I want to create a web application where a group of users could receive some data asynchronously whenever the C++ backend have something new. So, ideally, when a new user comes, he'll be added to the subscribers list and whenever the C++ backend has new data for the group, it will publish it to everybody.
libwebsockets seems to be a good library to use with C++. The only problem is that it seems that it's mainly designed on a callback system, so, apparently data is meant to be sent only if the client asks for it.
I found this post with a similar issue but I don't know if this is the best way to do that: How do I send async data via libwebsocket?
Any help will be appreciated.
Thank you.
found it!
libwebsockets lets you broadcast to all connected users to a specific protocol using libwebsocket_callback_on_writable_all_protocol(*protocol) which triggers LWS_CALLBACK_SERVER_WRITEABLE that will be handled by the protocol's callback function and that's where we could send the data.
So, typically, I use my second protocol (the non-http one) whenever I have some data to broadcast in any part of my code with libwebsocket_callback_on_writable_all_protocol(protocols + 1) and in the protocol's callback function
static int callback_linux_shell(struct libwebsocket_context * context, struct libwebsocket *wsi, enum libwebsocket_callback_reasons reason, void *user, void *in, size_t len) {
switch (reason) {
...
case LWS_CALLBACK_SERVER_WRITEABLE:
libwebsocket_write(wsi, my_data, my_data_size, LWS_WRITE_TEXT);
break;
...
}
}
Related
I am currently implementing a program that realizes TCP communication between PC program and external devices in QT. The problem I have is more general, but I will use this one as an example.
My class hierarchy looks like this:
Program
/ \
Server_A <--> Server_B <--- External system
|
Sockets[]
/ \
Commands Confirmations
/ | \
Interf1 Interf2 Interf3
I can get a command from device (Socket), my command gets into Confirmation class, realizes any Interface job, and returns confirmation back to Socket, which sends it back to device.
The problem occurs when I want to have a command send from an external system, which I also have to confirm.
I get a message on Server_B and pass it to Server_A with information about: socket to send command to and command to realize.
I pass a command to a particular socket
Socket sends a command to Commands, as there is logic for an External System commands.
Commands prepares a message, runs logic, and sends(through socket) message to device
Socket waits for response
Socket gets the response, understands that it was a response to an external system command, and passes it back to Commands
Commands realizes its logic.
Here it would all be fine, but the next step is:
Commands need to confirm the success(or failure) to external system.
So basically, what I have to do is pass a message from Commands to Server_B this way:
Commands->Socket->Server_A->Server_B. For all these classes, I would have to create an unnecessary method just to pass this one information. Is there a way to somehow solve this problem? During my programming, it often occurs that I have to pass something to the higher layer of my class structure, and it looks redundant to realize it through additional methods that only passes information further.
I have provided a sample pseudocode for this problem:
class Program
{
ServerB serverB;
ServerA serverA;
}
class ServerB
{
void send(QString msg);
}
class ServerA
{
QVector<MySocket*> sockets;
}
class MySocket
{
Commands commands;
Confirmations confirmations;
}
class Commands
{
void doLogic();
void sendToExternalSystem(QString message); //How to realize it?
}
My program is much bigger, but I hope it will give you a clue what I am trying to achieve. The simplest solution would be to add a method void sendToExternalSystem(QString message) into Sockets, Server_A and Server_B, aswell as providing a pointer for each parent during construction (commands will have access to sockets, sockets will have access to server_a, and server_a will have access to server_b)
Finally, I came up with a solution. It was necessary to implement ExternalCommand class, which instances were created in Server_B.
In the minimal solution, it has: 1. Field QString Message, 2. Method QString getMessage(), 3. Method void finish(QString), 4. Signal void sendToExternal(QString)
When I read the message sent from the external system in Server_B, I create an instance of this class, and connect it to the Server_B send method. In my code, it looks like that:
ExternalCommand::ExternalCommand(QString message, QObject* parent) : QObject(parent)
{
this->message=message;
}
QString ExternalCommand::getMessage()
{
return this->message;
}
void finish(QString outputMessage)
{
emit sendToExternal(outputMessage);
}
void Server_B::onReadyRead()
{
QTcpSocket *socket = dynamic_cast<QTcpSocket*>(sender());
QString message = socket->readAll();
ExternalCommand* cmd = new ExternalCommand(message);
connect(cmd, &ExternalCommand::sendToExternal, socket,
[socket](QString message) {socket->write(message.toUtf8());});
}
It was also necessary to implement some type of object destruction in ExternalCommand once the command is sent, but it isn't the point of this question.
So once this is implemented, instead of the message as QString, the message is passed to the lower levels as ExternalCommand*, and once an answer is got, it is possible to send it back to the External System, by calling ExternalCommand::finish(QString outputMessage);. Of course, this is just a minimal solution for this problem.
Thanks to #MatG for pointing me to Promise/Future pattern, which was helpful in finding this solution.
I'm trying to use libmosquitto to make a request (publish to a 'test/topic' topic) and I want to get a response based on the client (sender) id. So that means the client will publish to 'test/topic' and it will automatically subscribe 'test/topic/<client_id>'
The server has already subscribed on 'test/topic' and when it becomes the message, it will send a response (publish) to 'test/topic/<client_id>', which the client subscribed to receive that response in the first place.
The challenge here is how do I get the <client_id>, right. I already done this in python and js, where the client will send metadata or properties in the payload, which the server can unpack to get the client_id. However, I'm using C++ now and it's frustrating because I can't figure out how to get these properties.
Here is an example of how to do this in python. I just want to do the same with c++
I'm using the libmosquitto as I mentionned. I don't even have an example to show because I didn't find how to do this. There is literally no example on how to do this with the mosquitto c++ lib (which is confusing since mosquitto is a famous lib I guess).
I hope someone had a similar problem or can post an example for c++ and mosquitto lib. Thanks in advance.
When in doubt, look at the tests:
const char *my_client_id = ...;
mosquitto_property *proplist = NULL;
mosquitto_property_add_string_pair(&proplist, MQTT_PROP_USER_PROPERTY, "client_id", my_client_id);
mosquitto_publish_v5(mosq, &sent_mid, "test/topic", strlen("message"), "message", 0, false, proplist);
mosquitto_property_free_all(&proplist);
Since you asked in the comments, you can retrieve these properties from published messages by first setting an on_message callback using mosquitto_message_v5_callback_set and the implementing it like so:
void on_message(struct mosquitto *mosq, void *obj, const struct mosquitto_message *message, const mosquitto_property *props) {
std::string topic{message->topic};
if (topic == "test/topic") {
const char *client_id = nullptr;
mosquitto_property_read_string_pair(props, MQTT_PROP_USER_PROPERTY, nullptr, &client_id, false);
if (client_id) {
/* client_id contains a client id. */
}
}
In the unary RPC example provided in the grpc Github (client) and (server), is there any way to detect client's closed connection?
For example, in server.cc file:
std::string prefix("Hello ");
reply_.set_message(prefix + request_.name());
// And we are done! Let the gRPC runtime know we've finished, using the
// memory address of this instance as the uniquely identifying tag for
// the event.
status_ = FINISH;
int p = 0,i=0;
while(i++ < 1000000000) { // some dummy work
p = p + 10;
}
responder_.Finish(reply_, Status::OK, this);
With this dummy task before sending the response back to the client, server will take a few seconds. If we close the client (for example say with Ctrl+C), the server does not throw any error. It simply calls Finish and then deallocates the object as if the Finish is successful.
Is there any async feature (handler function) on the server-side to get us notified that the client has closed the connection or client is terminated?
Thank You!
Unfortunately, no.
But now guys from gRPC team works hard to implement callback mechanism into C++ implementation. As I understand it will work the same way as on Java implementation( https://youtu.be/5tmPvSe7xXQ?t=1843 ).
You can see how to work with future API with next examples: client_callback.cc and server_callback.cc
And the point of your interest there is ServerBidiReactor class from ::grpc::experimental namespace for server side. It have OnDone and OnCancel notification methods that maybe can help you.
Another interesting point there is that you can store a pointers to connection object and send notifications to client at any time.
But it still have many issue and I don't recommend to use this API in production code.
Current progress of C++ callbacks implementation you can see there: https://github.com/grpc/grpc/projects/12#card-12554506
I hope I don't have misunderstood the Thrift concept, but what I see from (example) questions like this, this framework is composed by different modular layers that can be enabled or disabled.
I'm mostly interesed in the "IDL part" of Thrift, so that I can create a common interface between my C++ code and an external Javascript application. I would like to call C++ functions using JS, with Binary data transmission, and I've already used the compiler for this.
But both my C++ (the server) and JS (client) application already exchange data using a C++ Webserver with Websockets support, it is not provided by Thrift.
So I was thinking to setup the following items:
In JS (already done):
TWebSocketTransport to send data to my "Websocket server" (with host ws://xxx.xxx.xxx.xxx)
TBinaryProtocol to encapsulate the data (using this JS implementation)
The compiled Thrift JS library with the correspondent C++ functions to call (done with the JS compiler)
In C++ (partial):
TBinaryProtocol to encode/decode the data
A TProcessor with handler to get the data from the client and process it
For now, the client is already able to sent requests to my websocket server, I see receiving them in binary form and I just need Thrift to:
Decode the input
Call the appropriate C++ function
Encode the output
My webserver will send the response to the client. So no "Thrift server" is needed here. I see there is the TProcessor->process() function, I'm trying to use it when I receive the binary data but it needs an in/out TProtocol. No problem here... but in order to create the TBinaryProtocol I also need a TTransport! If no Thrift server is expected... what Transport should I use?
I tried to set TTransport to NULL in TBinaryProtocol constructor, but once I use it it gives nullptr exception.
Code is something like:
Init:
boost::shared_ptr<MySDKServiceHandler> handler(new MySDKServiceHandler());
thriftCommandProcessor = boost::shared_ptr<TProcessor>(new MySDKServiceProcessor(handler));
thriftInputProtocol = boost::shared_ptr<TBinaryProtocol>(new TBinaryProtocol(TTransport???));
thriftOutputProtocol = boost::shared_ptr<TBinaryProtocol>(new TBinaryProtocol(TTransport???));
When data arrives:
this->thriftInputProtocol->writeBinary(input); // exception here
this->thriftCommandProcessor->process(this->thriftInputProtocol, this->thriftOutputProtocol, NULL);
this->thriftOutputProtocol->readBinary(output);
I've managed to do it using the following components:
// create the Processor using my compiled Thrift class (from IDL)
boost::shared_ptr<MySDKServiceHandler> handler(new MySDKServiceHandler());
thriftCommandProcessor = boost::shared_ptr<TProcessor>(new ThriftSDKServiceProcessor(handler));
// Transport is needed, I use the TMemoryBuffer so everything is kept in local memory
boost::shared_ptr<TTransport> transport(new apache::thrift::transport::TMemoryBuffer());
// my client/server data is based on binary protocol. I pass the transport to it
thriftProtocol = boost::shared_ptr<TProtocol>(new TBinaryProtocol(transport, 0, 0, false, false));
/* .... when the message arrives through my webserver */
void parseMessage(const byte* input, const int input_size, byte*& output, int& output_size)
{
// get the transports to write and read Thrift data
boost::shared_ptr<TTransport> iTr = this->thriftProtocol->getInputTransport();
boost::shared_ptr<TTransport> oTr = this->thriftProtocol->getOutputTransport();
// "transmit" my data to Thrift
iTr->write(input, input_size);
iTr->flush();
// make the Thrift work using the Processor
this->thriftCommandProcessor->process(this->thriftProtocol, NULL);
// the output transport (oTr) contains the called procedure result
output = new byte[MAX_SDK_WS_REPLYSIZE];
output_size = oTr->read(output, MAX_SDK_WS_REPLYSIZE);
}
My webserver will send the response to the client. So no "Thrift server" is needed here. I see there is the TProcessor->process() function, I'm trying to use it when I receive the binary data but it needs an in/out TProtocol. No problem here... but in order to create the TBinaryProtocol I also need a TTransport! If no Thrift server is expected... what Transport should I use?
The usual pattern is to store the bits somewhere and use that buffer or data stream as the input, same for the output. For certain languages there is a TStreamTransport available, for C++ the TBufferBase class looks promising to me.
i'm studying this source base. Basically this is an Anim server client for Symbian 3rd edition for the purpose of grabbing input events without consuming them in a reliable way.
If you spot this line of the server, here it is basically setting the RProperty value (apparently to an increasing counter); it seems no actual processing of the input is done.
inside this client line, the client is supposed to be receiving the notification data, but it only calls Attach.
my understanding is that Attach is only required to be called once, but is not clear in the client what event is triggered every time the server sets the RProperty
How (and where) is the client supposed to access the RProperty value?
After Attaching the client will somewhere Subscribe to the property where it passes a TRequestStatus reference. The server will signal the request status property via the kernel when the asynchronous event has happened (in your case the property was changed). If your example source code is implemented in the right way, you will find an active object (AO; CActive derived class) hanging around and the iStatus of this AO will be passed to the RProperty API. In this case the RunL function of the AO will be called when the property has been changed.
It is essential in Symbian to understand the active object framework and quite few people do it actually. Unfortunately I did not find a really good description online (they are explained quite well in Symbian OS Internals book) but this page at least gives you a quick example.
Example
In the ConstructL of your CMyActive subclass of CActive:
CKeyEventsClient* iClient;
RProperty iProperty;
// ...
void CMyActive::ConstructL()
{
RProcess myProcess;
TSecureId propertyCategory = myProcess.SecureId();
// avoid interference with other properties by defining the category
// as a secure ID of your process (perhaps it's the only allowed value)
TUint propertyKey = 1; // whatever you want
iClient = CKeyEventsClient::NewL(propertyCategory, propertyKey, ...);
iClient->OpenNotificationPropertyL(&iProperty);
// ...
CActiveScheduler::Add(this);
iProperty.Subscribe(iStatus);
SetActive();
}
Your RunL will be called when the property has been changed:
void CMyActive::RunL()
{
if (iStatus.Int() != KErrCancel) User::LeaveIfError(iStatus.Int());
// forward the error to RunError
// "To ensure that the subscriber does not miss updates, it should
// re-issue a subscription request before retrieving the current value
// and acting on it." (from docs)
iProperty.Subscribe(iStatus);
TInt value; // this type is passed to RProperty::Define() in the client
TInt err = iProperty.Get(value);
if (err != KErrNotFound) User::LeaveIfError(err);
SetActive();
}