1
0
mirror of https://github.com/mikf/gallery-dl.git synced 2024-11-23 11:12:40 +01:00
gallery-dl/gallery_dl/extractor/common.py

543 lines
18 KiB
Python
Raw Normal View History

2015-04-08 01:43:25 +02:00
# -*- coding: utf-8 -*-
# Copyright 2014-2020 Mike Fährmann
2015-04-08 01:43:25 +02:00
#
# 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.
"""Common classes and constants used by extractor modules."""
2017-07-24 18:33:42 +02:00
import re
2015-04-08 01:46:04 +02:00
import time
import netrc
2014-10-12 21:56:44 +02:00
import queue
2017-03-07 23:50:19 +01:00
import logging
import datetime
2014-10-12 21:56:44 +02:00
import requests
2015-04-08 01:46:04 +02:00
import threading
2015-11-24 19:47:51 +01:00
from .message import Message
from .. import config, text, util, exception, cloudflare
2015-04-08 01:46:04 +02:00
2015-04-08 01:43:25 +02:00
class Extractor():
category = ""
subcategory = ""
categorytransfer = False
directory_fmt = ("{category}",)
filename_fmt = "{filename}.{extension}"
2018-01-29 22:13:06 +01:00
archive_fmt = ""
cookiedomain = ""
2019-03-12 21:02:11 +01:00
root = ""
test = None
2015-11-29 23:41:43 +01:00
def __init__(self, match):
2015-04-08 01:43:25 +02:00
self.session = requests.Session()
2017-03-07 23:50:19 +01:00
self.log = logging.getLogger(self.category)
self.url = match.string
self._cookiefile = None
self._cookiejar = self.session.cookies
self._parentdir = ""
self._retries = self.config("retries", 4)
self._timeout = self.config("timeout", 30)
self._verify = self.config("verify", True)
2017-07-03 15:02:19 +02:00
2019-06-30 22:55:31 +02:00
if self._retries < 0:
self._retries = float("inf")
self._init_headers()
self._init_cookies()
self._init_proxies()
@classmethod
def from_url(cls, url):
if isinstance(cls.pattern, str):
cls.pattern = re.compile(cls.pattern)
match = cls.pattern.match(url)
return cls(match) if match else None
2014-10-12 21:56:44 +02:00
def __iter__(self):
2015-04-08 01:43:25 +02:00
return self.items()
def items(self):
yield Message.Version, 1
2014-10-12 21:56:44 +02:00
def skip(self, num):
return 0
def config(self, key, default=None):
return config.interpolate(
2019-11-23 23:50:16 +01:00
("extractor", self.category, self.subcategory), key, default)
def request(self, url, *, method="GET", session=None, retries=None,
encoding=None, fatal=True, notfound=None, **kwargs):
tries = 1
retries = self._retries if retries is None else retries
session = self.session if session is None else session
kwargs.setdefault("timeout", self._timeout)
kwargs.setdefault("verify", self._verify)
while True:
try:
response = session.request(method, url, **kwargs)
except (requests.exceptions.ConnectionError,
requests.exceptions.Timeout,
requests.exceptions.ChunkedEncodingError,
requests.exceptions.ContentDecodingError) as exc:
msg = exc
except (requests.exceptions.RequestException) as exc:
raise exception.HttpError(exc)
else:
code = response.status_code
if 200 <= code < 400 or fatal is None and \
(400 <= code < 500) or not fatal and \
(400 <= code < 429 or 431 <= code < 500):
if encoding:
response.encoding = encoding
return response
if notfound and code == 404:
raise exception.NotFoundError(notfound)
if cloudflare.is_challenge(response):
self.log.info("Solving Cloudflare challenge")
2019-12-01 01:07:24 +01:00
response, domain, cookies = cloudflare.solve_challenge(
2019-03-14 16:14:29 +01:00
session, response, kwargs)
2019-12-01 01:07:24 +01:00
if response.status_code >= 400:
continue
2019-03-14 16:14:29 +01:00
cloudflare.cookies.update(self.category, (domain, cookies))
2019-12-01 01:07:24 +01:00
return response
if cloudflare.is_captcha(response):
self.log.warning("Cloudflare CAPTCHA")
msg = "'{} {}' for '{}'".format(code, response.reason, url)
if code < 500 and code != 429 and code != 430:
break
self.log.debug("%s (%s/%s)", msg, tries, retries+1)
if tries > retries:
break
time.sleep(min(2 ** (tries-1), 1800))
tries += 1
raise exception.HttpError(msg)
2017-07-25 14:59:41 +02:00
2020-01-04 23:21:45 +01:00
def wait(self, *, seconds=None, until=None, reason=None, adjust=1):
now = datetime.datetime.now()
if seconds:
seconds = float(seconds)
until = now + datetime.timedelta(seconds=seconds)
elif until:
until = datetime.datetime.fromtimestamp(float(until))
seconds = (until - now).total_seconds()
else:
raise ValueError("Either 'seconds' or 'until' is required")
if reason:
t = until.time()
isotime = "{:02}:{:02}:{:02}".format(t.hour, t.minute, t.second)
2020-01-04 23:21:45 +01:00
self.log.info("Waiting until %s for %s.", isotime, reason)
time.sleep(seconds + adjust)
2017-07-25 14:59:41 +02:00
def _get_auth_info(self):
"""Return authentication information as (username, password) tuple"""
username = self.config("username")
password = None
if username:
password = self.config("password")
elif self.config("netrc", False):
try:
info = netrc.netrc().authenticators(self.category)
username, _, password = info
except (OSError, netrc.NetrcParseError) as exc:
self.log.error("netrc: %s", exc)
except TypeError:
self.log.warning("netrc: No authentication info")
return username, password
2019-03-12 21:02:11 +01:00
def _init_headers(self):
"""Initialize HTTP headers for the 'session' object"""
headers = self.session.headers
headers.clear()
headers["User-Agent"] = self.config(
"user-agent", ("Mozilla/5.0 (X11; Linux x86_64; rv:68.0) "
"Gecko/20100101 Firefox/68.0"))
headers["Accept"] = "*/*"
headers["Accept-Language"] = "en-US,en;q=0.5"
headers["Accept-Encoding"] = "gzip, deflate"
headers["Connection"] = "keep-alive"
headers["Upgrade-Insecure-Requests"] = "1"
2019-03-12 21:02:11 +01:00
def _init_proxies(self):
2019-01-30 17:09:32 +01:00
"""Update the session's proxy map"""
proxies = self.config("proxy")
if proxies:
if isinstance(proxies, str):
proxies = {"http": proxies, "https": proxies}
if isinstance(proxies, dict):
for scheme, proxy in proxies.items():
if "://" not in proxy:
proxies[scheme] = "http://" + proxy.lstrip("/")
self.session.proxies = proxies
else:
self.log.warning("invalid proxy specifier: %s", proxies)
2019-03-12 21:02:11 +01:00
def _init_cookies(self):
"""Populate the session's cookiejar"""
if self.cookiedomain is None:
return
cookies = self.config("cookies")
2017-07-25 14:59:41 +02:00
if cookies:
if isinstance(cookies, dict):
2019-03-12 21:02:11 +01:00
self._update_cookies_dict(cookies, self.cookiedomain)
elif isinstance(cookies, str):
cookiefile = util.expand_path(cookies)
2017-07-25 14:59:41 +02:00
try:
with open(cookiefile) as fp:
cookies = util.load_cookiestxt(fp)
except Exception as exc:
2017-07-25 14:59:41 +02:00
self.log.warning("cookies: %s", exc)
2019-03-12 21:02:11 +01:00
else:
self._update_cookies(cookies)
self._cookiefile = cookiefile
else:
self.log.warning(
"expected 'dict' or 'str' value for 'cookies' option, "
"got '%s' (%s)", cookies.__class__.__name__, cookies)
2019-03-12 21:02:11 +01:00
2019-03-14 16:14:29 +01:00
cookies = cloudflare.cookies(self.category)
if cookies:
domain, cookies = cookies
self._update_cookies_dict(cookies, domain)
def _store_cookies(self):
"""Store the session's cookiejar in a cookies.txt file"""
if self._cookiefile and self.config("cookies-update", True):
try:
with open(self._cookiefile, "w") as fp:
util.save_cookiestxt(fp, self._cookiejar)
except OSError as exc:
self.log.warning("cookies: %s", exc)
2019-03-12 21:02:11 +01:00
def _update_cookies(self, cookies, *, domain=""):
"""Update the session's cookiejar with 'cookies'"""
if isinstance(cookies, dict):
self._update_cookies_dict(cookies, domain or self.cookiedomain)
else:
setcookie = self._cookiejar.set_cookie
2019-03-12 21:02:11 +01:00
try:
cookies = iter(cookies)
except TypeError:
setcookie(cookies)
else:
for cookie in cookies:
setcookie(cookie)
def _update_cookies_dict(self, cookiedict, domain):
"""Update cookiejar with name-value pairs from a dict"""
setcookie = self._cookiejar.set
2019-03-12 21:02:11 +01:00
for name, value in cookiedict.items():
setcookie(name, value, domain=domain)
2014-10-12 21:56:44 +02:00
def _check_cookies(self, cookienames, *, domain=None):
2017-07-25 14:59:41 +02:00
"""Check if all 'cookienames' are in the session's cookiejar"""
2020-02-24 23:52:05 +01:00
if not self._cookiejar:
return False
if domain is None:
domain = self.cookiedomain
names = set(cookienames)
2020-02-24 23:52:05 +01:00
now = time.time()
for cookie in self._cookiejar:
2020-02-24 23:52:05 +01:00
if cookie.name in names and cookie.domain == domain:
if cookie.expires and cookie.expires < now:
self.log.warning("Cookie '%s' has expired", cookie.name)
else:
names.discard(cookie.name)
if not names:
return True
return False
2017-07-17 10:33:36 +02:00
def _get_date_min_max(self, dmin=None, dmax=None):
"""Retrieve and parse 'date-min' and 'date-max' config values"""
def get(key, default):
ts = self.config(key, default)
if isinstance(ts, str):
try:
ts = int(datetime.datetime.strptime(ts, fmt).timestamp())
except ValueError as exc:
self.log.warning("Unable to parse '%s': %s", key, exc)
ts = default
return ts
fmt = self.config("date-format", "%Y-%m-%dT%H:%M:%S")
return get("date-min", dmin), get("date-max", dmax)
def _dispatch_extractors(self, extractor_data, default=()):
""" """
extractors = {
data[0].subcategory: data
for data in extractor_data
}
include = self.config("include", default) or ()
if include == "all":
include = extractors
elif isinstance(include, str):
include = include.split(",")
result = [(Message.Version, 1)]
for category in include:
if category in extractors:
extr, url = extractors[category]
result.append((Message.Queue, url, {"_extractor": extr}))
return iter(result)
@classmethod
def _get_tests(cls):
"""Yield an extractor's test cases as (URL, RESULTS) tuples"""
2019-03-12 21:02:11 +01:00
tests = cls.test
if not tests:
return
if len(tests) == 2 and (not tests[1] or isinstance(tests[1], dict)):
tests = (tests,)
for test in tests:
if isinstance(test, str):
test = (test, None)
yield test
2015-04-08 01:43:25 +02:00
class GalleryExtractor(Extractor):
subcategory = "gallery"
filename_fmt = "{category}_{gallery_id}_{num:>03}.{extension}"
directory_fmt = ("{category}", "{gallery_id} {title}")
archive_fmt = "{gallery_id}_{num}"
enum = "num"
def __init__(self, match, url=None):
Extractor.__init__(self, match)
self.gallery_url = self.root + match.group(1) if url is None else url
def items(self):
self.login()
page = self.request(self.gallery_url, notfound=self.subcategory).text
data = self.metadata(page)
imgs = self.images(page)
if "count" in data:
images = zip(
range(1, data["count"]+1),
imgs,
)
else:
try:
data["count"] = len(imgs)
except TypeError:
pass
images = enumerate(imgs, 1)
yield Message.Version, 1
yield Message.Directory, data
for data[self.enum], (url, imgdata) in images:
if imgdata:
data.update(imgdata)
if "extension" not in imgdata:
text.nameext_from_url(url, data)
else:
text.nameext_from_url(url, data)
yield Message.Url, url, data
def login(self):
"""Login and set necessary cookies"""
def metadata(self, page):
"""Return a dict with general metadata"""
def images(self, page):
"""Return a list of all (image-url, metadata)-tuples"""
class ChapterExtractor(GalleryExtractor):
subcategory = "chapter"
directory_fmt = (
"{category}", "{manga}",
"{volume:?v/ />02}c{chapter:>03}{chapter_minor:?//}{title:?: //}")
filename_fmt = (
"{manga}_c{chapter:>03}{chapter_minor:?//}_{page:>03}.{extension}")
archive_fmt = (
"{manga}_{chapter}{chapter_minor}_{page}")
enum = "page"
class MangaExtractor(Extractor):
subcategory = "manga"
categorytransfer = True
chapterclass = None
reverse = True
def __init__(self, match, url=None):
Extractor.__init__(self, match)
self.manga_url = url or self.root + match.group(1)
if self.config("chapter-reverse", False):
self.reverse = not self.reverse
def items(self):
self.login()
page = self.request(self.manga_url).text
chapters = self.chapters(page)
if self.reverse:
chapters.reverse()
yield Message.Version, 1
2017-09-24 16:03:29 +02:00
for chapter, data in chapters:
data["_extractor"] = self.chapterclass
2017-09-24 16:03:29 +02:00
yield Message.Queue, chapter, data
def login(self):
"""Login and set necessary cookies"""
def chapters(self, page):
"""Return a list of all (chapter-url, metadata)-tuples"""
class AsynchronousMixin():
"""Run info extraction in a separate thread"""
def __iter__(self):
messages = queue.Queue(5)
thread = threading.Thread(
target=self.async_items,
args=(messages,),
daemon=True,
)
thread.start()
while True:
msg = messages.get()
if msg is None:
thread.join()
return
if isinstance(msg, Exception):
thread.join()
raise msg
yield msg
messages.task_done()
def async_items(self, messages):
try:
for msg in self.items():
messages.put(msg)
except Exception as exc:
messages.put(exc)
messages.put(None)
class SharedConfigMixin():
"""Enable sharing of config settings based on 'basecategory'"""
basecategory = ""
def config(self, key, default=None, *, sentinel=object()):
value = Extractor.config(self, key, sentinel)
return value if value is not sentinel else config.interpolate(
("extractor", self.basecategory, self.subcategory), key, default)
2019-03-07 22:55:26 +01:00
def generate_extractors(extractor_data, symtable, classes):
"""Dynamically generate Extractor classes"""
2019-11-23 23:50:16 +01:00
extractors = config.get(("extractor",), classes[0].basecategory)
2019-03-07 22:55:26 +01:00
ckey = extractor_data.get("_ckey")
prev = None
if extractors:
extractor_data.update(extractors)
for category, info in extractor_data.items():
if not isinstance(info, dict):
continue
root = info["root"]
domain = root[root.index(":") + 3:]
pattern = info.get("pattern") or re.escape(domain)
name = (info.get("name") or category).capitalize()
for cls in classes:
class Extr(cls):
pass
Extr.__module__ = cls.__module__
Extr.__name__ = Extr.__qualname__ = \
name + cls.subcategory.capitalize() + "Extractor"
Extr.__doc__ = \
"Extractor for " + cls.subcategory + "s from " + domain
Extr.category = category
Extr.pattern = r"(?:https?://)?" + pattern + cls.pattern_fmt
Extr.test = info.get("test-" + cls.subcategory)
Extr.root = root
if "extra" in info:
for key, value in info["extra"].items():
setattr(Extr, key, value)
if prev and ckey:
setattr(Extr, ckey, prev)
symtable[Extr.__name__] = prev = Extr
# Undo automatic pyOpenSSL injection by requests
pyopenssl = config.get((), "pyopenssl", False)
if not pyopenssl:
try:
from requests.packages.urllib3.contrib import pyopenssl # noqa
pyopenssl.extract_from_urllib3()
except ImportError:
pass
del pyopenssl
# Replace urllib3's default cipher list to avoid Cloudflare CAPTCHAs
2019-11-23 23:50:16 +01:00
ciphers = config.get((), "ciphers", True)
if ciphers:
if ciphers is True:
ciphers = (
# Firefox's list
"TLS_AES_128_GCM_SHA256:"
"TLS_CHACHA20_POLY1305_SHA256:"
"TLS_AES_256_GCM_SHA384:"
"ECDHE-ECDSA-AES128-GCM-SHA256:"
"ECDHE-RSA-AES128-GCM-SHA256:"
"ECDHE-ECDSA-CHACHA20-POLY1305:"
"ECDHE-RSA-CHACHA20-POLY1305:"
"ECDHE-ECDSA-AES256-GCM-SHA384:"
"ECDHE-RSA-AES256-GCM-SHA384:"
"ECDHE-ECDSA-AES256-SHA:"
"ECDHE-ECDSA-AES128-SHA:"
"ECDHE-RSA-AES128-SHA:"
"ECDHE-RSA-AES256-SHA:"
"DHE-RSA-AES128-SHA:"
"DHE-RSA-AES256-SHA:"
"AES128-SHA:"
"AES256-SHA:"
"DES-CBC3-SHA"
)
elif isinstance(ciphers, list):
ciphers = ":".join(ciphers)
from requests.packages.urllib3.util import ssl_ # noqa
ssl_.DEFAULT_CIPHERS = ciphers
del ssl_
del ciphers