def get_format_html(cls, html, final_url):
"""获得标准化的html"""
final_protocol = UrlUtil.get_protocol(final_url)
final_domain = UrlUtil.get_domain(final_url)
format_html = HtmlUtil.parse_protocol(html, final_protocol)
format_html = HtmlUtil.parse_domain(format_html, final_protocol, final_domain)
return format_html
def parse(self):
if not isinstance(self.downloader_item, MainItem):
logging.error("The param type is: " +
str(type(self.downloader_item)) +
", but it should be MainItem.")
return None
html = self.downloader_item.html
# 将downloader_item存库
with self.connection.cursor() as cursor:
sql = 'INSERT INTO snapshot (request_url, final_url, load_time, refer, get_time,' \
' task_id, send_ip, server_ip, deepth) VALUES (%s, %s, %s, %s, %s, %s, %s, ' \
'%s, %s);'
result = cursor.execute(sql, self.downloader_item.save_tuple())
if result != 1:
logging.error("snapshot插入记录" +
self.downloader_item.save_tuple() + "失败！")
# 拿到刚刚存库记录的id
with self.connection.cursor() as cursor:
sql = 'SELECT last_insert_id() as ss_id;'
cursor.execute(sql)
result = cursor.fetchone()
ss_id = result["ss_id"]
# 将页面内容存库
ss_html = SsHtmlItem(ss_id=ss_id, html=html)
with self.connection.cursor() as cursor:
sql = 'INSERT INTO ss_html (ss_id, html) VALUES (%s, %s);'
result = cursor.execute(sql, ss_html.save_tuple())
if result != 1:
logging.error("ss_html插入记录" + ss_html.save_tuple() + "失败！")
# 规范化一下页面内链接
final_protocol = UrlUtil.get_protocol(self.downloader_item.final_url)
final_domain = UrlUtil.get_domain(self.downloader_item.final_url)
format_html = HtmlUtil.parse_protocol(html, final_protocol)
format_html = HtmlUtil.parse_domain(format_html, final_protocol,
final_domain)
tree = etree.HTML(format_html)
hrefs = tree.xpath("//@href") # 拿到所有a标签中的链接对象
iframes = tree.xpath("//iframe/@src") # 拿到所有iframe的源链接
jss = tree.xpath("//script/@src") # 拿到所有的js链接
hrefs.extend(iframes)
hrefs.extend(jss)
if hrefs:
hrefs = href_clean(hrefs)
else:
hrefs = list()
inner_chains = set() # 内链列表，返回给引擎迭代
unknown_domains = set() # 可疑外链主域名列表，存库人工复查
request_top_domain = UrlUtil.get_top_domain(
self.downloader_item.request_url)
for href in hrefs:
this_top_domain = UrlUtil.get_top_domain(href)
if request_top_domain == this_top_domain and UrlUtil.get_url_suffix(
href) != "js":
inner_chains.add(href)
elif this_top_domain not in self.safe_chains and not UrlUtil.is_gov_or_edu(
href):
# 主域名不在白名单里而且不是政府或教育机构网站
unknown_domains.add(this_top_domain)
# 将须迭代的内链包装对象放入redis
logging.info("Length of inner_chains is " + str(len(inner_chains)))
dup_set_name = "engine:dup_set:" + str(self.downloader_item.task_id)
queue_name = "engine:queue:" + str(self.downloader_item.task_id)
for inner_chain in inner_chains:
if isinstance(self.redis_conn.ttl(dup_set_name), int):
sadd_re = self.redis_conn.sadd(dup_set_name, inner_chain)
if sadd_re == 1: # 等于1说明上条插入成功，没有重复，省了一次查重
new_main_item = MainItem(
inner_chain,
refer=self.downloader_item.final_url,
task_id=self.downloader_item.task_id,
deepth=self.downloader_item.deepth + 1)
self.redis_conn.lpush(
queue_name,
json.dumps(new_main_item, default=main_item_to_json))
# 将可疑外链存库
for unknown_domain in unknown_domains:
with self.connection.cursor() as cursor:
sql = "SELECT mydomain FROM malicious_domains;"
cursor.execute(sql)
malicious_records = cursor.fetchall()
malicious_domains = set([
malicious_record["mydomain"]
for malicious_record in malicious_records
])
if unknown_domain in malicious_domains:
suspicious_item = SuspiciousItem(
ss_id, unknown_domain, 1, 1,
time.strftime('%Y-%m-%d %H:%M:%S',
time.localtime(time.time())))
else:
suspicious_item = SuspiciousItem(ss_id, unknown_domain, 0, -1,
None)
with self.connection.cursor() as cursor:
sql = 'INSERT INTO suspicious_records (ss_id, unknown_domain, checked, result, ' \
'check_time) VALUES (%s, %s, %s, %s, %s)'
result = cursor.execute(sql, suspicious_item.save_tuple())
if result != 1:
logging.error("suspicious_records插入记录" +
suspicious_item.save_tuple() + "失败！")
self.connection.commit()
self.connection.close()
logging.info(self.downloader_item.request_url + " parse over.")