def crawler_web_seed_url(url):
'''
爬取种子页的所有a链接
:param url: 种子页url
:return: 无
'''
r = RequestUtil()
hu = HtmlUtil()
u = Util()
# 通过phandomjs 请求url，返回网页，包括网页的ajax请求
html = r.http_get_phandomjs(url)
#html = html.decode('utf-8').encode(sys.getfilesystemdomainencoding())
#print html
#可以从HTML或XML文件中提取数据的Python第三方库
soup = BeautifulSoup(html, 'lxml')
# a链接dom对象列表
a_docs = soup.find_all("a")
aset = set()
#获取domain
domain = hu.get_url_domain(url)
#获取host
host = hu.get_url_host(url)
print 'domain==>', domain
print 'host==>', host
for a in a_docs:
#获取a标签的href
a_href = hu.get_format_url(url, a, host)
#获取a标签的内容
a_title = a.get_text().strip()
if a_href == '' or a_title == '':
continue
if aset.__contains__(a_href):
continue
aset.add(a_href)
#获取a标签的host
a_host = hu.get_url_host(a_href)
#获取a标签href链接url的md5
a_md5 = u.get_md5(a_href)
#获取a标签所对应的xpath
a_xpath = hu.get_dom_parent_xpath_js_new(a)
print("%s\t%s\t%s\t%s\t%s") % (a_title.decode("utf-8"), a_href, a_host,
a_md5, a_xpath)
r.close_phandomjs()
def print_news_url_content(news_url):
'''
打印最终新闻页面内容
:param news_url:
:return:
'''
r = RequestUtil()
hu = HtmlUtil()
u = Util()
# 通过phandomjs 请求url，返回网页，包括网页的ajax请求
html = r.http_get_phandomjs(news_url)
#html = html.decode('utf-8').encode(sys.getfilesystemdomainencoding())
print html
r.close_phandomjs()
def test_beautiful():
# url = 'http://roll.news.qq.com'
url ='http://politics.gmw.cn/node_9844.htm'
r = RequestUtil()
hu = HtmlUtil()
html = r.http_get_phandomjs(url)
domain = get_tld(url)
host = hu.get_url_host(url)
u = Util()
print "domain:",domain,":host:",host
soup = BeautifulSoup(html, 'lxml')
a_docs = soup.find_all("a")
for a in a_docs:
a_href = get_format_url(url,a,host)
if a.text:
print a.text
if a_href:
xpath = hu.get_dom_parent_xpath_js(a)
print a_href,'_',xpath,u.get_md5(xpath)
def test_beautiful():
r = RequestUtil()
hu = HtmlUtil()
u = Util()
url = 'https://news.sina.com.cn/roll/#pageid=153&lid=2509&k=&num=50&page=1'
html = r.http_get_phandomjs(url)
#html = html.decode('utf-8').encode(sys.getfilesystemencoding())
#print html
#可以从HTML或XML文件中提取数据的Python第三方库
soup = BeautifulSoup(html, 'lxml')
a_docs = soup.find_all("a")
aset = set()
#获取domain
domain = get_fld(url)
#获取host
host = hu.get_url_host(url)
print 'domain==>', domain
print 'host==>', host
for a in a_docs:
#获取a标签的href
a_href = get_format_url(url, a, host)
#获取a标签的内容
a_title = a.get_text().strip()
if a_href == '' or a_title == '':
continue
if aset.__contains__(a_href):
continue
aset.add(a_href)
#获取a标签的host
a_host = hu.get_url_host(a_href)
#获取a标签href链接url的md5
a_md5 = u.get_md5(a_href)
#获取a标签所对应的xpath
a_xpath = hu.get_dom_parent_xpath_js(a)
print("%s\t%s\t%s\t%s\t%s") % (a_title.decode("utf-8"), a_href, a_host,
a_md5, a_xpath)
xpath_test.py
Created on 2019/3/16 20:23
Copyright (c) 2019/3/16, 海牛学院版权所有.
@author: 潘牛
'''
import mx.URL, sys
from tld import get_tld
from bs4 import BeautifulSoup
from lxml import etree
from commons.util.request_util import RequestUtil
from commons.util.html_util import HtmlUtil
from commons.util.util import Util
if __name__ == '__main__':
r = RequestUtil()
hu = HtmlUtil()
u = Util()
url = 'https://news.sina.com.cn/roll/#pageid=153&lid=2509&k=&num=50&page=1'
html = r.http_get_phandomjs(url)
dom_tree = etree.HTML(html)
###XPath匹配
a_text = dom_tree.xpath(
"//div[@id='d_list']/ul[5]/li[2]/span[contains(@class,'c_tit')]/a[1]/text()"
)
a_href = dom_tree.xpath("//div[@id='d_list']/ul[8]/li[3]/span[2]/a/@href")
print a_text[0]
print a_href[0]
def action(self):
is_success = True
t = TimeUtil()
u = Util()
hu = HtmlUtil()
r = RequestUtil()
in_values = []
ex_values = []
a_href = ''
main_md5 = u.get_md5(self.url)
now_time = datetime.now()
update_time = int(time.mktime(now_time.timetuple()))
create_time = update_time
create_day = int(t.now_day().replace('-', ''))
create_hour = int(t.now_hour())
try:
html = r.http_get_phandomjs(self.url)
domain = get_tld(self.url)
soup = BeautifulSoup(html, 'lxml')
a_docs = soup.find_all("a")
a_set = set()
a_param = {}
out_json_srt = ''
status = 0
host = hu.get_url_host(self.url)
for a in a_docs:
a_href = self.get_format_url(a,host)
a_title = a.get_text().strip()
if a_href == '' or a_title == '':
continue
if a_set.__contains__(a_href):
continue
a_set.add(a_href)
req = urllib2.Request(url=a_href)
a_host = req.get_host() if req.get_host() is not None else ''
a_md5 = u.get_md5(a_href)
if a_title != '':
a_param['title'] = a_title
out_json_srt = json.dumps(a_param,ensure_ascii=False)
a_xpath = hu.get_dom_parent_xpath_js(a)
insert_values = (main_md5,domain,host,a_md5,a_host,a_xpath,create_time,create_day,create_hour,update_time,status,
MySQLdb.escape_string(self.url),
MySQLdb.escape_string(a_href),
MySQLdb.escape_string(a_title),
out_json_srt)
if a_host.__contains__(domain):
in_values.append(insert_values)
else:
ex_values.append(insert_values)
in_table = 'hainiu_web_seed_internally'
ex_table = 'hainiu_web_seed_externally'
insert_sql = """
insert into <table> (md5,domain,host,a_md5,a_host,a_xpath,create_time,create_day,create_hour,update_time,status,url,a_url,a_title,param)
values (%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s) ON DUPLICATE KEY UPDATE update_time=update_time;
"""
try:
d = DBUtil(config._HAINIU_DB)
d.execute_no_commit("set NAMES utf8mb4;")
if in_values.__len__() != 0:
sql = insert_sql.replace('<table>',in_table)
d.executemany_no_commit(sql,in_values)
if ex_values.__len__() != 0:
sql = insert_sql.replace('<table>',ex_table)
d.executemany_no_commit(sql,ex_values)
d.commit()
except:
is_success = False
self.rl.exception()
self.rl.error(sql)
d.rollback()
finally:
d.close()
except:
is_success = False
self.rl.exception()
finally:
r.close_phandomjs()
return super(self.__class__, self).result(is_success, [main_md5,self.url,a_href,in_values.__len__(),ex_values.__len__(),self.queue_id])
def action(self):
#爬取 hainiu_queue 中符合要求的url 请求页面的所有 a标签url
r = RequestUtil()
hu = HtmlUtil()
u = Util()
#
is_success = True
db_util = DBUtil(_HAINIU_DB)
time_util = TimeUtil()
# 内外链表的列表
inner_list = []
exter_list = []
#获取种子的md5
md5 = u.get_md5(self.act)
try:
# 通过phandomjs 请求url，返回网页，包括网页的ajax请求
html = r.http_get_phandomjs(self.act)
#可以从HTML或XML文件中提取数据的Python第三方库
soup = BeautifulSoup(html, 'lxml')
# a链接dom对象列表
a_docs = soup.find_all("a")
if len(a_docs) == 0:
is_success = False
aset = set()
#获取种子的domain
domain = hu.get_url_domain(self.act)
#获取种子的host
host = hu.get_url_host(self.act)
# 时间（create_time、create_day、create_hour、update_time）
# create_time=time_util.get_timestamp()
#
# create_day = int(time_util.now_day().replace('-', ''))
# create_hour=int(time_util.now_hour())
# update_time=create_time
create_time = time_util.get_timestamp()
# 获取年月日格式
create_day = int(time_util.now_day(format='%Y%m%d'))
# 获取小时
create_hour = int(time_util.now_hour())
update_time = create_time
# params_json = json.dumps(self.params, ensure_ascii=False, encoding='utf-8')
for a_doc in a_docs:
#获取a标签的href
a_href = hu.get_format_url(self.act, a_doc, host)
#获取a标签的内容
a_title = a_doc.get_text().strip()
if a_href == '' or a_title == '':
continue
if aset.__contains__(a_href):
continue
aset.add(a_href)
#获取a标签的host
a_host = hu.get_url_host(a_href)
#获取a标签href链接url的md5
a_md5 = u.get_md5(a_href)
#获取a标签所对应的xpath
a_xpath = hu.get_dom_parent_xpath_js_new(a_doc)
# 一行数据
row_data = (self.act, md5, self.params, domain, host, a_href,
a_md5, a_host, a_xpath, a_title, create_time,
create_day, create_hour, update_time)
if a_href.__contains__(domain):
inner_list.append(row_data)
else:
exter_list.append(row_data)
# 并解析存入内链表或外链表，在存入时，如果url已存在，只做
# update 操作。（保证链接页面不会重复爬取）
if len(inner_list) > 0:
inner_insert_sql = """
insert into hainiu_web_seed_internally
(url,md5,param,domain,host,a_url,a_md5,a_host,a_xpath,a_title,create_time,
create_day,create_hour,update_time)
values (%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s)
ON DUPLICATE KEY UPDATE update_time=values(update_time);
"""
db_util.executemany_no_commit(inner_insert_sql, inner_list)
if len(exter_list) > 0:
exter_insert_sql = """
insert into hainiu_web_seed_externally
(url,md5,param,domain,host,a_url,a_md5,a_host,a_xpath,a_title,create_time,
create_day,create_hour,update_time)
values (%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s)
ON DUPLICATE KEY UPDATE update_time=values(update_time);
"""
db_util.executemany_no_commit(exter_insert_sql, exter_list)
db_util.commit()
except Exception, e:
is_success = False
db_util.rollback()
traceback.print_exc(e)
def action(self, *values):
# 插入内链表sql语句
insert_seed_internally='''
insert into web_seed_internally
(url,md5,param,domain,host,a_url,a_md5,a_host,a_xpath,a_title,create_time,create_day,create_hour,update_time,status)
values (%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s) ON DUPLICATE KEY UPDATE update_time=VALUES(update_time);
'''
# 插入外链表sql语句
insert_seed_externally='''
insert into web_seed_externally
(url,md5,param,domain,host,a_url,a_md5,a_host,a_xpath,a_title,create_time,create_day,create_hour,update_time,status)
values (%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s,%s) ON DUPLICATE KEY UPDATE update_time=VALUES(update_time);
'''
# 获取时间
a_time = TimeUtil()
db_util = DBUtil(_ZZ_DB)
# redis_d = RedisUtill()
total_count = 0
in_count = 0
ex_count = 0
try:
# 解析主网页信息
hu = HtmlUtil()
domain = hu.get_url_domain(self.act)
host = hu.get_url_host(self.act)
u = Util()
md5 = u.get_md5(self.act)
# 解析a标签信息
r = RequestUtil()
# 通过phandomjs 请求url，返回网页，包括网页的ajax请求
html = r.http_get_phandomjs(self.act)
# 可以从HTML或XML文件中提取数据的Python第三方库
soup = BeautifulSoup(html, 'lxml')
# a链接dom对象列表
aset = set()
# 获取host
a_host = hu.get_url_host(self.act)
# a_docs = soup.find_all("a",href=re.compile("^(/|.*"+domain+")"))
a_docs = soup.find_all("a")
for a in a_docs:
total_count += 1
# 获取a标签的href
a_url = hu.get_format_url(self.act,a,a_host)
# 获取a标签的内容
a_title = a.get_text().strip()
if a_url == '' or a_title == '':
continue
if aset.__contains__(a_url):
continue
aset.add(a_url)
# 获取a标签的host
a_host = hu.get_url_host(a_url)
# 获取a标签href链接url的md5
a_md5 = u.get_md5(a_url)
# 获取a标签所对应的xpath
a_xpath = hu.get_dom_parent_xpath_js_new(a)
create_time = a_time.get_timestamp()
create_day = int(a_time.now_day(format='%Y%m%d'))
create_hour = int(a_time.now_hour())
params_sql = [self.act,md5,self.params,domain,host,a_url,a_md5,a_host,a_xpath,a_title,create_time,create_day,create_hour,create_time,0]
if re.compile("^(/|.*"+domain+")").match(a_url) is not None:
db_util.execute(insert_seed_internally, params_sql)
#
# # redis
# redis_md5 = u.get_md5(md5+"\001"+a_md5)
# find_key = redis_d.get_value_for_key('seed:%s:a_url' % redis_md5)
# if find_key == None:
# # url,md5,param,domain,host,a_url,a_md5,a_host,a_xpath,a_title,create_time,create_day,create_hour,update_time,status
# dicts = {'seed:%s:param' % redis_md5 :self.params, 'seed:%s:a_url' % redis_md5 : a_url,
# 'seed:%s:md5' % redis_md5 : md5, 'seed:%s:a_md5' % redis_md5 :a_md5}
#
# dicts_temp = {'seed_temp:%s:param' % redis_md5 :self.params,'seed_temp:%s:a_url' % redis_md5 : a_url,
# 'seed_temp:%s:md5' % redis_md5 : md5, 'seed_temp:%s:a_md5' % redis_md5 : a_md5}
# redis_d.set_batch_datas(dicts)
# redis_d.set_batch_datas(dicts_temp)
in_count += 1
else:
db_util.execute(insert_seed_externally, params_sql)
ex_count += 1
r.close_phandomjs()
except Exception, err:
db_util.rollback()
traceback.print_exc(err)
def action(self):
is_success = True
t = TimeUtil()
f = FileUtil()
u = Util()
hu = HtmlUtil()
r = RequestUtil()
values = []
md5 = u.get_md5(self.url)
now_time = datetime.now()
update_time = int(time.mktime(now_time.timetuple()))
create_time = update_time
create_day = int(t.now_day().replace('-', ''))
create_hour = int(t.now_hour())
now_minute = int(t.now_min())
for i in xrange(60, -5, -5):
if now_minute >= i:
now_minute = i
break
now_minute = t.now_time(format='%Y%m%d%H') + (
'0%s' % (str(now_minute)) if now_minute < 10 else str(now_minute))
values.append(MySQLdb.escape_string(self.url))
values.append(md5)
values.append(create_time)
values.append(create_day)
values.append(create_hour)
values.append('')
values.append(MySQLdb.escape_string(self.param))
values.append(update_time)
try:
html = r.http_get_phandomjs(self.url)
domain = get_tld(self.url)
values[5] = domain
soup = BeautifulSoup(html, 'lxml')
title_doc = soup.find('title')
title = title_doc.contents[0] if title_doc is not None and len(
title_doc.contents) == 1 else ''
host = hu.get_url_host(self.url)
values.append(host)
values.append(MySQLdb.escape_string(title))
# k = KafkaUtil(config._KAFKA_CONFIG)
html = html.replace(content._SEQ1,
'').replace(content._SEQ2, content._SEQ4)
# push_str = content._SEQ3.join(('%s','%s')) % (self.url,html)
# push_str = content._SEQ3.join(('%s','%s')) % (u.get_md5(push_str),push_str)
# push_str = bytes(push_str)
# is_success = k.push_message(push_str)
is_success = True
if is_success:
self.save_file(create_time, f, now_minute, u, self.url, html)
else:
values.append('')
values.append('')
self.rl.error("kafka push error")
except:
is_success = False
values.append('')
values.append('')
self.rl.exception()
finally:
r.close_phandomjs()
try:
if is_success:
values.append(1)
insert_web_page_sql = """
insert into hainiu_web_page (url,md5,create_time,create_day,create_hour,domain,param,update_time,host,
title,status) values ("%s","%s",%s,%s,%s,"%s","%s",%s,"%s","%s",%s) on DUPLICATE KEY UPDATE update_time=values(update_time);
"""
else:
ip = u.get_local_ip()
values.append(ip)
values.append(2)
insert_web_page_sql = """
insert into hainiu_web_page (url,md5,create_time,create_day,create_hour,domain,param,update_time,host,
title,fail_ip,status) values ("%s","%s",%s,%s,%s,"%s","%s",%s,"%s","%s","%s",%s)
on DUPLICATE KEY UPDATE fail_times=fail_times+1,fail_ip=values(fail_ip);
"""
d = DBUtil(config._HAINIU_DB)
sql = insert_web_page_sql % tuple(values)
d.execute(sql)
except:
is_success = False
self.rl.exception()
self.rl.error(sql)
d.rollback()
d.commit()
finally:
d.close()
return super(self.__class__,
self).result(is_success,
[md5, self.url, update_time, self.queue_id])
def action(self):
logger = LogUtil().get_logger("download_action", "download_action")
#1）把队列中的url的HTML内容下载到文件中，每个消费线程每隔5分钟生成一个新的文件。
r = RequestUtil()
# hu = HtmlUtil()
u = Util()
db_util = DBUtil(_HAINIU_DB)
time_util = TimeUtil()
# 通过phandomjs 请求url，返回网页，包括网页的ajax请求
html = r.http_get_phandomjs(self.act)
# 拼接要写入的内容
html = html.replace("\r", "").replace("\n", "\002")
str1 = self.act + "\001" + html
str2 = u.get_md5(str1) + "\001" + str1
# 成功失败标记
is_success = True
# 获取时间
# now_time====>年月日时分秒
now_time = time.strftime("%Y%m%d,%H,%M,%S").split(",")
day = now_time[0]
hour = now_time[1]
minute = int(now_time[2])
for i in range(60, -5, -5):
if minute < i:
continue
minute = i
break
minute = '0%s' % minute if minute < 10 else minute
now_minute = '%s%s%s' % (day, hour, minute)
file_names = os.listdir(_LOCAL_DATA_DIR % ('tmp'))
logger.info("file_names:%s" % file_names)
thread_name = self.consumer_thread_name
logger.info("thread_name:%s" % thread_name)
last_file_name = ''
for file_name in file_names:
tmp = file_name.split("#")[0]
if tmp == thread_name:
last_file_name = file_name
break
now_file_name = "%s#%s" % (thread_name, now_minute)
try:
if last_file_name == '' or last_file_name != now_file_name:
# 移动老文件
# if last_file_name != '':
oldPath = _LOCAL_DATA_DIR % ("tmp/") + last_file_name
logger.info("oldPath:%s" % oldPath)
# if os.path.exists(oldPath) and os.path.getsize(oldPath) > 0:
if last_file_name != '':
done_file_name = last_file_name + "#" + str(
TimeUtil().get_timestamp())
logger.info("last_file_name:%s" % last_file_name)
newPath = _LOCAL_DATA_DIR % ("done/") + done_file_name
logger.info("newPath:%s" % newPath)
shutil.move(oldPath, newPath)
# 写入新文件
now_file_name = _LOCAL_DATA_DIR % ("tmp/") + now_file_name
# if not os.path.exists(_LOCAL_DATA_DIR+'tmp2/'):
# os.mkdir(_LOCAL_DATA_DIR+'tmp2/')
logger.info("now_file_name:%s" % now_file_name)
f = open(now_file_name, 'a+')
f.write(str2)
f.close()
else:
last_file_name = _LOCAL_DATA_DIR % ("tmp/") + last_file_name
logger.info("last_file_name:%s" % last_file_name)
# 写入老文件时进行换行
insert_str = "\n" + str2
f = open(last_file_name, 'a+')
f.write(insert_str)
f.close()
except Exception, e:
is_success = False
traceback.print_exc(e)