1
0
mirror of https://github.com/mikf/gallery-dl.git synced 2024-11-24 11:42:33 +01:00
gallery-dl/gallery_dl/extractor/recursive.py
2016-10-01 15:54:27 +02:00

31 lines
867 B
Python

# -*- coding: utf-8 -*-
# Copyright 2015, 2016 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.
"""Recursive extractor"""
import re
from .common import Extractor, Message
class RecursiveExtractor(Extractor):
category = "recursive"
pattern = ["r(?:ecursive)?:(.+)"]
test = [("recursive:https://pastebin.com/raw/FLwrCYsT", {
"url": "eee86d65c346361b818e8f4b2b307d9429f136a2",
})]
def __init__(self, match):
Extractor.__init__(self)
self.url = match.group(1)
def items(self):
page = self.request(self.url).text
yield Message.Version, 1
for match in re.finditer(r"https?://[^\s\"']+", page):
yield Message.Queue, match.group(0)