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

[2chan] fix metadata extraction

Browse Source
This commit is contained in:
Mike Fährmann 2019-12-03 16:53:08 +01:00
parent 173a93454e
commit 71acbdabf4
No known key found for this signature in database
GPG Key ID: 5680CA389D365A88
2 changed files with 7 additions and 5 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

10
gallery_dl/extractor/2chan.py
View File

@ -68,6 +68,8 @@ class _2chanThreadExtractor(Extractor):
def parse(self, post): def parse(self, post):
"""Build post-object by extracting data from an HTML post""" """Build post-object by extracting data from an HTML post"""
data = self._extract_post(post) data = self._extract_post(post)
if data["name"]:
data["name"] = data["name"].strip()
if '<a href="/' in post: if '<a href="/' in post:
self._extract_image(post, data) self._extract_image(post, data)
data["tim"], _, data["extension"] = data["filename"].partition(".") data["tim"], _, data["extension"] = data["filename"].partition(".")
@ -78,10 +80,10 @@ class _2chanThreadExtractor(Extractor):
@staticmethod @staticmethod
def _extract_post(post): def _extract_post(post):
return text.extract_all(post, ( return text.extract_all(post, (
("no" , 'name="', '"'), ("post", 'class="csb">' , '<'),
("post", '<b>', '</b>'), ("name", 'class="cnm">' , '<'),
("name", '<b>', ' </b>'), ("now" , 'class="cnw">' , '<'),
("now" , '</font> ', ' '), ("no" , 'class="cno">No.', '<'),
(None , '<blockquote', ''), (None , '<blockquote', ''),
("com" , '>', '</blockquote>'), ("com" , '>', '</blockquote>'),
))[0] ))[0]

2
gallery_dl/extractor/wikiart.py
View File

@ -93,7 +93,7 @@ class WikiartArtworksExtractor(WikiartExtractor):
directory_fmt = ("{category}", "Artworks by {group!c}", "{type}") directory_fmt = ("{category}", "Artworks by {group!c}", "{type}")
pattern = BASE_PATTERN + r"/paintings-by-([\w-]+)/([\w-]+)" pattern = BASE_PATTERN + r"/paintings-by-([\w-]+)/([\w-]+)"
test = ("https://www.wikiart.org/en/paintings-by-media/grisaille", { test = ("https://www.wikiart.org/en/paintings-by-media/grisaille", {
"url": "f92d55669fa949491c26a5437527adb14b35b8cc", "url": "228426a9d32b5bba9d659944c6b0ba73883af33f",
}) })
def __init__(self, match): def __init__(self, match):