bacow/AV_Data_Capture
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

check uri not start with /

Browse Source
This commit is contained in:
lededev
2021-06-06 04:57:24 +08:00
parent 5abeb360af
commit e2cd1f09df
1 changed files with 4 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
WebCrawler/javbus.py
View File

@@ -49,7 +49,10 @@ def getCover(htmlcode): #获取封面链接
image = doc('a.bigImage') image = doc('a.bigImage')
uri = image.attr('href') uri = image.attr('href')
if uri[0:4] != 'http': if uri[0:4] != 'http':
return "https://www.javbus.com" + uri if uri[0] == '/':
return "https://www.javbus.com" + uri
else:
return "https://www.javbus.com/" + uri
return uri return uri
def getRelease(htmlcode): #获取出版日期 def getRelease(htmlcode): #获取出版日期
html = etree.fromstring(htmlcode, etree.HTMLParser()) html = etree.fromstring(htmlcode, etree.HTMLParser())