Mirrors/gallery-dl
1
0
Fork 0
mirror of https://github.com/mikf/gallery-dl.git synced 2025-01-31 19:51:34 +01:00
Code Issues Releases Wiki Activity

[architizer] fix extraction (#4537)

Browse Source
This commit is contained in:
Mike Fährmann 2023-09-17 15:19:19 +02:00
parent 6178177227
commit cb4798f07a
No known key found for this signature in database
GPG Key ID: 5680CA389D365A88
1 changed files with 7 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
gallery_dl/extractor/architizer.py
View File

@ -29,11 +29,13 @@ class ArchitizerProjectExtractor(GalleryExtractor):
def metadata(self, page): def metadata(self, page):
extr = text.extract_from(page) extr = text.extract_from(page)
extr('id="Pages"', "")
return { return {
"title" : extr("data-name='", "'"), "title" : extr('data-name="', '"'),
"slug" : extr("data-slug='", "'"), "slug" : extr('data-slug="', '"'),
"gid" : extr("data-gid='", "'").rpartition(".")[2], "gid" : extr('data-gid="', '"').rpartition(".")[2],
"firm" : extr("data-firm-leaders-str='", "'"), "firm" : extr('data-firm-leaders-str="', '"'),
"location" : extr("<h2>", "<").strip(), "location" : extr("<h2>", "<").strip(),
"type" : text.unescape(text.remove_html(extr( "type" : text.unescape(text.remove_html(extr(
'<div class="title">Type</div>', '<br'))), '<div class="title">Type</div>', '<br'))),
@ -52,7 +54,7 @@ class ArchitizerProjectExtractor(GalleryExtractor):
return [ return [
(url, None) (url, None)
for url in text.extract_iter( for url in text.extract_iter(
page, "property='og:image:secure_url' content='", "?") page, 'property="og:image:secure_url" content="', "?")
] ]