dmoonfire/dosage
1
0
Fork 0
Code Issues Pull requests Packages Projects Releases Activity

Fix SinFest with an ugly hack

Browse source
This commit is contained in:
Tobias Gruetzmacher 2022-06-06 14:52:40 +02:00
parent e5c505e643
commit 1d632904e4
1 changed files with 11 additions and 4 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

15
dosagelib/plugins/s.py
View file

@ -6,7 +6,7 @@
from re import compile, escape, IGNORECASE, sub from re import compile, escape, IGNORECASE, sub
from os.path import splitext from os.path import splitext
from ..scraper import _BasicScraper, _ParserScraper from ..scraper import _BasicScraper, _ParserScraper, ParserScraper
from ..helpers import indirectStarter, bounceStarter, joinPathPartsNamer from ..helpers import indirectStarter, bounceStarter, joinPathPartsNamer
from ..util import tagre from ..util import tagre
from .common import ComicControlScraper, WordPressScraper, WordPressNavi, WordPressWebcomic from .common import ComicControlScraper, WordPressScraper, WordPressNavi, WordPressWebcomic
@ -237,15 +237,22 @@ class ShotgunShuffle(WordPressScraper):
firstStripUrl = url + 'comic/pilot/' firstStripUrl = url + 'comic/pilot/'
class SinFest(_ParserScraper): class SinFest(ParserScraper):
url = 'https://www.sinfest.net/' END_HTML_TAG = compile(r'</html>')
url = 'https://sinfest.xyz/'
stripUrl = url + 'view.php?date=%s' stripUrl = url + 'view.php?date=%s'
firstStripUrl = stripUrl % '2000-01-17' firstStripUrl = stripUrl % '2000-01-17'
imageSearch = '//img[contains(@src, "btphp/comics/")]' imageSearch = '//img[contains(@src, "btphp/comics/")]'
textSearch = imageSearch + '/@alt' textSearch = imageSearch + '/@alt'
prevSearch = '//a[./img[contains(@src, "images/prev")]]' prevSearch = '//a[d:class("prev")]'
help = 'Index format: yyyy-mm-dd' help = 'Index format: yyyy-mm-dd'
# Remove HTML end tag confusing our parser
def _parse_page(self, data):
data = self.END_HTML_TAG.sub('', data)
return super()._parse_page(data)
class SisterClaire(ComicControlScraper): class SisterClaire(ComicControlScraper):
url = 'https://www.sisterclaire.com/comic/' url = 'https://www.sisterclaire.com/comic/'