mikf/gallery-dl
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[sankakucomplex] fix 'date' values (#258)

Browse Source
This commit is contained in:
Mike Fährmann 2019-05-27 23:58:58 +02:00
parent bcd1801aa8
commit 2a085a5e96
No known key found for this signature in database
GPG Key ID: 5680CA389D365A88
1 changed files with 8 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
gallery_dl/extractor/sankakucomplex.py
View File

@ -34,11 +34,11 @@ class SankakucomplexArticleExtractor(SankakucomplexExtractor):
test = ( test = (
("https://www.sankakucomplex.com/2019/05/11/twitter-cosplayers", { ("https://www.sankakucomplex.com/2019/05/11/twitter-cosplayers", {
"url": "4a9ecc5ae917fbce469280da5b6a482510cae84d", "url": "4a9ecc5ae917fbce469280da5b6a482510cae84d",
"keyword": "35cd2a0aba712d6b0e27a9fa2a5e823199d10ca0", "keyword": "4b3b5766b277a5d0acbec90fa8f2343262b07efd",
}), }),
("https://www.sankakucomplex.com/2009/12/01/sexy-goddesses-of-2ch", { ("https://www.sankakucomplex.com/2009/12/01/sexy-goddesses-of-2ch", {
"url": "a1e249173fd6c899a8134fcfbd9c925588a63f7c", "url": "a1e249173fd6c899a8134fcfbd9c925588a63f7c",
"keyword": "8bf60e62fb5e9f2caabb29c16ed58d7e0dcf247f", "keyword": "f47a416d680717855bbc3e4f0cd44479f61d9aa4",
}), }),
) )
@ -47,16 +47,17 @@ class SankakucomplexArticleExtractor(SankakucomplexExtractor):
extr = text.extract_from(self.request(url).text) extr = text.extract_from(self.request(url).text)
data = { data = {
"title" : text.unescape( "title" : text.unescape(
extr('"og:title" content="', '"')), extr('property="og:title" content="', '"')),
"description": text.unescape( "description": text.unescape(
extr('"og:description" content="', '"')), extr('property="og:description" content="', '"')),
"date" : text.parse_datetime( "date" : text.parse_datetime(
extr('"og:updated_time" content="', '"')), extr('property="article:published_time" content="', '"')),
} }
imgs = self.images(extr) imgs = self.images(extr)
data["count"] = len(imgs) data["count"] = len(imgs)
data["tags"] = text.split_html(extr('="meta-tags">', '</div>'))[::2] data["tags"] = text.split_html(extr('="meta-tags">', '</div>'))[::2]
yield Message.Version, 1
yield Message.Directory, data yield Message.Directory, data
for img in imgs: for img in imgs:
img.update(data) img.update(data)
@ -94,6 +95,7 @@ class SankakucomplexTagExtractor(SankakucomplexExtractor):
test = ( test = (
("https://www.sankakucomplex.com/tag/cosplay/", { ("https://www.sankakucomplex.com/tag/cosplay/", {
"range": "1-50", "range": "1-50",
"count": 50,
"pattern": SankakucomplexArticleExtractor.pattern, "pattern": SankakucomplexArticleExtractor.pattern,
}), }),
("https://www.sankakucomplex.com/category/anime/"), ("https://www.sankakucomplex.com/category/anime/"),
@ -105,6 +107,7 @@ class SankakucomplexTagExtractor(SankakucomplexExtractor):
last = None last = None
data = {"_extractor": SankakucomplexArticleExtractor} data = {"_extractor": SankakucomplexArticleExtractor}
yield Message.Version, 1
while True: while True:
url = "{}/{}/page/{}/".format(self.root, self.path, pnum) url = "{}/{}/page/{}/".format(self.root, self.path, pnum)
response = self.request(url, expect=(404,)) response = self.request(url, expect=(404,))
@ -114,5 +117,4 @@ class SankakucomplexTagExtractor(SankakucomplexExtractor):
if url != last: if url != last:
last = url last = url
yield Message.Queue, url, data yield Message.Queue, url, data
return
pnum += 1 pnum += 1