public async Task <IActionResult> ExecuteAction(RequestDeviceAction requestDeviceAction)
{
ServiceResponse <string> response = await _deviceService.executeDeviceAction(requestDeviceAction);
if (response.Success)
{
return(Ok(response));
}
return(BadRequest(response));
}
public async Task <ServiceResponse <string> > executeDeviceAction(RequestDeviceAction requestDeviceAction)
{
ServiceResponse <string> serviceResponse = new ServiceResponse <string>();
var userClaims = _userService.getAuthenticatedUserClaims();
//Not sure if this big try block is correct, declaring a variable inside a try block in PHP
//doesn't limit its scope to the try block, so try blocks are smaller.
try {
var action = await _userDeviceActionRepository.getUserDeviceAction(
userClaims.Email,
requestDeviceAction.actionId,
requestDeviceAction.deviceId
);
Rol role = await _userRoleRepository.getUserRole(userClaims.Email);
if (!role.isParent() && action.Name.Contains("REQUEST"))
{
//here we should call a service that will send a notification to
//the main user email of the family (or push notification)
//asking for permission
serviceResponse.Message = "Action request sent";
return(serviceResponse);
}
//here we call the service with the connection to the Transport Layer API,
//and ask for the action to run.
serviceResponse.Message = "Action performed";
} catch (KeyNotFoundException) {
//This will be returned even if the endpoint receives a device id incorrect,
//we could split the queries to check if the device exists and return the correct
//message. It will be less performant, but will give a better UX.
//Anyways, the list of devices is gotten from the API, and FE should render it,
//so we should never receive a request with an invalid deviceId.
serviceResponse.Message = "Action forbbiden for user";
serviceResponse.Success = false;
} catch (UserRoleNotFoundException ex) {
serviceResponse.Message = ex.Message;
serviceResponse.Success = false;
}
return(serviceResponse);
}
