wsgi.py 文件源码

def write(self, data):
        """
        The WSGI I{write} callable returned by the I{start_response} callable.
        The given bytes will be written to the response body, possibly flushing
        the status and headers first.

        This will be called in a non-I/O thread.
        """
        # PEP-3333 states:
        #
        #   The server or gateway must transmit the yielded bytestrings to the
        #   client in an unbuffered fashion, completing the transmission of
        #   each bytestring before requesting another one.
        #
        # This write() method is used for the imperative and (indirectly) for
        # the more familiar iterable-of-bytestrings WSGI mechanism. It uses
        # C{blockingCallFromThread} to schedule writes. This allows exceptions
        # to propagate up from the underlying HTTP implementation. However,
        # that underlying implementation does not, as yet, provide any way to
        # know if the written data has been transmitted, so this method
        # violates the above part of PEP-3333.
        #
        # PEP-3333 also says that a server may:
        #
        #   Use a different thread to ensure that the block continues to be
        #   transmitted while the application produces the next block.
        #
        # Which suggests that this is actually compliant with PEP-3333,
        # because writes are done in the reactor thread.
        #
        # However, providing some back-pressure may nevertheless be a Good
        # Thing at some point in the future.

        def wsgiWrite(started):
            if not started:
                self._sendResponseHeaders()
            self.request.write(data)

        try:
            return blockingCallFromThread(
                self.reactor, wsgiWrite, self.started)
        finally:
            self.started = True