def get_origin_info(bw_id, logger):
try:
time.sleep(5)
url = 'https://m.weibo.cn/statuses/show?id=' + str(bw_id)
r = requests.get(url, headers=get_header(), proxies=get_proxy())
r.raise_for_status()
r.encoding = r.apparent_encoding
content = json.loads(r.text)
if content.get('ok') == 1:
# 默认为原创
origin = True
# 若包含被转发微博信息，判断为转发
if 'retweeted_status' in r.text:
origin = False
# 获取转发页数
rp_count = jsonpath(content, '$.data.reposts_count')[0]
if rp_count > 0:
rp_page = int(rp_count) / 10 + 1
else:
rp_page = 0
# 获取被转发用户信息
origin_user = jsonpath(content, '$.data.user')[0]
info_dict = {
'bw_id': bw_id,
'origin': origin,
'rp_count': rp_count,
'rp_page': rp_page,
'origin_user': origin_user
}
return info_dict
else:
return False
except Exception as e:
logger.error(f"Cannot get details of weibo {bw_id}. {e}")
def get_more_topic(query, epoch, topic_dir):
topic_list = []
page_count = 0
# 获取返回的总页数
base_url = 'https://m.weibo.cn/api/container/getIndex?containerid=100103type%3D38%26q%3D' + str(query) + '%26t%3D0&page_type=searchall'
try:
r = requests.get(base_url, headers=get_header(), proxies=get_proxy())
r.raise_for_status()
page = json.loads(r.text)['data']['cardlistInfo']['total'] / 10
print(f'[{time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())}] EPOCH: {epoch}. Keyword: {query}. Get {page} pages of new topics.')
except Exception:
time.sleep(60)
get_more_topic(query, epoch, topic_dir)
while(page_count <= page):
time.sleep(3)
page_count += 1
this_url = base_url + '&page=' + str(page_count)
print(f'[{time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())}] Crawling Topic. Page {page_count} of keyword {query}')
try:
r = requests.get(this_url, headers=get_header(), proxies=get_proxy())
r.raise_for_status()
r.encoding = r.apparent_encoding
content = json.loads(r.text)
if content['ok'] == 1:
items = jsonpath(content, '$..card_group..title_sub')
for item in items:
temp = item.strip('#')
if temp != query.strip():
topic_list.append([temp])
else:
continue
except Exception:
print(f'[{time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())}] Error happen in page --->" + str(page_count)')
# 结果写入文件
with open(topic_dir + 'Topics_' + str(epoch) + '.csv', 'a', encoding='utf-8', newline='') as f:
writer = csv.writer(f)
writer.writerows(topic_list)
# 获取元素输出
print(f'[{time.strftime("%Y-%m-%d %H:%M:%S", time.localtime())}] Finished Crawling Topic. Get {len(topic_list)} new topic for keyword {query}')
def get_query_info(wd, writer, logger, since_date=None):
if_crawl = True
page_count = 0
error = {}
# 将检索词编码，嵌入url得到不同词的url字典
# 爬取检索页面下热门栏的页面
base_url = 'https://m.weibo.cn/api/container/getIndex?containerid=100103type%3D60%26q%3D' + quote(
wd) + '%26t%3D0&page_type=searchall'
# 计算可获取的总页数
page = get_Page(wd, base_url, logger)
# 获取包含检索词的相关微博
while (page_count <= page):
result_list = []
page_count += 1
this_url = base_url + '&page=' + str(page_count)
# logger.info(f'Page {page_count}: {this_url}')
try:
time.sleep(3)
r = requests.get(this_url,
headers=get_header(),
proxies=get_proxy())
logger.info(f'Crawling Query. Page {page_count} of keyword {wd}')
r.raise_for_status()
r.encoding = r.apparent_encoding
content = json.loads(r.text)
if content.get('ok') == 1:
mblogs = jsonpath(content, '$.data.cards..mblog')
for mblog in mblogs:
# 含有该键的mblog表示该条微博不是原创微博
if mblog.get('retweeted_status'):
continue
mblog['created_at'] = standardize_date(mblog['created_at'])
this_topic, this_text = getText(mblog)
this_dict = {
'keyword': str(wd),
'user_id': mblog['user']['id'],
'screen_name': mblog['user']['screen_name'],
'bw_id': mblog['id'],
'repost_count': mblog['reposts_count'],
'topic': this_topic,
'content': this_text,
'created_at': mblog['created_at']
}
if since_date:
since_date = datetime.strptime(since_date, '%Y-%m-%d')
created_at = datetime.strptime(mblog['created_at'],
'%Y-%m-%d')
if (created_at > since_date):
if_crawl = False
else:
if_crawl = False
if not if_crawl:
result_list.append(this_dict)
# 将该页面符合规定时间的内容写入csv
writer.write_csv(result_list)
else:
continue
except Exception as e:
# 若第一次错误，则将url加入error，并休息60s
if error.get(this_url) is None:
error[this_url] = 1
page_count -= 1
time.sleep(60)
# 若第二次错误，则报错
else:
logger.error(f'Page {page_count} failed. {e}')
def get_repost_info(center_bw_id,
bw_id,
level,
writer,
logger,
temp_writer,
since_date=None):
error = {}
idList = []
# 获取原博主信息
origin_info = get_origin_info(bw_id, logger)
# 获取成功时：
if origin_info:
# 原创信息
origin = origin_info['origin']
# 用户信息
origin_user = origin_info['origin_user']
# 转发数和转发总页数
rp_count = origin_info['rp_count']
page = origin_info['rp_page']
# 可能出现微博删除或无法获取的情况，则不再获取该bw_id
else:
return None
# 转发信息爬取
if page == 0:
logger.info(
f'Center bw : {center_bw_id}. level: {level}. No repost of this bw {bw_id}.'
)
writer.write_csv(None,
END=True,
center_bw_id=center_bw_id,
origin_info=origin_info,
level=level)
else:
logger.info(
f'Center bw : {center_bw_id}. Get {page} pages of bw {bw_id}.')
base_url = 'https://m.weibo.cn/api/statuses/repostTimeline?id=' + str(
bw_id) + '&page='
page_count = 0
while (page_count <= page):
page_count += 1
result_list = []
try:
time.sleep(7)
this_url = base_url + str(page_count)
logger.info(
f'Center bw : {center_bw_id}. level: {level}. Crawling page {page_count} of bw {bw_id}.'
)
r = requests.get(this_url,
headers=get_header(),
proxies=get_proxy())
r.raise_for_status()
r.encoding = r.apparent_encoding
content = json.loads(r.text)
if content.get('ok') == 1:
datas = jsonpath(content, '$.data.data.*')
for data in datas:
data['created_at'] = standardize_date(
data['created_at'])
flag = checkLevel(level, origin_user['screen_name'],
data['raw_text'])
if flag:
this_dict = {
'center_bw_id': center_bw_id,
'user_id': origin_user['id'],
'screen_name': origin_user['screen_name'],
'bw_id': bw_id,
'origin': origin,
'repost_count': rp_count,
'fs_count': origin_user['followers_count'],
'fs_user_id': data['user']['id'],
'fs_screen_name': data['user']['screen_name'],
'fs_bw_id': data['id'],
'fs_fans_count':
data['user']['followers_count'],
'level': level,
'raw_text': data['raw_text'],
'created_at': data['created_at']
}
# 将待爬取id放入下一轮爬取的id列表(即其作为原博时)
idList.append({'bw_id': data['id']})
# 判断是否是规定时间之后产生的微博
if since_date:
since_date = datetime.strptime(
since_date, '%Y-%m-%d')
created_at = datetime.strptime(
data['created_at'], '%Y-%m-%d')
if (created_at > since_date):
if_crawl = False
else:
if_crawl = False
if not if_crawl:
result_list.append(this_dict)
else:
continue
# 将符合规定时间的内容写入csv
writer.write_csv(result_list)
else:
continue
except Exception as e:
if error.get(this_url) is None:
error[this_url] = 1
page_count -= 1
time.sleep(60)
else:
logger.error(
f"Cannot get page {page_count} of bw {bw_id}. {e}")
# 爬取完所有页数，将idList写入对应的level文件
if idList:
temp_writer.write_csv(idList)
def get_Page(wd, base_url, logger):
r = requests.get(base_url, headers=get_header(), proxies=get_proxy())
r.raise_for_status()
page = json.loads(r.text)['data']['cardlistInfo']['total'] / 10 + 1
logger.info(f'Keyword: {wd}. Get {page} pages of returned weibo.')
return page
