# -*- coding: utf-8 -*- # Copyright 2017-2020 Mike Fährmann # # This program is free software; you can redistribute it and/or modify # it under the terms of the GNU General Public License version 2 as # published by the Free Software Foundation. """Extractors for https://www.2chan.net/""" from .common import Extractor, Message from .. import text class _2chanThreadExtractor(Extractor): """Extractor for images from threads on www.2chan.net""" category = "2chan" subcategory = "thread" directory_fmt = ("{category}", "{board_name}", "{thread}") filename_fmt = "{tim}.{extension}" archive_fmt = "{board}_{thread}_{tim}" url_fmt = "https://{server}.2chan.net/{board}/src/{filename}" pattern = r"(?:https?://)?([^.]+)\.2chan\.net/([^/]+)/res/(\d+)" test = ("http://dec.2chan.net/70/res/4752.htm", { "url": "1c2d2ce8aea0fc71d94735cfc30009d628f33548", "keyword": "f508d6841ea2cb19ed799aac9dc580263ca50651", }) def __init__(self, match): Extractor.__init__(self, match) self.server, self.board, self.thread = match.groups() def items(self): url = "https://{}.2chan.net/{}/res/{}.htm".format( self.server, self.board, self.thread) page = self.request(url).text data = self.metadata(page) yield Message.Version, 1 yield Message.Directory, data for post in self.posts(page): if "filename" not in post: continue post.update(data) url = self.url_fmt.format_map(post) yield Message.Url, url, post def metadata(self, page): """Collect metadata for extractor-job""" title = text.extract(page, "