v4l3n71n/web_scrap
1
0
Fork 0
Code Issues Pull Requests Projects Releases 10 Wiki Activity

add private method album

Browse Source
This commit is contained in:
Valentin CZERYBA 2023-08-05 12:03:17 +02:00
parent 02f0c20bd0
commit 28b513e1b2
2 changed files with 15 additions and 4 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
lib/WPImport.py
View File

@ -54,6 +54,10 @@ class WPimport:
articlebody = soup.find_all("div", class_="articlebody")
if len(articlebody) > 0:
self._addOrUpdatePost(soup)
else:
albumbody = soup.find_all("div", class_="albumbody")
if len(albumbody) > 0:
self._addOrUpdateAlbum(soup)
else:
self._addOrUpdateFeaturedMedia(soup)
del webpage_content[first][second][i]
@ -99,6 +103,10 @@ class WPimport:
self._logger.debug("{0} : Number of article : {1}".format(self._name, len(articlebody)))
if len(articlebody) > 0:
self._addOrUpdatePost(soup)
else:
albumbody = soup.find_all("div", class_="albumbody")
if len(albumbody) > 0:
self._addOrUpdateAlbum(soup)
else:
self._addOrUpdateFeaturedMedia(soup)
@ -106,6 +114,9 @@ class WPimport:
# Private method
def _addOrUpdateAlbum(self, soup):
self._logger.info("{0} : Add/Update Album".format(self._name))
def _fromFileTmp(self):
try:
with open("{0}/{1}.json".format(self._tmp, self._name)) as file:

4
lib/WPMenu.py
View File

@ -86,8 +86,8 @@ class WPMenu:
idMenu = {"id":0, "type":"", "link":""}
soup = BeautifulSoup(content, self._parser)
articletitle = soup.find("h2", class_="articletitle").get_text()
if len(articletitle) > 0:
articletitle = soup.find("h2").get_text()
if len(articletitle) == 0:
articletitle = soup.find("div", class_="albumbody").find("h2").get_text()
exist = False
for index in range(1,10):
if exist is False: