def pixiv(page_url, options):
from pixivpy3 import AppPixivAPI
parsed = urlparse(page_url)
id = int(parsed.path.split("/")[-1])
api = AppPixivAPI()
secrets = h.get_secrets()["pixiv"]
api.login(secrets["username"], secrets["password"])
data = api.illust_detail(id)["illust"]
if len(data["meta_pages"]) == 0:
image_url = data["meta_single_page"]["original_image_url"]
elif options["album"]:
image_url = [
image["image_urls"]["original"] for image in data["meta_pages"]
]
else:
image_url = data["meta_pages"][
options["index"]]["image_urls"]["original"]
return Work(
data["title"],
(
data["user"]["account" if options["username"] else "name"],
data["user"]["name" if options["username"] else "account"],
),
None,
data["x_restrict"] > 0,
image_url,
page_url,
)
class Pixiv(DummySite):
def __init__(self, url, username, password, proxy=None):
proxies = get_proxy(proxy)
requests_kwargs = {
"timeout": (3, 10),
}
requests_kwargs.update(proxies)
self.api = AppPixivAPI(**requests_kwargs)
self._fetcher = PixivFetcher(**proxies)
self.api.login(username, password)
self._user_id = int(url.split("/")[-1])
self._dir_name = None
self._total_illustrations = 0
self._fetch_user_detail()
@property
def fetcher(self):
return self._fetcher
@property
def dir_name(self):
assert self._dir_name is not None
return self._dir_name
def _fetch_user_detail(self):
assert self._user_id is not None
profile = self.api.user_detail(self._user_id)
user = profile['user']
self._dir_name = "-".join([
user['name'],
user['account'],
str(user['id']),
])
self._dir_name = normalize_filename(self._dir_name)
self._total_illustrations = profile['profile']['total_illusts']
return self.dir_name
def _fetch_image_list(self, ):
ret = self.api.user_illusts(self._user_id)
while True:
for illustration in ret.illusts:
yield from parse_image_urls(illustration)
if ret.next_url is None:
break
ret = self.api.user_illusts(**self.api.parse_qs(ret.next_url))
def _fetch_single_image_url(self, illustration_id):
json_result = self.api.illust_detail(illustration_id)
illustration_info = json_result.illust
return illustration_info.image_urls['large']
@property
def tasks(self):
yield from self._fetch_image_list()
def pixiv_render(item, base_path, debug=False):
global pixiv_api
if pixiv_api is None:
pixiv_api = AppPixivAPI()
pixiv_api.login(pixiv_username, pixiv_password)
illust_id = get_illust_id(item.get_remote())
detail = pixiv_api.illust_detail(illust_id)
path = (str(detail['illust']['user']['name']) + '_' +
str(detail['illust']['user']['id']))
cpath(base_path + path)
urls = []
if detail['illust']['page_count'] > 1:
for page in detail['illust']['meta_pages']:
page_url = None
try:
page_url = page['image_urls']['original']
except (NameError, KeyError):
try:
page_url = list(page['image_urls'].values())[-1]
except (NameError, KeyError):
pass
if page_url is not None:
urls.append(page_url)
if len(urls) <= 0:
try:
urls.append(
detail['illust']['meta_single_page']['original_image_url'])
except (NameError, KeyError):
try:
urls.append(detail['illust']['image_urls']['large'])
except (NameError, KeyError):
pass
ret = []
for url in urls:
name = str(detail['illust']['title']) + '_' + str(
illust_id) + os.path.basename(url)
ret.append(path + '/' + name)
pixiv_api.download(url,
name=name,
path=os.path.abspath(base_path + path))
if debug: print('.', end='', flush=True)
return ret, detail
import json
import os
import re
from pixivpy3 import AppPixivAPI
with open('credentials.json') as cf:
credentials = json.load(cf)
with open('urls.txt') as uf:
urls = [u.split('/')[-1] for u in uf if 'pixiv.net' in u]
ids = [re.findall(r'\d+', id)[0] for id in urls]
api = AppPixivAPI()
api.login(credentials['email'], credentials['password'])
for id in ids:
json_result = api.illust_detail(id)
api.download(json_result.illust['meta_single_page']['original_image_url'])
class PixivDownloader:
def __init__(self,
client=None,
username=None,
password=None,
log_level=logging.WARNING):
if not client and (bool(username) != bool(password)):
raise AttributeError(
'If no client is given both username and password must be given'
)
if client:
self.api = client
else:
self.api = AppPixivAPI()
if not client and username and password:
self.api.login(username, password)
self.logger = logging.getLogger('PixivDownloader')
stdout = logging.StreamHandler()
self.logger.addHandler(stdout)
self.logger.setLevel(log_level)
def login(self, username=None, password=None, refresh_token=None):
if refresh_token:
self.logger.info('Loging in with refresh_token')
elif username:
self.logger.info('Loging in with username %s', username)
else:
self.logger.info('Loging')
return self.api.auth(username=username,
password=password,
refresh_token=refresh_token)
def logout(self):
self.logger.info('Logout')
self.api = AppPixivAPI()
def get_id_from_url(self, url):
path = urlparse(url).path
ids = re.findall('(\\d+)', path)
if not ids:
raise ValueError('Url does not contain post id')
return ids[0]
def download_by_id(self, post_id, output_dir):
data = self.api.illust_detail(post_id)
if data.get('error'):
raise PixivDownloaderError(
'Could not get post info or post doesn\'t exist.', data)
return self.download(data.illust, output_dir)
def download_by_url(self, url, output_dir):
return self.download_by_id(self.get_id_from_url(url), output_dir)
def download(self, post, output_dir):
output_dir = Path(output_dir).expanduser().absolute()
if not os.path.isdir(output_dir):
os.makedirs(output_dir)
self.logger.debug('Created dir "%s"', output_dir)
if post.type == 'illust' and not post.meta_pages:
downloader = self.download_illust
type = 'Image'
elif post.type == 'illust' and post.meta_pages:
downloader = self.download_illust_collection
type = 'Image Collection'
elif post.type == 'ugoira':
downloader = self.download_ugoira
type = 'Video'
elif post.type == 'manga':
downloader = self.download_manga
type = 'Manga'
else:
raise PixivDownloaderError(
f'Post type "{post.type}" not supported')
self.logger.info('Initialize "%s" downloader for post %s', type,
post.id)
return downloader(post, output_dir)
def download_illust(self, post, output_dir):
image_url = post.meta_single_page.get('original_image_url',
post.image_urls.large)
if '_webp' in image_url:
extension = 'webp'
else:
extension = os.path.splitext(image_url)[1].lstrip('.')
filename = self.get_filename(post, extension)
self.logger.info('Downloading "%s"', image_url)
self.api.download(image_url,
path=output_dir,
name=filename,
replace=True)
yield (Path(output_dir) / filename).absolute()
def download_illust_collection(self, post, output_dir):
output_dir = Path(output_dir)
yield from self._downloade_meta_pages(post, output_dir)
def download_manga(self, post, output_dir):
output_dir = Path(output_dir) / f'{post.title}-{post.user.account}'
if not output_dir.is_dir():
output_dir.mkdir(parents=True, exist_ok=True)
self.logger.debug('Created dir "%s"', output_dir)
yield from self._downloade_meta_pages(post, output_dir)
def _downloade_meta_pages(self, post, output_dir):
for index, image in enumerate(post.meta_pages, 1):
image_url = image.image_urls.get('original',
image.image_urls.large)
if '_webp' in image_url:
extension = 'webp'
else:
extension = os.path.splitext(image_url)[1].lstrip('.')
filename = self.get_filename(post,
extension,
suffix=f'-{index:0>2}')
self.logger.info('Downloading "%s"', image_url)
self.api.download(image_url,
path=str(output_dir),
name=filename,
replace=True)
yield (output_dir / filename).absolute()
def download_ugoira(self, post, output_dir):
ugoira_data = self.api.ugoira_metadata(post.id).ugoira_metadata
zip_url = ugoira_data.zip_urls.get('large',
ugoira_data.zip_urls.medium)
with TemporaryDirectory() as dir:
temp_dir = Path(dir)
filename = '{post.id}.zip'
self.logger.info('Downloading "%s"', zip_url)
self.api.download(zip_url, path=str(temp_dir), name=filename)
frames_dir = temp_dir / 'frames'
os.mkdir(frames_dir)
self._extract_zip(temp_dir / filename, frames_dir)
video_name = self.get_filename(post, 'mp4')
video_file = temp_dir / video_name
self._generate_mp4_from_frames(video_file, frames_dir,
ugoira_data.frames[0].delay)
final_path = (Path(output_dir) / video_name).absolute()
shutil.move(video_file, final_path)
yield final_path.absolute()
def get_filename(
self,
post,
extension,
prefix=None,
suffix=None,
):
suffix = suffix or ''
prefix = prefix or ''
filename = f'{prefix}{post.id}-{post.title}{suffix}.{extension}'.replace(
'/', '_').replace(' ', '_')
return filename
def _extract_zip(self, zip_file, output_dir):
self.logger.info('Extract "%s"', zip_file)
with ZipFile(zip_file, 'r') as zip_file:
zip_file.extractall(output_dir)
def _generate_mp4_from_frames(self, output_file, frames_dir, delay):
self.logger.info('Generate video to "%s"', output_file)
frames = sorted(
map(lambda file: os.path.join(str(frames_dir), file),
os.listdir(frames_dir)))
frames = list(map(imread, frames))
framerate = 1000 / delay
height, width, layers = frames[0].shape
video = VideoWriter(str(output_file), VideoWriter_fourcc(*'mp4v'),
framerate, (width, height))
for frame in frames:
video.write(frame)
destroyAllWindows()
video.release()
class APIHandler:
"""Singleton that handles all the API interactions in the program"""
def __init__(self):
self._api_thread = threading.Thread(target=self._login)
self._login_started = False
self._login_done = False
self._api = AppPixivAPI() # Object to login and request on
# Set in self.start() (because singleton is instantiated before config)
self._credentials: 'dict[str, str]'
self._response: 'Json'
def start(self, credentials):
"""Start logging in. The only setup entry point that is public"""
if not self._login_started:
self._credentials = credentials
self._api_thread.start()
self._login_started = True
def _await_login(self):
"""Wait for login to finish, then assign PixivAPI session to API"""
if not self._login_done:
self._api_thread.join()
self._login_done = True
def _login(self):
self._login_with_token()
def _login_with_token(self):
# TODO: refresh the token if it expired
try:
self._response = self._api.auth(
refresh_token=self._credentials['refresh_token'])
except PixivError as e:
print('')
print(e)
print('If this is a cloudflare captcha issue, just quit and retry')
print('It is not a problem with koneko or pixivpy')
print('Otherwise, please report to '
'https://github.com/akazukin5151/koneko/issues')
print("Press 'q' and enter to exit")
#else:
#print('Login success!')
# Public API requests for user id
def get_user_id(self) -> 'Json':
self._await_login()
return self._response['user']['id']
# Public API request functions for each mode
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
@utils.spinner('')
def artist_gallery(self, artist_user_id, offset) -> 'Json':
"""Mode 1"""
self._await_login()
return self._api.user_illusts(artist_user_id, offset=offset)
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
def protected_illust_detail(self, image_id) -> 'Json':
"""Mode 2"""
self._await_login()
return self._api.illust_detail(image_id)
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
def following_user_request(self, user_id, publicity, offset) -> 'Json':
"""Mode 3"""
self._await_login()
return self._api.user_following(user_id,
restrict=publicity,
offset=offset)
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
def search_user_request(self, searchstr, offset) -> 'Json':
"""Mode 4"""
self._await_login()
return self._api.search_user(searchstr, offset=offset)
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
@utils.spinner('')
def illust_follow_request(self, restrict, offset) -> 'Json':
"""Mode 5"""
self._await_login()
return self._api.illust_follow(restrict=restrict, offset=offset)
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
@utils.spinner('')
def illust_related_request(self, image_id, offset) -> 'Json':
"""Mode 15 (1.5 * 10 so it's an int)"""
self._await_login()
return self._api.illust_related(illust_id=image_id, offset=offset)
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
@utils.spinner('')
def illust_recommended_request(self, offset) -> 'Json':
"""Mode 6"""
self._await_login()
return self._api.illust_recommended(offset=offset)
# Download
@funcy.retry(tries=3, errors=(ConnectionError, PixivError))
def protected_download(self, url, path, name) -> 'IO':
"""Protect api download function with funcy.retry so it doesn't crash"""
self._await_login()
self._api.download(url, path=path, name=name)
class Pixiv:
def __init__(self):
self.enabled = config.getboolean('Pixiv', 'enabled', fallback=False)
self._refresh_token = config.get('Pixiv', 'refresh_token', fallback=None)
self._log = logging.getLogger(__name__)
self._pixiv = AppPixivAPI()
self._pixiv.set_accept_language(config.get('Pixiv', 'language', fallback='en-US'))
self._re_twitter = re.compile(r'^https?://(www.)?twitter.com/(?P<username>.+)$')
if self.enabled:
self._login()
def _login(self) -> None:
"""
Authenticate to Pixiv
Returns:
None
"""
self._log.debug(f'[PIXIV] Authenticating to Pixiv with the token {self._refresh_token}')
try:
self._pixiv.auth(refresh_token=self._refresh_token)
except Exception as error:
self._log.exception("[PIXIV] Failed to authenticate to Pixiv", exc_info=error)
def get_illust(self, illust_id: int) -> typing.Optional[dict]:
"""
Look up the provided illustration ID from SauceNao
Args:
illust_id (int):
Returns:
typing.Optional[dict]
"""
if not self.enabled:
return None
illust = self._pixiv.illust_detail(illust_id)
if 'error' in illust and 'invalid_grant' in illust['error']['message']:
self._log.warning(f'Pixiv login session is no longer valid; re-authenticating with saved token')
self._login()
illust = self._pixiv.illust_detail(illust_id)
return illust['illust'] if illust and 'illust' in illust else None
def get_author(self, author_id: int) -> typing.Optional[dict]:
"""
Get the author for the specified illustration
Args:
author_id (int):
Returns:
typing.Optional[dict]
"""
if not self.enabled:
return None
user = self._pixiv.user_detail(author_id)
if 'error' in user and 'invalid_grant' in user['error']['message']:
self._log.info(f'Re-Authenticating to Pixiv with the saved refresh token')
self._login()
user = self._pixiv.user_detail(author_id)
return user
def get_author_twitter(self, author_id: int) -> typing.Optional[str]:
"""
Get the Pixiv artists Twitter page, if available
Args:
author_id (int):
Returns:
typing.Optional[str]
"""
if not self.enabled:
return None
user = self.get_author(author_id)
twitter_url = user['profile']['twitter_url'] if (user and 'profile' in user) else None
if twitter_url:
match = self._re_twitter.match(twitter_url)
if match and match.group('username'):
return f"@{match.group('username')}"
def check_log(instance, **kwargs):
log = instance
title = 'test'
if log.nick != 'maobot' and log.command == 'PRIVMSG' and log.nick != 'maobot_php':
url_pat = re.compile(r"https?://[a-zA-Z0-9\-./?@&=:~_#]+")
url_list = re.findall(url_pat, log.message)
for url in url_list:
r = requests.get(url)
content_type_encoding = r.encoding if r.encoding != 'ISO-8859-1' else None
soup = BeautifulSoup(r.content, 'html.parser', from_encoding=content_type_encoding)
try:
title = soup.title.string
Log(command='NOTICE', channel=log.channel,
nick='maobot', message=title, is_irc=False).save()
except (AttributeError, TypeError, HTTPError):
pass
# image dl
nicoseiga_pat = re.compile(
'http://seiga.nicovideo.jp/seiga/[a-zA-Z]+([0-9]+)')
pixiv_pat = re.compile(
'https://www.pixiv.net/member_illust.php/?\?([a-zA-Z0-9\-./?@&=:~_#]+)')
twitter_pat = re.compile(
'https://twitter.com/[a-zA-Z0-9_]+/status/\d+')
image_format = ["jpg", "jpeg", "gif", "png"]
if twitter_pat.match(url):
try:
images = soup.find("div", {"class": "permalink-tweet-container"}).find("div", {"class": "AdaptiveMedia-container"}).findAll("div", {"class": "AdaptiveMedia-photoContainer"})
except AttributeError:
images = soup.findAll("div", {"class": "media"})
for image in images:
try:
image_url = image.find('img')['src']
img = image_from_response(requests.Session().get(image_url),
Image(original_url=url, related_log=log, caption=title))
img.save()
Log.objects.filter(id=log.id).update(attached_image=img.thumb)
except:
pass
elif nicoseiga_pat.match(url):
seiga_login = '******'
seiga_id = nicoseiga_pat.search(url).group(1)
seiga_source = 'http://seiga.nicovideo.jp/image/source/%s' % seiga_id
login_post = {'mail_tel': SECRET_KEYS['nicouser'],
'password': SECRET_KEYS['nicopass']}
session = requests.Session()
session.post(seiga_login, data=login_post)
soup = BeautifulSoup(session.get(seiga_source).text, 'lxml')
image_url = 'http://lohas.nicoseiga.jp%s' % soup.find(
'div', {'class': 'illust_view_big'})['data-src']
img = image_from_response(requests.Session().get(image_url),
Image(original_url=url, related_log=log))
img.save()
Log.objects.filter(id=log.id).update(attached_image=img.thumb)
elif pixiv_pat.match(url):
from pixivpy3 import AppPixivAPI
from urllib.parse import parse_qs
api = AppPixivAPI()
api.login(SECRET_KEYS['pixiuser'], SECRET_KEYS['pixipass'])
pixiv_query = pixiv_pat.search(url).group(1)
pixiv_dict = parse_qs(pixiv_query)
pixiv_id = pixiv_dict['illust_id']
pixiv_illust = api.illust_detail(pixiv_id, req_auth=True).illust
if 'meta_pages' in pixiv_illust and len(pixiv_illust.meta_pages) != 0:
image_urls = []
if 'page' in pixiv_dict:
image_urls.append(pixiv_illust.meta_pages[int(pixiv_dict['page'][0])].image_urls.large)
else:
for i in pixiv_illust.meta_pages:
image_urls.append(i.image_urls.large)
else:
image_urls = [pixiv_illust.image_urls.large]
for image_url in image_urls:
response = api.requests_call('GET', image_url,
headers={'Referer': 'https://app-api.pixiv.net/'},
stream=True)
img = image_from_response(response,
Image(original_url=url, related_log=log, caption=pixiv_illust.title))
img.save()
Log.objects.filter(id=log.id).update(attached_image=img.thumb)
elif url.split(".")[-1] in image_format:
img = image_from_response(requests.Session().get(url),
Image(original_url=url, related_log=log))
img.save()
Log.objects.filter(id=log.id).update(attached_image=img.thumb)
class CustomPixivPy:
"""
A wrapper around PixivAPI and AppPixivAPI to facilitate automatic re-authentication
(for required methods) and custom result format
"""
TOKEN_LIFESPAN = datetime.timedelta(seconds=3600)
MAX_PIXIV_RESULTS = 3000
RESULTS_PER_QUERY = 50
MAX_RETRIES = 5
def __init__(self, **kwargs):
super().__init__(**kwargs)
# forces reauth() to trigger if any method is called:
self.last_auth = datetime.datetime.fromtimestamp(0)
self.refresh_token = ""
self.aapi = AppPixivAPI(**kwargs)
self.papi = PixivAPI(**kwargs)
def login(self, refresh_token):
self.refresh_token = refresh_token
self.aapi.auth(refresh_token=refresh_token)
self.papi.auth(refresh_token=refresh_token)
self.last_auth = datetime.datetime.now()
logger.debug('Pyxiv login done')
return self # allows chaining
@retry
def illust_ranking(self, mode='daily', offset=None):
self.reauth()
offset = (offset or 0) // self.RESULTS_PER_QUERY + 1
return self.papi.ranking('illust',
mode,
offset,
include_stats=False,
image_sizes=['medium', 'large'])
@retry
def search_illust(self,
word,
search_target='text',
sort='date',
offset=None):
self.reauth()
offset = (offset or 0) // self.RESULTS_PER_QUERY + 1
return self.papi.search_works(word,
offset,
mode=search_target,
types=['illustration'],
sort=sort,
include_stats=False,
image_sizes=['medium', 'large'])
@retry
def illust_detail(self, illust_id, req_auth=True):
self.reauth()
return self.aapi.illust_detail(illust_id, req_auth)
def reauth(self):
"""Re-authenticates with pixiv if the last login was more than TOKEN_LIFESPAN ago"""
if datetime.datetime.now() - self.last_auth > self.TOKEN_LIFESPAN:
self.login(self.refresh_token)
self.papi.auth(refresh_token=self.refresh_token)
logger.debug("Reauth successful")
self.last_auth = datetime.datetime.now()
def get_pixiv_results(self, offset=None, *, query="", nsfw=False):
"""
Get results from Pixiv as a dict
If no parameters are given, SFW daily ranking is returned
:param offset: Optional. page offset
:param query: Optional. Specify a search query
:param nsfw: Whether to allow NSFW illustrations, false by default
:return: list of dicts containing illustration information
"""
json_result, last_error = None, None
for attempt in range(1, self.MAX_RETRIES + 1):
try:
json_result = self.search_illust(query, offset=offset, sort='popular') \
if query else self.illust_ranking('daily_r18' if nsfw else 'daily', offset=offset)
except PixivError as e:
if attempt == self.MAX_RETRIES:
logger.warning("Failed fetching Pixiv data: %s", e)
raise e from None
else:
break
results = []
if json_result.get('has_error'):
return results
it = json_result.response if query else (
x['work'] for x in json_result.response[0]['works'])
for img in it:
if not nsfw and img['sanity_level'] == 'black':
continue # white = SFW, semi_black = questionable, black = NSFW
results.append({
'url':
img['image_urls']['large'],
'thumb_url':
img['image_urls']['medium'],
'title':
img['title'],
'user_name':
img['user']['name'],
'user_link':
f"https://www.pixiv.net/en/users/{img['user']['id']}"
})
logger.debug(results[-1])
return results