14601512_黑帮往事（上）

admin

"""
This module provides a pool manager that uses Google App Engine's
`URLFetch Service <https://cloud.google.com/appengine/docs/python/urlfetch>`_.

Example usage::

    from pip._vendor.urllib3 import PoolManager
    from pip._vendor.urllib3.contrib.appengine import AppEngineManager, is_appengine_sandbox

    if is_appengine_sandbox():
        # AppEngineManager uses AppEngine's URLFetch API behind the scenes
        http = AppEngineManager()
    else:
        # PoolManager uses a socket-level API behind the scenes
        http = PoolManager()

    r = http.request('GET', 'https://google.com/')

There are `limitations <https://cloud.google.com/appengine/docs/python/\
urlfetch/#Python_Quotas_and_limits>`_ to the URLFetch service and it may not be
the best choice for your application. There are three options for using
urllib3 on Google App Engine:

1. You can use :class:`AppEngineManager` with URLFetch. URLFetch is
   cost-effective in many circumstances as long as your usage is within the
   limitations.
2. You can use a normal :class:`~urllib3.PoolManager` by enabling sockets.
   Sockets also have `limitations and restrictions
   <https://cloud.google.com/appengine/docs/python/sockets/\
   #limitations-and-restrictions>`_ and have a lower free quota than URLFetch.
   To use sockets, be sure to specify the following in your ``app.yaml``::

        env_variables:
            GAE_USE_SOCKETS_HTTPLIB : 'true'

3. If you are using `App Engine Flexible
<https://cloud.google.com/appengine/docs/flexible/>`_, you can use the standard
:class:`PoolManager` without any configuration or special environment variables.
"""

from __future__ import absolute_import

import io
import logging
import warnings

from ..exceptions import (
    HTTPError,
    HTTPWarning,
    MaxRetryError,
    ProtocolError,
    SSLError,
    TimeoutError,
)
from ..packages.six.moves.urllib.parse import urljoin
from ..request import RequestMethods
from ..response import HTTPResponse
from ..util.retry import Retry
from ..util.timeout import Timeout
from . import _appengine_environ

try:
    from google.appengine.api import urlfetch
except ImportError:
    urlfetch = None


log = logging.getLogger(__name__)


class AppEnginePlatformWarning(HTTPWarning):
    pass


class AppEnginePlatformError(HTTPError):
    pass


class AppEngineManager(RequestMethods):
    """
    Connection manager for Google App Engine sandbox applications.

    This manager uses the URLFetch service directly instead of using the
    emulated httplib, and is subject to URLFetch limitations as described in
    the App Engine documentation `here
    <https://cloud.google.com/appengine/docs/python/urlfetch>`_.

    Notably it will raise an :class:`AppEnginePlatformError` if:
        * URLFetch is not available.
        * If you attempt to use this on App Engine Flexible, as full socket
          support is available.
        * If a request size is more than 10 megabytes.
        * If a response size is more than 32 megabytes.
        * If you use an unsupported request method such as OPTIONS.

    Beyond those cases, it will raise normal urllib3 errors.
    """

    def __init__(
        self,
        headers=None,
        retries=None,
        validate_certificate=True,
        urlfetch_retries=True,
    ):
        if not urlfetch:
            raise AppEnginePlatformError(
                "URLFetch is not available in this environment."
            )

        warnings.warn(
            "urllib3 is using URLFetch on Google App Engine sandbox instead "
            "of sockets. To use sockets directly instead of URLFetch see "
            "https://urllib3.readthedocs.io/en/1.26.x/reference/urllib3.contrib.html.",
            AppEnginePlatformWarning,
        )

        RequestMethods.__init__(self, headers)
        self.validate_certificate = validate_certificate
        self.urlfetch_retries = urlfetch_retries

        self.retries = retries or Retry.DEFAULT

    def __enter__(self):
        return self

    def __exit__(self, exc_type, exc_val, exc_tb):
        # Return False to re-raise any potential exceptions
        return False

    def urlopen(
        self,
        method,
        url,
        body=None,
        headers=None,
        retries=None,
        redirect=True,
        timeout=Timeout.DEFAULT_TIMEOUT,
        **response_kw
    ):

        retries = self._get_retries(retries, redirect)

        try:
            follow_redirects = redirect and retries.redirect != 0 and retries.total
            response = urlfetch.fetch(
                url,
                payload=body,
                method=method,
                headers=headers or {},
                allow_truncated=False,
                follow_redirects=self.urlfetch_retries and follow_redirects,
                deadline=self._get_absolute_timeout(timeout),
                validate_certificate=self.validate_certificate,
            )
        except urlfetch.DeadlineExceededError as e:
            raise TimeoutError(self, e)

        except urlfetch.InvalidURLError as e:
            if "too large" in str(e):
                raise AppEnginePlatformError(
                    "URLFetch request too large, URLFetch only "
                    "supports requests up to 10mb in size.",
                    e,
                )
            raise ProtocolError(e)

        except urlfetch.DownloadError as e:
            if "Too many redirects" in str(e):
                raise MaxRetryError(self, url, reason=e)
            raise ProtocolError(e)

        except urlfetch.ResponseTooLargeError as e:
            raise AppEnginePlatformError(
                "URLFetch response too large, URLFetch only supports"
                "responses up to 32mb in size.",
                e,
            )

        except urlfetch.SSLCertificateError as e:
            raise SSLError(e)

        except urlfetch.InvalidMethodError as e:
            raise AppEnginePlatformError(
                "URLFetch does not support method: %s" % method, e
            )

        http_response = self._urlfetch_response_to_http_response(
            response, retries=retries, **response_kw
        )

        # Handle redirect?
        redirect_location = redirect and http_response.get_redirect_location()
        if redirect_location:
            # Check for redirect response
            if self.urlfetch_retries and retries.raise_on_redirect:
                raise MaxRetryError(self, url, "too many redirects")
            else:
                if http_response.status == 303:
                    method = "GET"

                try:
                    retries = retries.increment(
                        method, url, response=http_response, _pool=self
                    )
                except MaxRetryError:
                    if retries.raise_on_redirect:
                        raise MaxRetryError(self, url, "too many redirects")
                    return http_response

                retries.sleep_for_retry(http_response)
                log.debug("Redirecting %s -> %s", url, redirect_location)
                redirect_url = urljoin(url, redirect_location)
                return self.urlopen(
                    method,
                    redirect_url,
                    body,
                    headers,
                    retries=retries,
                    redirect=redirect,
                    timeout=timeout,
                    **response_kw
                )

        # Check if we should retry the HTTP response.
        has_retry_after = bool(http_response.headers.get("Retry-After"))
        if retries.is_retry(method, http_response.status, has_retry_after):
            retries = retries.increment(method, url, response=http_response, _pool=self)
            log.debug("Retry: %s", url)
            retries.sleep(http_response)
            return self.urlopen(
                method,
                url,
                body=body,
                headers=headers,
                retries=retries,
                redirect=redirect,
                timeout=timeout,
                **response_kw
            )

        return http_response

    def _urlfetch_response_to_http_response(self, urlfetch_resp, **response_kw):

        if is_prod_appengine():
            # Production GAE handles deflate encoding automatically, but does
            # not remove the encoding header.
            content_encoding = urlfetch_resp.headers.get("content-encoding")

            if content_encoding == "deflate":
                del urlfetch_resp.headers["content-encoding"]

        transfer_encoding = urlfetch_resp.headers.get("transfer-encoding")
        # We have a full response's content,
        # so let's make sure we don't report ourselves as chunked data.
        if transfer_encoding == "chunked":
            encodings = transfer_encoding.split(",")
            encodings.remove("chunked")
            urlfetch_resp.headers["transfer-encoding"] = ",".join(encodings)

        original_response = HTTPResponse(
            # In order for decoding to work, we must present the content as
            # a file-like object.
            body=io.BytesIO(urlfetch_resp.content),
            msg=urlfetch_resp.header_msg,
            headers=urlfetch_resp.headers,
            status=urlfetch_resp.status_code,
            **response_kw
        )

        return HTTPResponse(
            body=io.BytesIO(urlfetch_resp.content),
            headers=urlfetch_resp.headers,
            status=urlfetch_resp.status_code,
            original_response=original_response,
            **response_kw
        )

    def _get_absolute_timeout(self, timeout):
        if timeout is Timeout.DEFAULT_TIMEOUT:
            return None  # Defer to URLFetch's default.
        if isinstance(timeout, Timeout):
            if timeout._read is not None or timeout._connect is not None:
                warnings.warn(
                    "URLFetch does not support granular timeout settings, "
                    "reverting to total or default URLFetch timeout.",
                    AppEnginePlatformWarning,
                )
            return timeout.total
        return timeout

    def _get_retries(self, retries, redirect):
        if not isinstance(retries, Retry):
            retries = Retry.from_int(retries, redirect=redirect, default=self.retries)

        if retries.connect or retries.read or retries.redirect:
            warnings.warn(
                "URLFetch only supports total retries and does not "
                "recognize connect, read, or redirect retry parameters.",
                AppEnginePlatformWarning,
            )

        return retries


# Alias methods from _appengine_environ to maintain public API interface.

is_appengine = _appengine_environ.is_appengine
is_appengine_sandbox = _appengine_environ.is_appengine_sandbox
is_local_appengine = _appengine_environ.is_local_appengine
is_prod_appengine = _appengine_environ.is_prod_appengine
is_prod_appengine_mvms = _appengine_environ.is_prod_appengine_mvms
离开凳子还想闪躲，却低估了持续不断的春药的效力，当即脚下一软，跪在地上，紧接着赵宜之的鞭子劈头盖脸打过来，打在胸部、小腹、大腿、之类的敏感处。想象中的剧痛没有到来，只有暧昧的疼痛感，在痛苦中夹杂着快感....那份深藏在心底的记忆逐渐清晰起来。澜不再闪躲，任由赵宜之发泄般把鞭子抽打在自己身上，低下头抵着地面，屁股高高翘起，发出狗一般细碎的呜咽。
　　“以前你被调教过吧？”仿佛被击碎了幻想，澜一下子爬起来，捂着胸部又回到了怒目而视的样子：只是那发红的脸颊和轻微的喘息，配上满身的鞭痕确实是缺少了作为一方大佬的威势。赵宜之扔掉鞭子，冲过去掐住澜的脖子，把她死死压在地面上，在惊怒的脸上扇了一巴掌：“原本想着随便玩玩也就算了，现在看来不让你知道知道厉害不行了。”赵宜之抬起头打了个响指：“不过