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

add media in body html

Browse Source
This commit is contained in:
Valentin CZERYBA 2023-03-31 00:05:11 +02:00
parent c92f24e6af
commit 90881eb037
1 changed files with 11 additions and 13 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
insert_wordpress.py
View File

@ -49,6 +49,8 @@ class WPimport:
href_img = img[0].get("src")
page_img = requests.get(href_img)
img_break = False
new_img["old_src"]=href_img
new_img["old_href"]=href_a
if page_img.status_code == 404:
href_img = href_a
img_break = True
@ -70,11 +72,13 @@ class WPimport:
r = requests.post("http://{0}/wp-json/wp/v2/media".format(self.wordpress), auth=self.basic, headers=headers, data=data)
if r.status_code == 201:
res = r.json()
new_img["old_src"]=href_img
new_img["old_href"]=href_a
new_img["id"]=res["id"]
new_img["new_src"]=res["guid"]["rendered"]
id_res = res["id"]
rendered = res["guid"]["rendered"]
else:
id_res = res[0]["id"]
rendered = res[0]["guid"]["rendered"]
new_img["id"]=id_res
new_img["new_src"]=rendered
new_img["break"]=img_break
list_img.append(new_img)
@ -126,14 +130,8 @@ class WPimport:
bodyhtml = bodyhtml + "</p>"
for i in list_img:
o = urlparse(i["new_src"])
if i == True:
print(i["old_href"])
bodyhtml = bodyhtml.replace(i["old_href"], o.path)
else:
print(i["old_src"])
bodyhtml = bodyhtml.replace(i["old_src"], o.path)
print(bodyhtml)
exit(0)
bodyhtml = bodyhtml.replace(i["old_href"], o.path)
bodyhtml = bodyhtml.replace(i["old_src"], o.path)
hour = articledate[0].text
time = dateheader[0].text.split(" ")