
publish_time = tools.timestamp_to_date(publish_timestamp) if publish_timestamp else None
pics_url = content.xpath('.//img/@src|.//img/@data-src').extract()
biz = tools.get_param(req_url, '__biz')
digest = selector.re_first('var msg_desc = "(.*?)"')
cover = selector.re_first('var cover = "(.*?)";') or selector.re_first('msg_cdn_url = "(.*?)"')
source_url = selector.re_first("var msg_source_url = '(.*?)';")
content_html = content.extract_first(default='')
comment_id = selector.re_first('var comment_id = "(\d+)"')
article_data = {
'account': account,
'title': title,
'url': req_url,
'author': author,
'publish_time': publish_time,
'__biz': biz,
'digest': digest,
'cover': cover,
"pics_url": pics_url,
"content_html": content_html,
"source_url": source_url,
"comment_id": comment_id,
"sn": sn,
"spider_time": tools.get_current_date()
}
# 入库
if article_data and data_pipeline.save_article(article_data) is not None:
self._task_manager.update_article_task_state(sn, 1)
return self._task_manager.get_task()
def get_article_content(self, data, req_url):
log.debug('获取文章内容')
if data: # 被验证不详实的文章 首次不反回内容，跳转到https://mp.weixin.qq.com/mp/rumor
req_url = req_url.replace('amp;', '')
mid = tools.get_param(req_url, 'mid') or tools.get_param(
req_url, 'appmsgid') # 图文消息id 同一天发布的图文消息 id一样
idx = tools.get_param(req_url, 'idx') or tools.get_param(
req_url, 'itemidx') # 第几条图文消息 从1开始
article_id = mid + idx # 用mid和idx 拼接 确定唯一一篇文章 如mid = 2650492260 idx = 1，则article_id = 26504922601
WechatAction._current_aritcle_id = article_id # 记录当前文章的id 为获取评论信息时找对应的文章id使用
print('当前id' + WechatAction._current_aritcle_id)
regex = '(<div class="rich_media_content ".*?)<script nonce'
content = tools.get_info(data, regex, fetch_one=True)
if content:
# 缓存文章内容
WechatAction._article_info[article_id]['content'] = content
# 取公众号名
regex = '<title>(.*?)</title>'
account = tools.get_info(data, regex, fetch_one=True)
WechatAction._article_info[article_id]['account'] = account
else: # 被验证不实的文章，不会请求观看点赞数，此时直接入库
regex = '<title>(.*?)</title>'
content = tools.get_info(data, regex, fetch_one=True)
WechatAction._article_info[article_id]['content'] = content
# 入库
print('被验证不实的文章，不会请求观看点赞数，此时直接入库')
WechatAction._wechat_service.add_article_info(
WechatAction._article_info.pop(article_id))
# 如果下一页是文章列表的链接， 替换文章列表中的appmsg_token,防止列表链接过期
if (len(WechatAction._todo_urls)
== 1) and ('/mp/profile_ext'
in WechatAction._todo_urls[-1]):
regex = 'appmsg_token = "(.*?)"'
appmsg_token = tools.get_info(data, regex,
fetch_one=True).strip()
WechatAction._todo_urls[-1] = tools.replace_str(
WechatAction._todo_urls[-1], 'appmsg_token=.*?&',
'appmsg_token=%s&' % appmsg_token)
return self.__open_next_page()
else:
# 无文章内容
pass
def get_comment(self, data, req_url):
log.debug('获取评论信息')
req_url = req_url.replace('amp;', '')
mid = tools.get_param(req_url, 'appmsgid') # 图文消息id 同一天发布的图文消息 id一样
idx = tools.get_param(req_url, 'idx') # 第几条图文消息 从1开始
article_id = mid + idx # 用mid和idx 拼接 确定唯一一篇文章 如mid = 2650492260 idx = 1，则article_id = 26504922601
data = tools.get_json(data)
comment = data.get('elected_comment', []) # 精选留言
# 缓存文章评论信息
WechatAction._article_info[article_id]['comment'] = comment
WechatAction._wechat_service.add_article_info(WechatAction._article_info.pop(article_id))
def __parse_account_info(self, data, req_url):
__biz = tools.get_param(req_url, "__biz")
regex = 'id="nickname">(.*?)</strong>'
account = tools.get_info(data, regex, fetch_one=True).strip()
regex = 'profile_avatar">.*?<img src="(.*?)"'
head_url = tools.get_info(data, regex, fetch_one=True)
regex = 'class="profile_desc">(.*?)</p>'
summary = tools.get_info(data, regex, fetch_one=True).strip()
# 认证信息（关注的账号直接点击查看历史消息，无认证信息）
regex = '<i class="icon_verify success">.*?</i>(.*?)</span>'
verify = tools.get_info(data, regex, fetch_one=True)
verify = verify.strip() if verify else ""
# 二维码
regex = 'var username = "" \|\| "(.*?)";' # || 需要转译
qr_code = tools.get_info(data, regex, fetch_one=True)
qr_code = "http://open.weixin.qq.com/qr/code?username="******"__biz": __biz,
"account": account,
"head_url": head_url,
"summary": summary,
"qr_code": qr_code,
"verify": verify,
"spider_time": tools.get_current_date(),
}
if account_data:
data_pipeline.save_account(account_data)
def deal_article_dynamic_info(self, req_data, text):
"""
取文章动态信息 阅读 点赞 评论
:param req_data: post 请求的data str格式
:param text:
:return:
"""
data = tools.get_json(text)
dynamic_data = dict(
sn=tools.get_param(req_data, "sn"),
__biz=tools.get_param(req_data, "__biz").replace("%3D", "="),
read_num=data.get("appmsgstat", {}).get("read_num"),
like_num=data.get("appmsgstat", {}).get("like_num"),
comment_count=data.get("comment_count"),
spider_time=tools.get_current_date(),
)
if dynamic_data:
data_pipeline.save_article_dynamic(dynamic_data)
def deal_article_dynamic_info(self, req_data, text):
"""
取文章动态信息 阅读 点赞 评论
:param req_data: post 请求的data str格式
:param text:
:return:
"""
data = tools.get_json(text)
dynamic_data = dict(
sn=tools.get_param(req_data, 'sn'),
__biz=tools.get_param(req_data, '__biz').replace('%3D', '='),
read_num=data.get('appmsgstat', {}).get('read_num'),
like_num=data.get('appmsgstat', {}).get('like_num'),
comment_count=data.get('comment_count'),
spider_time=tools.get_current_date()
)
if dynamic_data:
data_pipeline.save_article_dynamic(dynamic_data)
def __parse_account_info(self, data, req_url):
'''
@summary:
---------
@param data:
---------
@result:
'''
__biz = tools.get_param(req_url, '__biz')
WechatAction._current_account_biz = __biz
regex = 'id="nickname">(.*?)</strong>'
account = tools.get_info(data, regex, fetch_one=True).strip()
regex = 'profile_avatar">.*?<img src="(.*?)"'
head_url = tools.get_info(data, regex, fetch_one=True)
regex = 'class="profile_desc">(.*?)</p>'
summary = tools.get_info(data, regex, fetch_one=True).strip()
# 认证信息（关注的账号直接点击查看历史消息，无认证信息）
regex = '<i class="icon_verify success">.*?</i>(.*?)</span>'
verify = tools.get_info(data, regex, fetch_one=True)
verify = verify.strip() if verify else ''
# 二维码
regex = 'var username = "" \|\| "(.*?)";' # || 需要转译
qr_code = tools.get_info(data, regex, fetch_one=True)
qr_code = 'http://open.weixin.qq.com/qr/code?username='******'__biz':
__biz,
'account':
account,
'head_url':
head_url,
'summary':
summary,
'qr_code':
qr_code,
'verify':
verify,
'account_id':
WechatAction._account_info.pop(__biz)
if __biz in WechatAction._account_info.keys() else '',
'record_time':
tools.get_current_date()
}
if not WechatAction._wechat_service.is_exist('wechat_account', __biz):
WechatAction._wechat_service.add_account_info(account_info)
def parse_article_info(article_info, comm_msg_info):
if not article_info:
return
# log.debug(tools.dumps_json(article_info))
title = article_info.get("title")
digest = article_info.get("digest")
url = article_info.get("content_url").replace("\\", "").replace(
"amp;", "")
source_url = article_info.get("source_url").replace("\\",
"") # 引用的文章链接
cover = article_info.get("cover").replace("\\", "")
subtype = article_info.get("subtype")
is_multi = article_info.get("is_multi")
author = article_info.get("author")
copyright_stat = article_info.get("copyright_stat")
duration = article_info.get("duration")
del_flag = article_info.get("del_flag")
type = comm_msg_info.get("type")
publish_time = tools.timestamp_to_date(
comm_msg_info.get("datetime"))
sn = tools.get_param(url, "sn")
if sn:
# 缓存文章信息
article_data = {
"title": title,
"digest": digest,
"url": url,
"source_url": source_url,
"cover": cover,
"subtype": subtype,
"is_multi": is_multi,
"author": author,
"copyright_stat": copyright_stat,
"duration": duration,
"del_flag": del_flag,
"type": type,
"publish_time": publish_time,
"sn": sn,
"__biz": __biz,
"spider_time": tools.get_current_date(),
}
return article_data
def parse_article_info(article_info, comm_msg_info):
if not article_info:
return
# log.debug(tools.dumps_json(article_info))
title = article_info.get('title')
digest = article_info.get('digest')
url = article_info.get('content_url').replace('\\', '').replace('amp;', '')
source_url = article_info.get('source_url').replace('\\', '') # 引用的文章链接
cover = article_info.get('cover').replace('\\', '')
subtype = article_info.get('subtype')
is_multi = article_info.get('is_multi')
author = article_info.get('author')
copyright_stat = article_info.get('copyright_stat')
duration = article_info.get('duration')
del_flag = article_info.get('del_flag')
type = comm_msg_info.get('type')
publish_time = tools.timestamp_to_date(comm_msg_info.get('datetime'))
sn = tools.get_param(url, 'sn')
if sn:
# 缓存文章信息
article_data = {
'title': title,
'digest': digest,
'url': url,
'source_url': source_url,
'cover': cover,
'subtype': subtype,
'is_multi': is_multi,
'author': author,
'copyright_stat': copyright_stat,
'duration': duration,
'del_flag': del_flag,
'type': type,
'publish_time': publish_time,
'sn': sn,
'__biz': __biz,
'spider_time': tools.get_current_date()
}
return article_data
def __parse_account_info(self, data, req_url):
'''
@summary:
---------
@param data:
---------
@result:
'''
__biz = tools.get_param(req_url, '__biz')
regex = 'id="nickname">(.*?)</strong>'
account = tools.get_info(data, regex, fetch_one=True).strip()
regex = 'profile_avatar">.*?<img src="(.*?)"'
head_url = tools.get_info(data, regex, fetch_one=True)
regex = 'class="profile_desc">(.*?)</p>'
summary = tools.get_info(data, regex, fetch_one=True).strip()
# 认证信息（关注的账号直接点击查看历史消息，无认证信息）
regex = '<i class="icon_verify success">.*?</i>(.*?)</span>'
verify = tools.get_info(data, regex, fetch_one=True)
verify = verify.strip() if verify else ''
# 二维码
regex = 'var username = "" \|\| "(.*?)";' # || 需要转译
qr_code = tools.get_info(data, regex, fetch_one=True)
qr_code = 'http://open.weixin.qq.com/qr/code?username='******'__biz': __biz,
'account': account,
'head_url': head_url,
'summary': summary,
'qr_code': qr_code,
'verify': verify,
'spider_time': tools.get_current_date()
}
if account_data:
data_pipeline.save_account(account_data)
def __parse_article_list(self, article_list, req_url):
'''
@summary: 解析文章列表
---------
@param article_list: 文章列表信息 str
{
"list":[
{
"comm_msg_info":{
"id":1000000513,
"type":49,
"datetime":1511354167,
"fakeid":"3082125093",
"status":2,
"content":""
},
"app_msg_ext_info":{
"title":"Python 内存优化",
"digest":"实际项目中，pythoner更加关注的是Python的性能问题。本文，关注的是Python的内存优化，一般说来，如果不发生内存泄露，运行在服务端的Python代码不用太关心内存，但是如果运行在客户端，那还是有优化的必要。",
"content":"",
"fileid":505083208,
"content_url":"http:\/\/mp.weixin.qq.com\/s?__biz=MzA4MjEyNTA5Mw==&mid=2652566858&idx=1&sn=d2a76f4a601f94d8acc7b436d18e9648&chksm=8464dd00b313541684c14f974325ea6ae725ffc901fd9888cc00d1acdd13619de3297a5d9a35&scene=27#wechat_redirect",
"source_url":"http:\/\/www.cnblogs.com\/xybaby\/p\/7488216.html",
"cover":"http:\/\/mmbiz.qpic.cn\/mmbiz_jpg\/fhujzoQe7TpODTuicia4geCiaIj1AbZwVQQVbRHy3FhzwMHEvCvtzXVicHTaPEu8jZ2pgkCAgBqEHugYMvzg3tpoww\/0?wx_fmt=jpeg",
"subtype":9,
"is_multi":1,
"multi_app_msg_item_list":[
{
"title":"面向对象：With the wonder of your love, the sun above always shines",
"digest":"With the wonder of your love, the sun above always shines",
"content":"",
"fileid":505083209,
"content_url":"http:\/\/mp.weixin.qq.com\/s?__biz=MzA4MjEyNTA5Mw==&mid=2652566858&idx=2&sn=97f223783da7748080f8103654447c99&chksm=8464dd00b313541601938565a41487ea76209331fd6f4c8996a2ff5572f4fd465de9fa4cbaac&scene=27#wechat_redirect",
"source_url":"https:\/\/mp.weixin.qq.com\/s\/_uD9jY4nXQQ6CtA__dsN8w?scene=25#wechat_redirect",
"cover":"http:\/\/mmbiz.qpic.cn\/mmbiz_jpg\/fhujzoQe7TpODTuicia4geCiaIj1AbZwVQQ5ukvwH1GPq5zlWxv05WvRiaw6BiaeyGRD1w17nAPGTlQgEvvDuZnB9HA\/0?wx_fmt=jpeg",
"author":"",
"copyright_stat":101,
"del_flag":1
}
],
"author":"",
"copyright_stat":100,
"del_flag":1
}
}
]
}
---------
@result:
'''
# log.debug(tools.dumps_json(article_list))
# 解析json内容里文章信息
def parse_article_info(article_info, release_time):
if not article_info:
return
# log.debug(tools.dumps_json(article_info))
title = article_info.get('title')
summary = article_info.get('digest')
url = article_info.get('content_url').replace('\\', '').replace('amp;', '')
source_url = article_info.get('source_url').replace('\\', '') # 引用的文章链接
cover = article_info.get('cover').replace('\\', '')
author = article_info.get('author')
if url and url.startswith('http://mp.weixin.qq.com/'):# 被发布者删除的文章 无url和其他信息， 此时取不到mid 且不用入库, 或者商城类的url不入库
mid = tools.get_param(url, 'mid') or tools.get_param(url, 'appmsgid') # 图文消息id 同一天发布的图文消息 id一样
idx = tools.get_param(url, 'idx') or tools.get_param(url, 'itemidx')# 第几条图文消息 从1开始
article_id = mid + idx # 用mid和idx 拼接 确定唯一一篇文章 如mid = 2650492260 idx = 1，则article_id = 26504922601
# 判断该文章库中是否已存在
if WechatAction._wechat_service.is_exist('wechat_article', article_id) or (ONLY_TODAY_MSG and release_time < tools.get_current_date('%Y-%m-%d')):
self._is_need_get_more = False
return # 不往下进行 舍弃之后的文章
__biz = tools.get_param(url, '__biz') # 用于关联公众号
# 缓存文章信息
WechatAction._article_info[article_id] = {
'article_id':int(article_id),
'title' : title,
'summary' : summary,
'release_time':release_time,
'url' : url,
'source_url' : source_url,
'cover' : cover,
'account':'',
'author' : author,
'__biz' : __biz,
'read_num' : None,
'like_num' : None,
'content' : '',
'comment' : [],
'record_time':tools.get_current_date()
}
# 将文章url添加到待抓取队列
WechatAction._todo_urls.append(url)
# log.debug(tools.dumps_json(article_list))
article_list = tools.get_json(article_list)
article_list = article_list.get('list', [])
for article in article_list:
article_type = article.get('comm_msg_info', {}).get('type')
if article_type != 49: # 49为常见的图文消息、其他消息有文本、语音、视频，此处不采集，格式不统一
continue
release_time = article.get('comm_msg_info', {}).get('datetime')
release_time = tools.timestamp_to_date(release_time)
# 微信公众号每次可以发多个图文消息
# 第一个图文消息
app_msg_ext_info = article.get('app_msg_ext_info', {})
parse_article_info(app_msg_ext_info, release_time)
if not self._is_need_get_more:
break
# 同一天附带的图文消息
multi_app_msg_item_list = app_msg_ext_info.get('multi_app_msg_item_list')
for multi_app_msg_item in multi_app_msg_item_list:
parse_article_info(multi_app_msg_item, release_time)
if not self._is_need_get_more:
break
# 将更新公众号为做完的回调加入到队列中
__biz = tools.get_param(req_url, '__biz') # 用于关联公众号
WechatAction._todo_urls.append(lambda: WechatAction._wechat_service.update_account_article_num(__biz))
def get_article_list(self, data, req_url):
'''
@summary: 获取文章列表
分为两种
1、第一次查看历史消息 返回的是html格式 包含公众号信息
2、下拉显示更多时 返回json格式
但是文章列表都是json格式 且合适相同
抓取思路：
1、如果是第一种格式，直接解析文章内容，拼接下一页json格式的地址
2、如果是第二种格式，
---------
@param data:
---------
@result:
'''
try:
# 判断是否为被封的账号， 被封账号没有文章列表
if 'list' in data:
# 取html格式里的文章列表
if 'action=home' in req_url:
# 解析公众号信息
self.__parse_account_info(data, req_url)
# 解析文章列表
regex = "msgList = '(.*?})';"
article_list = tools.get_info(data, regex, fetch_one=True)
article_list = article_list.replace('"', '"')
self.__parse_article_list(article_list)
#判断是否还有更多文章 没有跳转到下个公众号，有则下拉显示更多
regex = "can_msg_continue = '(\d)'"
can_msg_continue = tools.get_info(data,
regex,
fetch_one=True)
if can_msg_continue == '0': # 无更多文章
pass
elif self._is_need_get_more:
# 以下是拼接下拉显示更多的历史文章 跳转
# 取appmsg_token 在html中
regex = 'appmsg_token = "(.*?)";'
appmsg_token = tools.get_info(data,
regex,
fetch_one=True)
# 取其他参数 在url中
__biz = tools.get_param(req_url, '__biz')
pass_ticket = tools.get_param(req_url, 'pass_ticket')
next_page_url = 'https://mp.weixin.qq.com/mp/profile_ext?action=getmsg&__biz={__biz}&f=json&offset={offset}&count=10&is_ok=1&scene=124&uin=777&key=777&pass_ticket={pass_ticket}&wxtoken=&appmsg_token={appmsg_token}&x5=0&f=json'.format(
__biz=__biz,
offset=10,
pass_ticket=pass_ticket,
appmsg_token=appmsg_token)
WechatAction._todo_urls.append(next_page_url)
else: # json格式
data = tools.get_json(data)
article_list = data.get('general_msg_list', {})
self.__parse_article_list(article_list)
#判断是否还有更多文章 没有跳转到下个公众号，有则下拉显示更多
can_msg_continue = data.get('can_msg_continue')
if not can_msg_continue: # 无更多文章
pass
elif self._is_need_get_more:
# 以下是拼接下拉显示更多的历史文章 跳转
# 取参数 在url中
__biz = tools.get_param(req_url, '__biz')
pass_ticket = tools.get_param(req_url, 'pass_ticket')
appmsg_token = tools.get_param(req_url, 'appmsg_token')
# 取offset 在json中
offset = data.get('next_offset', 0)
next_page_url = 'https://mp.weixin.qq.com/mp/profile_ext?action=getmsg&__biz={__biz}&f=json&offset={offset}&count=10&is_ok=1&scene=124&uin=777&key=777&pass_ticket={pass_ticket}&wxtoken=&appmsg_token={appmsg_token}&x5=0&f=json'.format(
__biz=__biz,
offset=offset,
pass_ticket=pass_ticket,
appmsg_token=appmsg_token)
WechatAction._todo_urls.append(next_page_url)
else: # 该__biz 账号已被封
pass
except Exception as e:
log.error(e)
return self.__open_next_page()
def deal_article_list(self, req_url, text):
"""
@summary: 获取文章列表
分为两种
1、第一次查看历史消息 返回的是html格式 包含公众号信息
2、下拉显示更多时 返回json格式
但是文章列表都是json格式 且合适相同
抓取思路：
1、如果是第一种格式，直接解析文章内容，拼接下一页json格式的地址
2、如果是第二种格式，
---------
@param data:
---------
@result:
"""
try:
# 判断是否为被封的账号， 被封账号没有文章列表
__biz = tools.get_param(req_url, "__biz")
if "list" in text:
# 取html格式里的文章列表
if "action=home" in req_url:
# 解析公众号信息
self.__parse_account_info(text, req_url)
# 解析文章列表
regex = "msgList = '(.*?})';"
article_list = tools.get_info(text, regex, fetch_one=True)
article_list = article_list.replace(""", '"')
publish_time = self.__parse_article_list(
article_list, __biz, is_first_page=True)
# 判断是否还有更多文章 没有跳转到下个公众号，有则下拉显示更多
regex = "can_msg_continue = '(\d)'"
can_msg_continue = tools.get_info(text,
regex,
fetch_one=True)
if can_msg_continue == "0": # 无更多文章
log.info("抓取到列表底部 无更多文章，公众号 {} 抓取完毕".format(__biz))
new_last_publish_time = self._task_manager.get_new_last_article_publish_time(
__biz)
if not new_last_publish_time:
# 标记成僵尸号
log.info("公众号 {} 为僵尸账号 不再监控".format(__biz))
self._task_manager.sign_account_is_zombie(__biz)
else:
self._task_manager.update_account_last_publish_time(
__biz, new_last_publish_time)
elif publish_time:
# 以下是拼接下拉显示更多的历史文章 跳转
# 取appmsg_token 在html中
regex = 'appmsg_token = "(.*?)";'
appmsg_token = tools.get_info(text,
regex,
fetch_one=True)
# 取其他参数 在url中
__biz = tools.get_param(req_url, "__biz")
pass_ticket = tools.get_param(req_url, "pass_ticket")
next_page_url = "https://mp.weixin.qq.com/mp/profile_ext?action=getmsg&__biz={__biz}&f=json&offset={offset}&count=10&is_ok=1&scene=124&uin=777&key=777&pass_ticket={pass_ticket}&wxtoken=&appmsg_token={appmsg_token}&x5=0&f=json".format(
__biz=__biz,
offset=10,
pass_ticket=pass_ticket,
appmsg_token=appmsg_token,
)
return self._task_manager.get_task(
next_page_url,
tip="正在抓取列表 next_offset {} 抓取到 {}".format(
10, publish_time),
)
else: # json格式
text = tools.get_json(text)
article_list = text.get("general_msg_list", {})
publish_time = self.__parse_article_list(
article_list, __biz)
# 判断是否还有更多文章 没有跳转到下个公众号，有则下拉显示更多
can_msg_continue = text.get("can_msg_continue")
if not can_msg_continue: # 无更多文章
log.info("抓取到列表底部 无更多文章，公众号 {} 抓取完毕".format(__biz))
new_last_publish_time = self._task_manager.get_new_last_article_publish_time(
__biz)
self._task_manager.update_account_last_publish_time(
__biz, new_last_publish_time)
pass
elif publish_time:
# 以下是拼接下拉显示更多的历史文章 跳转
# 取参数 在url中
__biz = tools.get_param(req_url, "__biz")
pass_ticket = tools.get_param(req_url, "pass_ticket")
appmsg_token = tools.get_param(req_url, "appmsg_token")
# 取offset 在json中
offset = text.get("next_offset", 0)
next_page_url = "https://mp.weixin.qq.com/mp/profile_ext?action=getmsg&__biz={__biz}&f=json&offset={offset}&count=10&is_ok=1&scene=124&uin=777&key=777&pass_ticket={pass_ticket}&wxtoken=&appmsg_token={appmsg_token}&x5=0&f=json".format(
__biz=__biz,
offset=offset,
pass_ticket=pass_ticket,
appmsg_token=appmsg_token,
)
return self._task_manager.get_task(
next_page_url,
tip="正在抓取列表 next_offset {} 抓取到 {}".format(
offset, publish_time),
)
else: # 该__biz 账号已被封
self._task_manager.sign_account_is_zombie(__biz)
pass
except Exception as e:
log.exception(e)
return self._task_manager.get_task()
def deal_article(self, req_url, text):
"""
解析文章
:param req_url:
:param text:
:return:
"""
sn = tools.get_param(req_url, "sn")
if not text:
self._task_manager.update_article_task_state(sn, -1)
return None
selector = Selector(text)
content = selector.xpath(
'//div[@class="rich_media_content "]|//div[@class="rich_media_content"]|//div[@class="share_media"]'
)
title = (selector.xpath('//h2[@class="rich_media_title"]/text()').
extract_first(default="").strip())
account = (selector.xpath('//a[@id="js_name"]/text()').extract_first(
default="").strip())
author = (selector.xpath(
'//span[@class="rich_media_meta rich_media_meta_text"]//text()').
extract_first(default="").strip())
publish_timestamp = selector.re_first('n="(\d{10})"')
publish_timestamp = int(
publish_timestamp) if publish_timestamp else None
publish_time = (tools.timestamp_to_date(publish_timestamp)
if publish_timestamp else None)
pics_url = content.xpath(".//img/@src|.//img/@data-src").extract()
biz = tools.get_param(req_url, "__biz")
digest = selector.re_first('var msg_desc = "(.*?)"')
cover = selector.re_first('var cover = "(.*?)";') or selector.re_first(
'msg_cdn_url = "(.*?)"')
source_url = selector.re_first("var msg_source_url = '(.*?)';")
content_html = content.extract_first(default="")
comment_id = selector.re_first('var comment_id = "(\d+)"')
article_data = {
"account": account,
"title": title,
"url": req_url,
"author": author,
"publish_time": publish_time,
"__biz": biz,
"digest": digest,
"cover": cover,
"pics_url": pics_url,
"content_html": content_html,
"source_url": source_url,
"comment_id": comment_id,
"sn": sn,
"spider_time": tools.get_current_date(),
}
# 入库
if article_data and data_pipeline.save_article(
article_data) is not None:
self._task_manager.update_article_task_state(sn, 1)
return self._task_manager.get_task()