我的目标是向网页发送 HTTP 请求
您展示的代码并没有尝试实现HTTP protocol 的任何相似之处，甚至没有接近。
一方面，如果您更仔细地查看自己的示例，您会发现 GET 请求（顺便说一句，由于您使用 HTTP 1.1，缺少必需的 Host 标头）包含 2 行休息，但cin.getline()（为什么不std::getline()？）一次只读取1行。所以，你读了一行，发送它，然后等待一个没有到达的响应，因为你还没有完成发送一个完整的请求。这可以解释为什么您的 while 循环挂起。
如果您希望用户输入完整的 HTTP 请求，然后按原样发送，则必须读取用户的整个请求，然后将其完全发送到服务器，然后才能尝试接收服务器的响应。这意味着您必须处理各个消息头之间的换行符，处理将消息头与消息正文分开的终止换行符，并检测正文数据的结尾。
我建议不要依赖用户按原样输入完整的 HTTP 请求。我建议您提示用户输入相关内容并让用户键入普通文本，然后您的代码可以根据需要将该文本格式化为适当的 HTTP 请求。
当您读取服务器的响应时，您不能只是盲目地读取任意数据块。您必须根据 HTTP 协议的规则处理您阅读的内容。这对于确定您何时到达响应结束并需要停止阅读尤其重要。如RFC 2616 Section 4.4 Message Length 中所述，可以通过多种不同方式之一发出响应结束的信号。
一般来说，您在处理 TCP 时也会犯一些常见的新手错误。 TCP 是一种流传输，您没有考虑到 send() 和 recv() 可以发送/接收比请求更少的字节。或者 recv() 不返回以 null 结尾的数据。
话虽如此，尝试这样的事情：
void sendAll(SOCKET sckt, const void *buf, int buflen)
{
// send all bytes until buflen has been sent,
// or an error occurs...
const char *pbuf = static_cast<const char*>(buf);
while (buflen > 0)
{
int numSent = send(sckt, pbuf, buflen, 0);
if (numSent < 0) {
std::ostringstream errMsg;
errMsg << "Error sending to socket: " << WSAGetLastError();
throw std::runtime_error(errMsg.str());
}
pbuf += numSent;
buflen -= numSent;
}
}
int readSome(SOCKET sckt, void *buf, int buflen)
{
// read as many bytes as possible until buflen has been received,
// the socket is disconnected, or an error occurs...
char *pbuf = static_cast<char*>(buf);
int total = 0;
while (buflen > 0)
{
int numRecvd = recv(sckt, pbuf, buflen, 0);
if (numRecvd < 0) {
std::ostringstream errMsg;
errMsg << "Error receiving from socket: " << WSAGetLastError();
throw std::runtime_error(errMsg.str());
}
if (numRecvd == 0) break;
pbuf += numRecvd;
buflen -= numRecvd;
total += numRecvd;
}
return total;
}
void readAll(SOCKET sckt, void *buf, int buflen)
{
// read all bytes until buflen has been received,
// or an error occurs...
if (readSome(sckt, buf, buflen) != buflen)
throw std::runtime_error("Socket disconnected unexpectedly");
}
std::string readLine(SOCKET sckt)
{
// read a line of characters until a line break is received...
std::string line;
char c;
do
{
readAll(sckt, &c, 1);
if (c == '\r')
{
readAll(sckt, &c, 1);
if (c == '\n') break;
line.push_back('\r');
}
else if (c == '\n') {
break;
}
line.push_back(c);
}
while (true);
return line;
}
...
inline void ltrim(std::string &s) {
// erase whitespace on the left side...
s.erase(s.begin(), std::find_if(s.begin(), s.end(), [](int ch) {
return !std::isspace(ch);
}));
}
inline void rtrim(std::string &s) {
// erase whitespace on the right side...
s.erase(std::find_if(s.rbegin(), s.rend(), [](int ch) {
return !std::isspace(ch);
}).base(), s.end());
}
inline void trim(std::string &s) {
// erase whitespace on both sides...
ltrim(s);
rtrim(s);
}
inline void upperCase(std::string &s)
{
// translate all characters to upper-case...
std::transform(s.begin(), s.end(), s.begin(), ::toupper);
}
...
std::string makeRequest(const std::string &host, const std::string &method, const std::string &resource, const std::vector<std::string> &extraHeaders, const void *body, int bodyLength)
{
std::ostringstream oss;
oss << method << " " << resource << " HTTP/1.1\r\n";
oss << "Host: " << host << "\r\n";
oss << "Content-Length: " << bodyLength << "\r\n";
for(auto &hdr : extraHeaders)
{
// TODO: ignore Host and Content-Length...
oss << hdr << "\r\n";
}
oss << "\r\n";
oss.write(static_cast<const char*>(body), bodyLength);
return oss.str();
}
bool getHeaderValue(const std::vector<std::string> &headers, const std::string &headerName, std::string &value)
{
value.clear();
std::string toFind = headerName;
upperCase(toFind);
// find the requested header by name...
for(auto &s : headers)
{
std::string::size_type pos = s.find(':');
if (pos != std::string::npos)
{
std::string name = s.substr(0, pos-1);
trim(name);
upperCase(name);
if (name == toFind)
{
// now return its value...
value = s.substr(pos+1);
trim(value);
return true;
}
}
}
// name not found
return false;
}
...
std::cout << "Connected to " << hostName << std::endl;
try
{
std::string method, resource, hdr, data;
std::string status, version, reason;
std::vector<std::string> headers;
int statusCode, rec;
do
{
headers.clear();
data.clear();
// get user input
std::cout << "Method > " << std::flush;
if (!std::getline(std::cin, method))
throw std::runtime_error("Error reading from stdin");
upperCase(method);
std::cout << "Resource > " << std::flush;
if (!std::getline(std::cin, resource))
throw std::runtime_error("Error reading from stdin");
std::cout << "Extra Headers > " << std::flush;
while (std::getline(std::cin, hdr) && !hdr.empty())
headers.push_back(hdr);
if (!std::cin)
throw std::runtime_error("Error reading from stdin");
std::cout << "Data > " << std::flush;
// use Ctrl-Z or Ctrl-D to end the data, depending on platform...
std::ios_base::fmtflags flags = std::cin.flags();
std::cin >> std::noskipws;
std::copy(std::istream_iterator<char>(std::cin), std::istream_iterator<char>(), std::back_inserter(data));
if (!std::cin)
throw std::runtime_error("Error reading from stdin");
std::cin.flags(flags);
std::cin.clear();
// send request
std::string request = makeRequest(hostName, method, resource, headers, data.c_str(), data.length());
std::cout << "Sending request: << std::endl << request << std::endl;
// TODO: reconnect to hostName if previous request disconnected...
sendAll(connectSocket, request.c_str(), request.length());
// receive response
headers.clear();
data.clear();
// read the status line and parse it...
status = readLine(connectSocket);
std::cout << status << std::endl;
std::getline(std::istringstream(status) >> version >> statusCode, reason);
upperCase(version);
// read the headers...
do
{
hdr = readLine(connectSocket);
std::cout << hdr << std::endl;
if (hdr.empty()) break;
headers.push_back(hdr);
}
while (true);
// The transfer-length of a message is the length of the message-body as
// it appears in the message; that is, after any transfer-codings have
// been applied. When a message-body is included with a message, the
// transfer-length of that body is determined by one of the following
// (in order of precedence):
// 1. Any response message which "MUST NOT" include a message-body (such
// as the 1xx, 204, and 304 responses and any response to a HEAD
// request) is always terminated by the first empty line after the
// header fields, regardless of the entity-header fields present in
// the message.
if (((statusCode / 100) != 1) &&
(statusCode != 204) &&
(statusCode != 304) &&
(method != "HEAD"))
{
// 2. If a Transfer-Encoding header field (section 14.41) is present and
// has any value other than "identity", then the transfer-length is
// defined by use of the "chunked" transfer-coding (section 3.6),
// unless the message is terminated by closing the connection.
if (getHeaderValue(headers, "Transfer-Encoding", hdr))
upperCase(hdr);
if (!hdr.empty() && (hdr != "IDENTITY"))
{
std::string chunk;
std::string::size_type oldSize, size;
do
{
chunk = readLine(connectSocket);
std::istringstream(chunk) >> std::hex >> size;
if (size == 0) break;
oldSize = data.size();
chunkData.resize(oldSize + size);
readAll(connectSocket, &data[oldSize], size);
std::cout.write(&data[oldSize], size);
readLine(connectSocket);
}
while (true);
std::cout << std::endl;
do
{
hdr = readLine(connectSocket);
std::cout << hdr << std::endl;
if (hdr.empty()) break;
headers.push_back(hdr);
}
while (true);
}
// 3. If a Content-Length header field (section 14.13) is present, its
// decimal value in OCTETs represents both the entity-length and the
// transfer-length. The Content-Length header field MUST NOT be sent
// if these two lengths are different (i.e., if a Transfer-Encoding
// header field is present). If a message is received with both a
// Transfer-Encoding header field and a Content-Length header field,
// the latter MUST be ignored.
else if (getHeaderValue(headers, "Content-Length", hdr))
{
std::string::size_type size;
if ((std::istringstream(hdr) >> size) && (size > 0))
{
data.resize(size);
readAll(connectSock, &data[0], size);
std::cout << data;
}
}
// 4. If the message uses the media type "multipart/byteranges", and the
// transfer-length is not otherwise specified, then this self-
// delimiting media type defines the transfer-length. This media type
// MUST NOT be used unless the sender knows that the recipient can parse
// it; the presence in a request of a Range header with multiple byte-
// range specifiers from a 1.1 client implies that the client can parse
// multipart/byteranges responses.
else if (getHeaderValue(headers, "Content-Type", hdr) &&
(hdr.compare(0, 10, "multipart/") == 0))
{
// TODO: extract 'boundary' attribute and read from
// socket until the terminating boundary is reached...
}
// 5. By the server closing the connection.
else
{
do
{
rec = readSome(connectSocket, recvBuf, sizeof(recvBuf));
if (rec == 0) break;
data.append(recvBuf, rec);
std::cout.write(recvBuf, rec);
}
while (rec == sizeof(recvBuf));
}
}
std::cout << std::endl;
// use status, headers, and data as needed ...
getHeaderValue(headers, "Connection", hdr);
upperCase(hdr);
if (version == "HTTP/1.0")
{
if (hdr != "KEEP-ALIVE")
break;
}
else
{
if (hdr == "CLOSE")
break;
}
}
while (true);
}
catch (const std::exception &e)
{
std::cerr << e.what() << std::endl;
}
closesocket(connectSocket);
std::cout << "Disconnected from " << hostName << std::endl;
std::system("pause");
HTTP 不是很有趣吗？ :-) 到目前为止，这还不是一个完整的 HTTP 实现，但它应该可以帮助您入门。但是，如您所见，从头开始实施 HTTP 可能非常复杂，并且您必须遵守许多规则和限制。你最好不要手动实现 HTTP。有大量可用于 C++ 的 3rd 方 HTTP 库。改用其中一个，让他们为您处理繁重的工作，这样您就可以专注于自己的业务逻辑。