C++ Own Observer Pattern - c++

I'm designing an observer pattern which should work this way: observer calls AddEventListener method of EventDispatcher and passes a string which is the name of the event, PointerToItself and a PointerToItsMemberMethod
After that event happens inside of the EventDispatcher; it looks through the list of subscriptions and if there are some, assigned to this event calls the action method of the observer.
I've come to this EventDispatcher.h. CAUTION contains bit of pseudo-code.
The are two questions:
How do I define the type of action in struct Subscription?
Am I moving the right way?
PS: No, I'm not gonna use boost or any other libraries .
#pragma once
#include <vector>
#include <string>
using namespace std;
struct Subscription
{
void* observer;
string event;
/* u_u */ action;
};
class EventDispatcher
{
private:
vector<Subscription> subscriptions;
protected:
void DispatchEvent ( string event );
public:
void AddEventListener ( Observer* observer , string event , /* u_u */ action );
void RemoveEventListener ( Observer* observer , string event , /* u_u */ action );
};
This header implements like this in EventDispatcher.cpp
#include "EventDispatcher.h"
void EventDispatcher::DispatchEvent ( string event )
{
int key = 0;
while ( key < this->subscriptions.size() )
{
Subscription subscription = this->subscriptions[key];
if ( subscription.event == event )
{
subscription.observer->subscription.action;
};
};
};
void EventDispatcher::AddEventListener ( Observer* observer , string event , /* */ action )
{
Subscription subscription = { observer , event , action );
this->subscriptions.push_back ( subscription );
};
void EventDispatcher::RemoveEventListener ( Observer* observer , string event , /* */ action )
{
int key = 0;
while ( key < this->subscriptions.size() )
{
Subscription subscription = this->subscriptions[key];
if ( subscription.observer == observer && subscription.event == event && subscription.action == action )
{
this->subscriptions.erase ( this->subscriptions.begin() + key );
};
};
};

You could either define an Action class or pass a lambda function (C++11). In the latter case, action could be defined as
function<void (EventDispatcher*)> action;
and you would register the observer as follows
Observer * me = this;
observable->AddEventListener (this, "EventName", [me] (EventDispatcher* dispatcher) {
// code here; me is available
});
You should probably use smart weak pointers to store the Observers in the EventDispatcher, such that you do not have to care for un-registering.
Edit: Added following example (just one subscription possible, but should illustrate the idea -- you have to be careful that you do not reference an object that does no longer exist)
struct Observable {
std::weak_ptr<function<void (const Observable&)>> action;
void AddEventListener (std::weak_ptr<function<void (const Observable&)>> theAction) {
action = theAction;
}
void EventRaised () {
if (!action.expired ()) {
auto theAction = action.lock ();
(*theAction) (*this);
}
}
};
struct Observer {
...
void CallOnEvent (const Observable & observable) {
// do something
}
// field to store the action as long as it is needed
std::shared_ptr<function<void (const Observable&)>> action;
void ... {
auto me = this;
action = std::make_shared<function<void (const Observable&)>> (
[me] (const Observable& observable) {
me->CallOnEvent (observable);
}
);
// we could have as well used std::bind
observable.AddEventListener (action);
}
};

Perhaps you should just create a class to be derived by "users":
class Action {
public:
friend class EventDispatcher;
virtual SomeResultType DoThis() = 0;
private:
/* Some common data */
};
Just pass some derived-from-class-Action typed variable to AddEventListener. When the corresponding event is triggered, just fill in the common data and call the DoThis() method.
void EventDispatcher::DispatchEvent ( string event )
{
int key = 0;
while ( key < this->subscriptions.size() )
{
Subscription subscription = this->subscriptions[key];
if ( subscription.event == event )
{
subscription->action();
};
};
};
For AddEventListener:
void EventDispatcher::AddEventListener ( Observer* observer , string event , Action* action )
{
Subscription subscription = { observer , event , action );
this->subscriptions.push_back ( subscription );
};
An example of a Action derived class:
class myAction: public Action {
public:
// Implement the DoThis() method
void SomeResultType DoThis() {
cout << "Hello World!";
return SomeValue;
}
};
// To use the action,
myAction* act = new myAction;
myEventDispatcher.AddEventListener(someObserver, "HelloWorld", act);
This is one of the safest way to implement actions (and callbacks).

In its simplest form u_u could be a function pointer e.g.
typedef void (*u_u)(void*); // or whatever arguments u like
then you just supply a function that is called whenever the event is triggered.
void myaction(void* arg)
{
...
}
Subscription s;
...
s.action = myaction;

Related

Passing a TForm as an argument to a function

I have an application with several Forms. Two of them are quite similar, they have features in the form of VCL objects (labels, images, etc...) in common, which I named the same.
I want to have a function in a specific class which can accept one of these two Form as a parameter in order to modify the parameters that they have in common. The solution I came around does not seem to work.
As my application is quite big and complicated, I replicated the problem using a small example.
First, below is an example of my MainForm :
And an example of one subForm (they are all arranged in a similar way)
I have an additionnal class which is used to fill in the Edits on the subForms. The code for this class is the following:
#pragma hdrstop
#include "master_class.h"
#include "sub_Form2.h"
#include "sub_Form3.h"
#include "sub_Form4.h"
//---------------------------------------------------------------------------
#pragma package(smart_init)
Master::Master(void)
{
}
Master::~Master(void)
{
}
void Master::WriteToForm(TForm* Form)
{
TForm2* curForm = static_cast<TForm2*>(Form);
TForm3* self = dynamic_cast<TForm3*>(Form);
TForm2* self2 = dynamic_cast<TForm2*>(Form);
if (self != NULL && self2 == NULL) {
TForm3* curForm = static_cast<TForm3*>(Form);
}
else if (self == NULL && self2 != NULL) {
TForm2* curForm = static_cast<TForm2*>(Form);
}
curForm -> Edit1 -> Text = "blablabla_1";
curForm -> Edit2 -> Text = "blablabla_2";
}
And in the MainForm, the code for the "Fill Form2" button is the following:
Master1 -> WriteToForm(Form2);
where Master1 is just an object of the Master class.
This works very well for Form2 :
But for Form3, which is filled up using Master1 -> WriteToForm(Form3), here is what I get, which the same pb than in my real application:
So what should go to the Edit, is misplaced. I think the main pb comes from the fact that I did not create every label, edit, etc... on the same order. I did that on purpose to mimic my real application. To verify this, I created a 3rd subForm, where this time the VCL objects were created in the same order as my first subForm, and this works:
So I would suspect that this comes from the initial cast
TForm2* curForm = static_cast<TForm2*>(Form);
When I pass Form3 as an argument, Form3 is somewhat casted into the "shape" of Form2, which is not defined in the same order. Maybe this could be corrected by modifying directly the DFM file, but it is not a realistic approach for my main app.
I do this initial cast otherwise I get a compilation error saying that curForm is not known at the first line
curForm -> Edit1 -> Text = "blablabla_1";
So, is there a better way to pass the Form as an argument to the WriteToForm function?
Just because two types are similar does not mean they are related. Your code does not work because your two Form classes are not related to each other in any way. You can't just cast one to the other arbitrarily.
To solve this, you have several options:
code for both Form classes separately, eg:
void Master::WriteToForm(TForm* Form)
{
TForm2* curForm2 = dynamic_cast<TForm2*>(Form);
TForm3* curForm3 = dynamic_cast<TForm3*>(Form);
if (curForm2)
{
curForm2->Edit1->Text = _D("blablabla_1");
curForm2->Edit2->Text = _D("blablabla_2");
}
else if (curForm3)
{
curForm3->Edit1->Text = _D("blablabla_1");
curForm3->Edit2->Text = _D("blablabla_2");
}
}
Or:
void WriteToForm(TForm2* Form);
void WriteToForm(TForm3* Form);
...
void Master::WriteToForm(TForm2* Form)
{
Form->Edit1->Text = _D("blablabla_1");
Form->Edit2->Text = _D("blablabla_2");
}
void Master::WriteToForm(TForm3* Form)
{
Form->Edit1->Text = _D("blablabla_1");
Form->Edit2->Text = _D("blablabla_2");
}
Make your function use a template (however, be aware of this: Why can templates only be implemented in the header file?):
template<typename T>
void WriteToForm(T* Form);
...
void Master::WriteToForm<T>(T* Form)
{
Form->Edit1->Text = _D("blablabla_1");
Form->Edit2->Text = _D("blablabla_2");
}
make the two Form classes derive from a common base class or interface, eg:
class TBaseForm : public TForm
{
public:
inline __fastcall TBaseForm(TComponent *Owner) : TForm(Owner) {}
virtual void SetEdit1(const String &Text) = 0;
virtual void SetEdit2(const String &Text) = 0;
};
...
class TForm2 : public TBaseForm
{
...
public:
__fastcall TForm2(TComponent *Owner);
...
void SetEdit1(const String &NewText);
void SetEdit2(const String &NewText);
};
__fastcall TForm2::TForm2(TComponent *Owner)
: TBaseForm(Owner)
{
...
}
void TForm2::SetEdit1(const String &NewText)
{
Edit1->Text = NewText;
}
void TForm2::SetEdit2(const String &NewText)
{
Edit2->Text = NewText;
}
...
repeat for TForm3...
...
void Master::WriteToForm(TBaseForm* Form)
{
Form->SetEdit1(_D("blablabla_1"));
Form->SetEdit2(_D("blablabla_2"));
}
Or:
__interface INTERFACE_UUID("{E900785E-0151-480F-A33A-1F1452A431D2}")
IMyIntf : public IInterface
{
public:
virtual void SetEdit1(const String &Text) = 0;
virtual void SetEdit2(const String &Text) = 0;
};
...
class TForm2 : public TForm, public IMyIntf
{
...
public:
__fastcall TForm2(TComponent *Owner);
...
void SetEdit1(const String &NewText);
void SetEdit2(const String &NewText);
};
__fastcall TForm2::TForm2(TComponent *Owner)
: TForm(Owner)
{
...
}
void TForm2::SetEdit1(const String &NewText)
{
Edit1->Text = NewText;
}
void TForm2::SetEdit2(const String &NewText)
{
Edit2->Text = NewText;
}
...
repeat for TForm3...
...
void Master::WriteToForm(IMyIntf* Intf)
{
Intf->SetEdit1(_D("blablabla_1"));
Intf->SetEdit2(_D("blablabla_2"));
}
use RTTI to access the fields, eg:
#include <System.Rtti.hpp>
void Master::WriteToForm(TForm* Form)
{
TRttiContext Ctx;
TRttiType *FormType = Ctx.GetType(Form->ClassType());
TRttiField *Field = FormType->GetField(_D("Edit1"));
if (Field)
{
TValue value = Field->GetValue(Form);
if( (!value.Empty) && (value.IsObject()) )
{
TObject *Obj = value.AsObject();
// Either:
static_cast<TEdit*>(Obj)->Text = _D("blablabla_1");
// Or:
TRttiProperty *Prop = Ctx.GetType(Obj->ClassType())->GetProperty(_D("Text"));
if (Prop) Prop->SetValue(Obj, String(_D("blablabla_1")));
}
}
Field = FormType->GetField(_D("Edit2"));
if (Field)
{
TValue value = Field->GetValue(Form);
if( (!value.Empty) && (value.IsObject()) )
{
TObject *Obj = value.AsObject();
// Either:
static_cast<TEdit*>(Obj)->Text = _D("blablabla_2");
// Or:
TRttiProperty *Prop = Ctx.GetType(Obj->ClassType())->GetProperty(_D("Text"));
if (Prop) Prop->SetValue(Obj, String(_D("blablabla_2")));
}
}
}

capnproto : Can I get an AnyPointer from a struct?

Given this schema
struct TestObject
{
value1 #0 : Int32 = -5;
value2 #1 : Float32 = 9.4;
}
struct TestContainer
{
object #0: TestObject;
}
Is it possible to get an AnyPointer::Builder from the TestObject::Builder in c++ code?
This is what I am trying to do:
::capnp::MallocMessageBuilder message;
auto container = message.initRoot<TestContainer>();
TestObject::Builder objBuilder = container.initObject();
//Get an AnyPointer
capnp::AnyPointer::Builder anyBuilder = capnp::toAny( objBuilder )(); //No this does not work.
MyTestObject test( 41, 643.7f );
test.serialise( anyBuilder );
What I am trying to do is have an abstract interface with a single argument type
eg.
class ISerialisable
{
virtual void serialise(capnp::AnyPointer::Builder& any) = 0;
}
class MyTestObject: public ISerialisable
{
void serialise(capnp::AnyPointer::Builder& any) override
{
auto testObjBuilder = any.getAs<TestObject>(); or should initAs be used?
testObject.setValue1( whatever1);
testObject.setValue2( whatever2);
}
}
Is it possible to go down this route?

unit test - mockito testing to check if another method is invoked

Summary: I am trying to test if a method is invoked once I call one method.
What this class does is, displays information of wrong spelled words and provides u with buttons to 'ignore' or 'ignore all' or 'add to dictionary', etc.
Over here 'ignore' is a JButton declared above.
I am trying to write one test for this method ->
public class SpellCheckerDialog extends JDialog implements ActionListener {
...
..
public void actionPerformed( ActionEvent ev ) {
Object source = ev.getSource();
if( source == ignore ) {
searchNext();
}
}
...
}
Here is what it is invoking, I am testing to see if this method is being invoked or not.
...
//inside same class
public boolean searchNext() {
String wordStr;
while( true ) {
wordStr = tok.nextInvalidWord();
if( wordStr == null ) {
dispose();
String title = SpellChecker.getApplicationName();
if(title == null){
title = this.getTitle();
}
SpellChecker.getMessageHandler().handleInformation( getParent(), title, Utils.getResource( "msgFinish" ) );
return false;
}
if( ignoreWords.contains( wordStr ) ) {
continue;
}
String changeTo = changeWords.get( wordStr );
if( changeTo != null ) {
replaceWord( wordStr, changeTo );
continue;
}
break;
}
word.setText( wordStr );
notFound.setText( wordStr );
List<Suggestion> list = dictionary.searchSuggestions( wordStr );
boolean needCapitalization = tok.isFirstWordInSentence() && Utils.isFirstCapitalized( wordStr );
Vector<String> suggestionsVector = new Vector<String>();
for( int i = 0; i < list.size() && i < options.getSuggestionsLimitDialog(); i++ ) {
Suggestion sugestion = list.get( i );
String newWord = sugestion.getWord();
if( needCapitalization ) {
newWord = Utils.getCapitalized( newWord );
}
if( i == 0 )
word.setText( newWord );
suggestionsVector.add( newWord );
}
suggestionsList.setListData( suggestionsVector );
addToDic.setEnabled( true );
return true;
}
What I have tried until now, tried using Mockito and calling the verify method, but this code snippet seems to not working or have lots of dependencies that I am struggling to get around.
Inside my TestClass, I have this - >
Dialog fr = Mockito.mock(Dialog.class);
SpellCheckerDialog sD = new SpellCheckerDialog(fr);
sD.searchNext();
Mockito.verify(sD, Mockito.times(1)).thenReturn(searchNext());
I don't know if I should be making a stub for my (ActionEvent ev) or ...
Verifications must be done on the mocks created by Mockito, because the framework can't possibly know what happens with objects that it does not manage. This being said, your searchNext() method is part of your class under test so you probably want to spy on it just like in the example below:
public class SpyTest {
class MyClass {
public void callDoSomething(){
doSomething();
}
public void doSomething(){
// whatever
}
}
#Test
public void shouldSpyAndVerifyMethodCall(){
MyClass objectUnderTest = new MyClass();
MyClass spy = Mockito.spy(objectUnderTest);
spy.callDoSomething();
Mockito.verify(spy, Mockito.times(1)).doSomething();
}
}
My advice is to go through the Mockito documentation and examples from the link above, as they're pretty straight-forward and should give you a good starting point.
EDIT as per your comment:
public class SpyTest {
class MyClass {
private JButton myButtton;
public void actionPerformed(ActionEvent event){
if(event.getSource() == myButtton) {
searchNext();
}
}
public void searchNext(){
// whatever
}
}
#Mock // define a mock for the button to "hack" the source check
private JButton mockButton;
#InjectMocks // inject the mock in our object under test
private MyClass objectUnderTest;
#Test
public void shouldSpyAndVerifyMethodCall(){
// spy on our object so we can query various interactions
MyClass spy = spy(objectUnderTest);
// event mock
ActionEvent mockEvent = mock(ActionEvent.class);
// "hack" the source check
when(mockEvent.getSource()).thenReturn(mockButton);
// call main logic
spy.actionPerformed(mockEvent);
// verify interactions
verify(spy).searchNext(); // times(1) not needed because it's the implicit/default setting, see David's comment
}
}

How to send a message to the class that created the object

I would like to send back data to class that create this object.
It's game related.
The enemy objects have a threaded function and move on their own in the scene.
It generates lots of errors if you include the header file from the class that creates to the objects into the object itself ... to pass pointers.
Enemy Class:
Class Enemy
{
private:
void (*iChange)(DWORD &);
}:
Enemy::Enemy(void (*iChangeHandler)(DWORD &) ) : iChange(0)
{
this->iChange = iChangeHandler;
}
void Enemy::Draw(D3DGraphics& gfx)
{
this->iChange(this->dwThreadID); // send a message back to the class that created me
gfx.PutPixel(this->my_position_x + 0,this->my_position_y,this->red,this->blue,this->green);
this->grafix->DrawCircle(this->my_position_x + 0,this->my_position_y, this->radius, this->red,this->blue,this->green);
(sprintf)( this->enemy_buffer, "X: %d, Y: %d", this->my_position_x , this->my_position_y);
this->grafix->DrawString( this->enemy_buffer, this->my_position_x , this->my_position_y, &fixedSys, D3DCOLOR_XRGB(255, 0, 0) );
}
Game Class:
struct enemies_array_ARRAY {
std::string name;
DWORD ID;
Enemy* enemy;
} enemies_array[25];
void Game::EnemyEvent(DWORD &thread_id)
{
enemies_array[0]...... // i want to acces this struct array
}
Game::Game(HWND hWnd)
{
enemies_array[0].name = "john Doe";
enemies_array[0].ID = NULL;
enemies_array[0].enemy = new Enemy(&Game::EnemyEvent);
// error: C2664:
// another attemp
enemies_array[0].name = "john Doe";
enemies_array[0].ID = NULL;
enemies_array[0].enemy = new Enemy(Game::EnemyEvent);
// error C3867:
}
If I understand correctly, you want to call a function on the Game object. This means you need to pass a pointer to the Game object in order to correctly call a non static member function pointer(iChange) on it.
Make the changes shown below and you should be able to do what you want
enemies_array[0].enemy =  new Enemy(this,&Game::EnemyEvent);   
typedef void (Game::*ChangeFunc)(DWORD &)
Class Enemy
{
private:
ChangeFunc iChange;
Game *pGame;
}:
Enemy(Game *pCreatorGame, ChangeFunc iChangeHandler )
{
iChange = iChangeHandler;
pGame = pCreatorGame;
}
void Enemy::Draw(D3DGraphics& gfx)
{
(pGame->*iChange)(this->dwThreadID);

Memory Alignment issue with std::map on Linux

I have run into a problem while working with c++ over Linux.
I have a base Message class which looks like this:
class MsgBase
{
public:
MsgBase( unsigned int msgid );
map < unsigned int, MSGIEBase* > messageIE_Map; // map for IEs
unsigned int messageId; // 32 bit message id
};
class Der1 is derived from MsgBase and looks like:
class Der1 : public MsgBase
{
public:
Der1 ();
virtual ~Der1 ();
// IEs
MSGIE_UINT32 ueId;
MSGIE_String configFileName;
};
Here MSGIE_UINT32 and MSGIE_String are classes derived from MSGIEBase and therefore their address can be stored in the map defined in base class above.
When Der1 is constructed the address of ueId and configFileName is stored in the map.
Here if I print the size of map ( through gdb and in the program ) it comes to be 24.
[ _M_header = 16, _M_node_count = 4, _M_key_compare = 1, 3 byte padding I suppose ].
Till here everything is fine. Now the Der1 object pointer is put inside an event object and the event is post into a queue. The event class looks like:
class Event
{
public:
MsgBase* msgPtr;
};
A different thread removes the event from the queue, extracts the msgPtr and casts it into Der1 Pointer and this is where the problem starts.
Here if I print the size of the map in the program it is 21. That means the address of the next member in the MsgBase class i.e. messageId gets shifted by 3 bytes and so the value of messageId changes completely. (when seen through gdb, the address is still intact and so is the size of the map i.e. 24 ).
This is a word alignment issue to the best of my knowledge but why is the memory alignment not consistent in different functions and why does the address of a member of a class chage when the memory to the class has been allocated using new. I am using Linux 2.6.27. , gcc version 4.1.1. , RHEL-4.
For ruling out non-virtual destructor/copy/assignment problems, please add the following to MsgBase:
public:
virtual ~MsgBase();
private:
MsgBase(MsgBase const& other);
MsgBase& operator=(MsgBase const& other);
I will try to provide all the required information step by step:
Information 1 : The relevant code.
//Step 1: Create msg and fill message Id
MsgBase*msgPtr = new Der1();
// C'tor of Der1 is as follows:
Der1::Der1 ()
: MsgBase ( ATS_SUTD_EPCTESTER_ATTACH_SCENARIO_MsgId ), // msgid is 13( 0xd )
ueId ( IE_UE_KEY, "UE", false ),
configFileName ( IE_CONFIG_FILE_NAME_KEY, "Configuration File Name", false )
{
// Insert the IEs in the map
this->addIEEntry ( IE_UE_KEY, &ueId ); // this puts entries in the map
this->addIEEntry ( IE_CONFIG_FILE_NAME_KEY, &configFileName );
}
// Step 2: Declare event and post the event
Event* event = new Event ( eventId, "Event" );
event->setData( msgPtr, hdr);
// check the message id at this stage (
cout << "msgId = " << ( ( (Der1* )msgPtr )->messageId )<< endl; // Here it comes out
// to be 0xd which is correct
// post the event
AppClass::getInstance()->addEventAndSchedule ( event );
//The queue is a member of AppClass and has been defined as
std::list <EventBase* > eventQueue;
// The code which inserts data into the queue is as follows:
bool AppClass::addEventAndSchedule ( EventBase* ev )
{
if ( ev == NULL ) return false;
this->eventQueueMutex.acquireLock();
this->eventQueue.push_back( ev );
this->eventQueueMutex.releaseLock();
// Submit Job to Scheduler
bool status = JobScheduler::getInstance()->scheduleJob( this );
return status;
}
// The event class is
class Event: public EventBase
{
public:
Event ();
virtual ~Event ();
Event ( int evId );
Event ( int evId, string evName );
MsgBase* getMessagePtr ();
void setData ( MsgBase* mPtr, Header* hPtr )
private:
// Prevent copying
Event& operator= ( Event& ev );
Event ( Event& evBase );
MsgBase* msgPtr;
Header* hdrPtr;
};
void Event::setData ( MsgBase* mPtr, Header* hPtr )
{
this->msgPtr = mPtr;
this->hdrPtr = hPtr;
}
Step 3 : Extract the event and re-print the message Id
// The code which extracts data from the queue is as follows:
void AppClass::process ()
{
EventBase* beventPtr = NULL;
this->eventQueueMutex.acquireLock();
if ( !this->eventQueue.empty() )
{
beventPtr = (EventBase* )( this->eventQueue.front() );
this->eventQueue.pop_front();
}
else
{
isQueueEmpty = true;
}
this->eventQueueMutex.releaseLock();
Event* eventPtr = ( Event* )beventPtr ;
Der1* msgPtr = (Der1* )( eventPtr->getMessagePtr()) ;
cout << "msgId = " << msgPtr->messageId << endl; // This value
//comes out to be incorrect it is now 0xd000000 i.e. a 3 byte shift
}
Information 2 : Exact problem.
The exact problem is that the 'messasgeId' is getting changed in transition. Initially it is 0xd but after popping from the queue it becomes 0xd000000. Because of this all the processing stops. The address of this parameter also changes from 0x82bd7cc to 0x82bd7c9 when printed in the program. However when seen from gdb it is still 0x82bd7cc and the value is still 0xd.
Information 3 : Compiler Flags.
Compiler Flags are same for all the files and they are:
-O0 -g3 -Wall -fmessage-length=0