Mirrors/gallery-dl
1
0
Fork 0
mirror of https://github.com/mikf/gallery-dl.git synced 2024-11-22 10:42:34 +01:00
Code Issues Releases Wiki Activity

[8chan] fix image urls

Browse Source
This commit is contained in:
Mike Fährmann 2015-02-03 00:22:39 +01:00
parent 54d40b85fb
commit ac002cc4b3
1 changed files with 4 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

5
gallery_dl/extractor/8chan.py
View File

@ -4,7 +4,8 @@ import re
class Extractor(BasicExtractor): class Extractor(BasicExtractor):
thread_url_fmt = "https://www.8ch.net/{0}/res/{1}.html" url_base = "https://8ch.net"
thread_url_fmt = url_base + "/{0}/res/{1}.html"
regex = r'>File: <a href="([^"]+)">([^<]+)\.[^<]+<.*?<span class="postfilename"( title="([^"]+)")?>([^<]+)<' regex = r'>File: <a href="([^"]+)">([^<]+)\.[^<]+<.*?<span class="postfilename"( title="([^"]+)")?>([^<]+)<'
def __init__(self, match, config): def __init__(self, match, config):
@ -18,4 +19,6 @@ class Extractor(BasicExtractor):
text = self.request(url).text text = self.request(url).text
for match in re.finditer(self.regex, text): for match in re.finditer(self.regex, text):
url, prefix, fullname, name = match.group(1, 2, 4, 5) url, prefix, fullname, name = match.group(1, 2, 4, 5)
if url.startswith("/"):
url = self.url_base + url
yield (url, prefix + "-" + unquote(fullname or name)) yield (url, prefix + "-" + unquote(fullname or name))