深入tornado中的http1connection

前言

　　tornado中http1connection文件的作用极其重要，他实现了http1.x协议。

　　本模块基于gen模块和iostream模块实现异步的处理请求或者响应。

　　阅读本文需要一些基础的http知识。

正文：

　　http协议是建立在tcp基础上的应用层协议，tcp层由TCPServer，IOStream负责，对http报文的读取与解析则由http1connection负责。当http报文被解析完成再交由给某个delegate类实例负责进行后续处理。

接下来说一下大体的过程(tornado作为服务器端)：

　　如果客户端要向服务端发送http请求，首先要建立tcp连接，

　　tornado在连接建立完成后会将连接封装为一个IOStream对象，这个对象可以异步的从连接中读写数据

　　tornado中又实现了HTTP1ServerConnection与HTTP1Connection两个类，他们依赖于底层的IOStream从套接字中读写，并共同合作完成了http1.x协议。

　　HTTP1Connection实际主要是用来处理http事务(http权威指南：http事务是由一条请求以及对应该请求的响应组成)，当然他自己实现了前半部分，也就是对报文起始行、首部、主体进行读，解析；后半部分需要配合HTTPMessageDelegate进行工作。

　　HTTPMessageDelegate对经过HTTP1Connection解析后的报文进行分配，然后由其他类(比如说RequestHandler)执行具体的业务逻辑。

　　其他类生成响应，并且把响应发送到IOStream中，这时就表示着这条http事务已经完成，我们需要根据情况判断是否关闭连接。

　　HTTP1ServerConnection则是不断的生成HTTP1Connection实例，也就是不断的处理http事务，直到连接关闭。

HTTP1Connection类：

　　先看HTTP1Connection这个类。这个类实际上主要完成的工作可以概括为两个：

　　　　1  读取并解析报文消息

　　　　2  写入报文

1 读取并解析报文

　　当请求或者响应到来的时候，read_response是解析消息的入口，尽管该方法读起来好像仅仅是针对响应，但因为不管是请求或者是响应格式是相差不大的，所以不管是请求或者是响应他都是可以处理的。read_response中主要调用了_read_message方法，解析报文的逻辑也都是在_read_message方法中，另外，本文主要是对其作为服务端时进行分析

先来说一下_read_mssage的大体逻辑：

　　首先HTTP1Connection基于iostream读取请求报文，并对请求报文进行解析，分离出起始行 请求首部，并根据请求首部判定是否读取消息主体以及消息主体的长度。

　　在这个过程中，分析起始行的信息然后委托给代理(HTTPMessageDelegate)获取对应的RequestHandler(这一步主要是_RequestDispatcher类实现的)，并实例化，根据起始行的method调用相关方法，在调用方法执行业务逻辑时可能会用到模板语言，cookie，csrf等等其他东西，但最终会产生响应，并将响应发送到IOStream中。这些工作都是delegate干的。

　　最后HTTP1Connection等待响应发送完成(这一步操作是异步的)，根据是否支持keep-alive决定是否处理完后关闭连接

    def read_response(self, delegate):
        """Read a single HTTP response.
        """
        if self.params.decompress:
            delegate = _GzipMessageDelegate(delegate, self.params.chunk_size)
        return self._read_message(delegate)

    @gen.coroutine
    def _read_message(self, delegate):    # self是HTTP1Connection实例对象，delegate是_ServerRequestAdapter实例对象
        need_delegate_close = False
        try:
            header_future = self.stream.read_until_regex(b"
?

?
", max_bytes=self.params.max_header_size)
            # 两种方式来等待请求头(在服务器模式下是请求，客户端模式下是响应)的读，第一种是什么时候发送过来什么时候读，
            # 第二种是设置超时时间时长的定时任务，如果这段时间内没有发送过来那么就关闭连接
            if self.params.header_timeout is None: # 第一种
                header_data = yield header_future  # 获取起始行以及头部信息的bytes流
            else:    # 第二种
                try:
                    header_data = yield gen.with_timeout(
                        self.stream.io_loop.time() + self.params.header_timeout,
                        header_future,
                        io_loop=self.stream.io_loop,
                        quiet_exceptions=iostream.StreamClosedError)
                except gen.TimeoutError:
                    self.close()
                    raise gen.Return(False)
            start_line, headers = self._parse_headers(header_data)    # 获取起始行以及头部信息
            if self.is_client:
                start_line = httputil.parse_response_start_line(start_line)
                self._response_start_line = start_line
            else:
                start_line = httputil.parse_request_start_line(start_line)
                self._request_start_line = start_line
                self._request_headers = headers

            self._disconnect_on_finish = not self._can_keep_alive(start_line, headers) # 如果不是keep alive在响应结束后关闭连接
            need_delegate_close = True
            with _ExceptionLoggingContext(app_log):
                # 这一步会做了很多东西，如果是服务器端，这一步会设置request对象，并根据请求中的url选择对应handler
                header_future = delegate.headers_received(start_line, headers)
                if header_future is not None:
                    yield header_future
            if self.stream is None:
                # We've been detached.
                need_delegate_close = False
                raise gen.Return(False)
            skip_body = False
            if self.is_client:  # 作为client
                if (self._request_start_line is not None and
                        self._request_start_line.method == 'HEAD'):  # 如果方法是HEAD，那么默认是没有主体的。即使有主体也会被忽略掉
                    skip_body = True
                code = start_line.code
                # 如果客户端发送了一个带条件的GET 请求且该请求已被允许，而文档的内容（自上次访问以来或者根据请求的条件）并没有改变，则服务器应当返回这个304状态码
                if code == 304:  
                    # 304报文可能会包含content-length首部属性，但实际上是没有消息主体的
                    # http://tools.ietf.org/html/rfc7230#section-3.3
                    skip_body = True
                if code >= 100 and code < 200: # 临时的响应。客户端在收到常规响应之前，应准备接收一个或多个1XX响应
                    # 1xx 报文是不能包含主体信息的
                    if ('Content-Length' in headers or
                            'Transfer-Encoding' in headers):
                        raise httputil.HTTPInputError(
                            "Response code %d cannot have body" % code)
                    # 我们所需要的真正的响应还没有接收到，所以继续接收
                    yield self._read_message(delegate)
            else:
                # 1、Expect:100-continue 用于客户端在发送POST数据给服务器前，征询服务器情况，看服务器是否处理POST的数据，
                # 如果不处理，客户端则不上传POST数据，如果处理，则POST上传数据。在现实应用中，通过在POST大数据时，才会使用100-continue协议。
                # http://www.cnblogs.com/tekkaman/archive/2013/04/03/2997781.html
                if (headers.get("Expect") == "100-continue" and not self._write_finished):
                    # 默认是支持的，所以收到请求后，返回100。
                    self.stream.write(b"HTTP/1.1 100 (Continue)

")
            if not skip_body:
                # 这一步读取消息主体
                body_future = self._read_body(start_line.code if self.is_client else 0, headers, delegate)
                if body_future is not None:
                    if self._body_timeout is None:
                        yield body_future
                    else:
                        try:
                            yield gen.with_timeout(
                                self.stream.io_loop.time() + self._body_timeout,
                                body_future, self.stream.io_loop,
                                quiet_exceptions=iostream.StreamClosedError)
                        except gen.TimeoutError:
                            gen_log.info("Timeout reading body from %s", self.context)
                            self.stream.close()
                            raise gen.Return(False)
            self._read_finished = True
            if not self._write_finished or self.is_client:
                need_delegate_close = False
                with _ExceptionLoggingContext(app_log):
                    # 如果是服务器端，这一步会生成对应的handler实例，然后执行业务逻辑，最后将响应写入IOStream中
                    delegate.finish()
            # If we're waiting for the application to produce an asynchronous
            # response, and we're not detached, register a close callback
            # on the stream (we didn't need one while we were reading)
            # 等待异步响应完成，所有数据都写入 fd，才继续后续处理，详细见 _finish_request/finish 方法实现。
            # 当异步写完成，在HTTPServerRequest中调用当前对象的finish方法，finish方法则会调用_finisth_request方法，该方法内部会对_finish_future对象set_result
            if (not self._finish_future.done() and self.stream is not None and not self.stream.closed()):
                self.stream.set_close_callback(self._on_connection_close)
                yield self._finish_future
            # 判定是否关闭连接，服务器端一般等待客户端主动关闭，而如果是客户端则根据是否持久连接进行关闭
            if self.is_client and self._disconnect_on_finish:    
                self.close()
            if self.stream is None:
                raise gen.Return(False)
        except httputil.HTTPInputError as e:
            gen_log.info("Malformed HTTP message from %s: %s", self.context, e)
            self.close()
            raise gen.Return(False)
        finally:
            if need_delegate_close:
                with _ExceptionLoggingContext(app_log):
                    delegate.on_connection_close()
            self._clear_callbacks()
        raise gen.Return(True)

值得注意的是，该方法中读取报文主体的几种方式：
　　1 假设没有开启keep-alive，那么我们将连接结束作为报文终止的标志

　　2 而如果开启了keep-alive，那么我们根据Content-Length确定当前报文的终止位置

　　3 如果开启了分块传输编码(Transfer-Encoding:chunked，这时候Content-Length就不起作用了，实际上在tornado中如果Content-Length以及分块传输编码都指定则会返回错误)那么就会根据分块传输编码的格式一直读取，直到读取到b"0 "时就可以确定当前报文已终止

_read_body方法则根据情况选择读取报文主体的方式，以上三种选择分别对应于以下三种方法：

　　1 _read_body_until_close

　　2 _read_fixed_body

　　3 _read_chunked_body

 来看一下代码：

    def _read_body(self, code, headers, delegate):
        # https://imququ.com/post/transfer-encoding-header-in-http.html
        if "Content-Length" in headers:
            if "Transfer-Encoding" in headers:
                # Response cannot contain both Content-Length and
                # Transfer-Encoding headers.
                # If a message is received with both a Transfer-Encoding and a
                # Content-Length header field, the Transfer-Encoding overrides the
                # Content-Length.  
                # http://tools.ietf.org/html/rfc7230#section-3.3.3
                raise httputil.HTTPInputError("Response with both Transfer-Encoding and Content-Length")
            if "," in headers["Content-Length"]:
                # Proxies sometimes cause Content-Length headers to get
                # duplicated.  If all the values are identical then we can
                # use them but if they differ it's an error.
                pieces = re.split(r',s*', headers["Content-Length"])
                if any(i != pieces[0] for i in pieces):
                    raise httputil.HTTPInputError(
                        "Multiple unequal Content-Lengths: %r" %
                        headers["Content-Length"])
                headers["Content-Length"] = pieces[0]

            try:
                content_length = int(headers["Content-Length"])
            except ValueError:
                # Handles non-integer Content-Length value.
                raise httputil.HTTPInputError(
                    "Only integer Content-Length is allowed: %s" % headers["Content-Length"])

            if content_length > self._max_body_size:
                raise httputil.HTTPInputError("Content-Length too long")
        else:
            content_length = None

        if code == 204:        # 状态码204（无内容）
            # This response code is not allowed to have a non-empty body,
            # and has an implicit length of zero instead of read-until-close.
            # http://www.w3.org/Protocols/rfc2616/rfc2616-sec4.html#sec4.3
            if ("Transfer-Encoding" in headers or content_length not in (None, 0)):
                raise httputil.HTTPInputError(
                    "Response with code %d should not have body" % code)
            content_length = 0

        if content_length is not None: # 而如果开启了keep-alive，那么我们根据Content-Length确定当前报文的终止位置
            return self._read_fixed_body(content_length, delegate)
        if headers.get("Transfer-Encoding") == "chunked": # 开启了分块传输编码
            return self._read_chunked_body(delegate)
        if self.is_client: # 非持久连接
            return self._read_body_until_close(delegate)
        return None

    @gen.coroutine
    def _read_fixed_body(self, content_length, delegate):
        while content_length > 0:
            body = yield self.stream.read_bytes(min(self.params.chunk_size, content_length), partial=True)
            content_length -= len(body)
            if not self._write_finished or self.is_client:
                with _ExceptionLoggingContext(app_log):
                    ret = delegate.data_received(body)
                    if ret is not None:
                        yield ret

    @gen.coroutine
    def _read_chunked_body(self, delegate):
        # TODO: "chunk extensions" http://tools.ietf.org/html/rfc2616#section-3.6.1
        total_size = 0
        while True:
            # 先读取chunk长度
            chunk_len = yield self.stream.read_until(b"
", max_bytes=64)
            chunk_len = int(chunk_len.strip(), 16)
            # 如果chunk长度为0，表示分块传输的终止
            if chunk_len == 0:
                return
            total_size += chunk_len
            # 检测长度是否过大
            if total_size > self._max_body_size:
                raise httputil.HTTPInputError("chunked body too large")
            bytes_to_read = chunk_len
            while bytes_to_read:
                # 读取该长度对应的data
                chunk = yield self.stream.read_bytes(
                    min(bytes_to_read, self.params.chunk_size), partial=True)
                bytes_to_read -= len(chunk)
                if not self._write_finished or self.is_client:
                    with _ExceptionLoggingContext(app_log):
                        # 读取的消息主体要交给代理(HTTPMessageDelegate)处理
                        ret = delegate.data_received(chunk)
                        if ret is not None:
                            yield ret
            # chunk ends with 

            # 每一个data后面都有一个CRLF
            crlf = yield self.stream.read_bytes(2)
            assert crlf == b"
"

    @gen.coroutine
    def _read_body_until_close(self, delegate):
        body = yield self.stream.read_until_close()
        if not self._write_finished or self.is_client:
            with _ExceptionLoggingContext(app_log):
                delegate.data_received(body)

 2 写入报文

　　写入报文主要可以分两步：

　　　　1  写入报文起始行以及头部

　　　　2  写入报文主体

与之相关的方法主要有三个，来看源码：

    def write_headers(self, start_line, headers, chunk=None, callback=None):
        """Implements `.HTTPConnection.write_headers`.写入起始行和消息头"""
        lines = []
        if self.is_client:    # 客户端，那就是发送请求了
            self._request_start_line = start_line
            lines.append(utf8('%s %s HTTP/1.1' % (start_line[0], start_line[1])))
            # Client requests with a non-empty body must have either a
            # Content-Length or a Transfer-Encoding.
            self._chunking_output = (
                start_line.method in ('POST', 'PUT', 'PATCH') and
                'Content-Length' not in headers and
                'Transfer-Encoding' not in headers)
        else:    # 服务端，那就是发送响应了
            self._response_start_line = start_line
            lines.append(utf8('HTTP/1.1 %d %s' % (start_line[1], start_line[2])))
            self._chunking_output = (
                # TODO: should this use
                # self._request_start_line.version or
                # start_line.version?
                self._request_start_line.version == 'HTTP/1.1' and
                # 304 responses have no body (not even a zero-length body), and so
                # should not have either Content-Length or Transfer-Encoding.
                # headers.
                start_line.code not in (204, 304) and
                # No need to chunk the output if a Content-Length is specified.
                'Content-Length' not in headers and
                # Applications are discouraged from touching Transfer-Encoding,
                # but if they do, leave it alone.
                'Transfer-Encoding' not in headers)
            # If a 1.0 client asked for keep-alive, add the header.
            if (self._request_start_line.version == 'HTTP/1.0' and
                (self._request_headers.get('Connection', '').lower() ==
                 'keep-alive')):
                headers['Connection'] = 'Keep-Alive'
        # tornado无论作为客户端还是服务端默认是支持分块传输的
        if self._chunking_output:
            headers['Transfer-Encoding'] = 'chunked'
        # 这下这种情况消息主体应为空
        if (not self.is_client and
            (self._request_start_line.method == 'HEAD' or
             start_line.code == 304)):
            self._expected_content_remaining = 0
        elif 'Content-Length' in headers:
            self._expected_content_remaining = int(headers['Content-Length'])
        else:
            self._expected_content_remaining = None
        # TODO: headers are supposed to be of type str, but we still have some
        # cases that let bytes slip through. Remove these native_str calls when those
        # are fixed.
        header_lines = (native_str(n) + ": " + native_str(v) for n, v in headers.get_all())
        if PY3:
            lines.extend(l.encode('latin1') for l in header_lines)
        else:
            lines.extend(header_lines)
        for line in lines:
            if b'
' in line:
                raise ValueError('Newline in header: ' + repr(line))
        future = None
        if self.stream.closed():
            future = self._write_future = Future()
            future.set_exception(iostream.StreamClosedError())
            future.exception()
        else:
            if callback is not None:
                self._write_callback = stack_context.wrap(callback)
            else:
                future = self._write_future = Future()
            data = b"
".join(lines) + b"

"
            if chunk:
                data += self._format_chunk(chunk)
            self._pending_write = self.stream.write(data)
            self._pending_write.add_done_callback(self._on_write_complete)
        return future

    def _format_chunk(self, chunk):
        '''如果采用了分块传输编码，则将参数chunk转换为http中规定的格式。如果没有使用分块传输编码则原样返回'''
        if self._expected_content_remaining is not None:
            self._expected_content_remaining -= len(chunk)
            if self._expected_content_remaining < 0:
                # Close the stream now to stop further framing errors.
                self.stream.close()
                raise httputil.HTTPOutputError(
                    "Tried to write more data than Content-Length")
        if self._chunking_output and chunk:
            # Don't write out empty chunks because that means END-OF-STREAM
            # with chunked encoding
            return utf8("%x" % len(chunk)) + b"
" + chunk + b"
"
        else:
            return chunk

    def write(self, chunk, callback=None):
        """Implements `.HTTPConnection.write`. 写入报文主体

        For backwards compatibility is is allowed but deprecated to
        skip `write_headers` and instead call `write()` with a
        pre-encoded header block.
        """
        future = None
        if self.stream.closed():
            future = self._write_future = Future()
            self._write_future.set_exception(iostream.StreamClosedError())
            self._write_future.exception()
        else:
            if callback is not None:
                self._write_callback = stack_context.wrap(callback)
            else:
                future = self._write_future = Future()
            self._pending_write = self.stream.write(self._format_chunk(chunk))
            self._pending_write.add_done_callback(self._on_write_complete)
        return future

HTTP1ServerConnection类

HTTP1ServerConnection比较简单，主要实现了服务端处理逻辑：

　　在本条tcp连接上，不停的处理http事务(当然也有可能客户端不支持持久连接所以处理完一条http事务后，tcp连接被关闭)

　　当发生异常时，关闭连接。

其中start_serving方法是入口，其内部调用了_server_request_loop，来看代码

def start_serving(self, delegate):
        """
            Starts serving requests on this connection.

            :arg delegate: a `.HTTPServerConnectionDelegate`
        """
        assert isinstance(delegate, httputil.HTTPServerConnectionDelegate) 
        self._serving_future = self._server_request_loop(delegate)
        # Register the future on the IOLoop so its errors get logged.
        self.stream.io_loop.add_future(self._serving_future, lambda f: f.result())

    @gen.coroutine
    def _server_request_loop(self, delegate):
        try:
            while True:
                # 不断处理http事务，知道连接关闭或者出现异常
                conn = HTTP1Connection(self.stream, False, self.params, self.context)
                request_delegate = delegate.start_request(self, conn)    
                try:
                    ret = yield conn.read_response(request_delegate)
                except (iostream.StreamClosedError, iostream.UnsatisfiableReadError): # 连接关闭
                    return
                except _QuietException:
                    # This exception was already logged.
                    conn.close()
                    return
                except Exception:
                    gen_log.error("Uncaught exception", exc_info=True)
                    conn.close()
                    return
                if not ret:
                    return
                yield gen.moment
        finally:
            delegate.on_close(self)

参考：

　　http://www.cnblogs.com/tekkaman/archive/2013/04/03/2997781.html

　　http://strawhatfy.github.io/2015/11/02/tornado.http1connection.HTTP1Connection/