Using GLFW to capture mouse dragging? [C++] - c++

So i'm trying to capture mouse dragging in my OpenGL application. I've done the following so far:
glfwSetMouseButtonCallback(window, mouse_callback);
static void mouse_callback(GLFWwindow* window, int button, int action, int mods)
{
if (button == GLFW_MOUSE_BUTTON_LEFT) {
double x;
double y;
glfwGetCursorPos(window, &x, &y);
if (previous_y_position - y > 0)
{
camera_translation.y -= 1.0f;
previous_y_position = y;
}
else
{
camera_translation.y += 1.0f;
previous_y_position = y;
}
}
}
The problem with this though is if I would like to zoom in I need to move my mouse upwards and then click repeatedly. For some reason if I press down on the left mouse button and drag upwards it does nothing.

In cursor_pos_callback, just confirm if the button is pressed, and that works.
void mouse_cursor_callback( GLFWwindow * window, double xpos, double ypos)
{
if (glfwGetMouseButton(window, GLFW_MOUSE_BUTTON_LEFT) == GLFW_RELEASE)
{
return;
}
// `write your drag code here`
}

mouse_callback is stateless. It receives events, momentary "actions".
You need to make your program to "remember" that mouse button is pressed. So that when button is pressed in a frame 1, you can refer to this information in all the frames after that and before mouse button is released.
The simple way is to flip a boolean flag on press/release:
static void mouse_callback(GLFWwindow* window, int button, int action, int mods)
{
if (button == GLFW_MOUSE_BUTTON_LEFT) {
if(GLFW_PRESS == action)
lbutton_down = true;
else if(GLFW_RELEASE == action)
lbutton_down = false;
}
if(lbutton_down) {
// do your drag here
}
}
Schematically:
state released pressed released
timeline -------------|------------------------------|---------------
^ ^
mouse_callback calls GLFW_PRESS GLFW_RELEASE
The hard way is to use a state machine (especially if you need more complicated combinations of states of input controllers).

Related

OpenGL GLUT left mouse click overrides holding right mouse

I'm adapting some skeleton code to learn how OpenGL works and have in SphericalCameraManipulator.cpp which allows me to pan and tilt the camera while I hold down right mouse:
void SphericalCameraManipulator::handleMouseMotion(int x, int y)
{
//Motion
float xMotion = (float)x - previousMousePosition[0];
float yMotion = (float)y - previousMousePosition[1];
if(reset)
{
xMotion = yMotion = 0.0;
reset = false;
}
this->previousMousePosition[0] = (float)x;
this->previousMousePosition[1] = (float)y;
//Right Button Action
if(this->currentButton == GLUT_RIGHT_BUTTON && this->currentState == GLUT_DOWN)
{
this->pan -= xMotion*this->panTiltMouseStep ;
this->tilt += yMotion*this->panTiltMouseStep ;
}
//Enforce Ranges
this->enforceRanges();
}
I deleted the left mouse action as I don't want it to control the camera, and it now actions a command in the main code body.
//Left Button Action
if(button == GLUT_LEFT_BUTTON && state == GLUT_DOWN)
{
... //game action
}
My problem is that when I click the left mouse, the press of the right mouse is cancelled, and I must release and repress the right mouse to continue controlling the camera.
Is there a way to prevent this from happening? It interrupts the flow of the game. I'm using GLUT
Rabbid76's comment saying there is no mouse hold event set me on the right path.
I wrote a small function that simply recorded the last state of mouse button 2 (the 'camera look around' button):
void SphericalCameraManipulator::handleMouse(int button, int state)
{
this->currentButton = button;
this->currentState = state;
//to enable mouse look around
if (this->currentButton == 2) //right mouse
this->lastLookState = this->currentState;
if(this->currentState == GLUT_UP)
{
reset = true;
}
}
This mean even when the current state was regarding the left mouse button (game action) it was still possiblew to look around since a 'mouse two button up' event had not occured:
this->currentButton = 2;
this->currentState = 1;

Why does mouse down block repainting of a component?

I am trying to create a custom component that captures mouse events, especially MouseMove.
I derive from TWinControl, but I have also tried with TGraphicControl, TCustomControl, TTrackBar, etc.
My problem is when I hold down the mouse on the component, it is not being repainted.
The Paint() method is not called until I release the mouse button, even if I call Invalidate().
A TrackBar is the closest component I want to make. You select the tick, and move it around with the mouse. But you do not have to release the mouse to see the tick move at the same time (the component is drawn again).
If I directly call Paint(), it works, but the background is not erased.
What am I missing?
EDIT :
I tried again and i confirm if i held the mouse down, Invalidate(); call are take in account only when i release the mouse.
Try your self with my code below, paint is only call on release :
__fastcall TMyCustomComponent::TMyCustomComponent(TComponent* Owner)
: TCustomTransparentControl(Owner)
{
mValue = 0;
}
void __fastcall TMyCustomComponent::MouseDown(System::Uitypes::TMouseButton Button, System::Classes::TShiftState Shift, int X, int Y)
{
if (Button == mbLeft)
{
mValueStart = 0;
}
}
void __fastcall TMyCustomComponent::MouseMove(System::Classes::TShiftState Shift, int X, int Y)
{
Invalidate();
}
void __fastcall TMyCustomComponent::Paint(void)
{
TGraphicControl::Paint();
Canvas->Font->Name = "Arial";
Canvas->Font->Size = 8;
Canvas->Font->Style = TFontStyles() << fsBold;
Canvas->Font->Color = clInfoText;
Canvas->Brush->Color = clInfoBk;
Canvas->FillRect(TRect(0, 0, 104, 21));
mValue++;
Canvas->TextOut(0, 2, AnsiString(mValue));
Canvas->Brush->Color = clBtnShadow;
}
The following works fine for me:
__fastcall TMyCustomComponent::TMyCustomComponent(TComponent* Owner)
: TCustomTransparentControl(Owner)
{
mValue = 0;
InterceptMouse = true; // <-- needed for TCustomTransparentControl to trigger Mouse...() methods!
}
void __fastcall TMyCustomComponent::MouseDown(System::Uitypes::TMouseButton Button, System::Classes::TShiftState Shift, int X, int Y)
{
if (Button == mbLeft)
{
mValue = 0;
Invalidate();
}
TCustomTransparentControl::MouseDown(Button, Shift, X, Y);
}
void __fastcall TMyCustomComponent::MouseMove(System::Classes::TShiftState Shift, int X, int Y)
{
++mValue;
Invalidate();
TCustomTransparentControl::MouseMove(Shift, X, Y);
}
void __fastcall TMyCustomComponent::Paint()
{
TCustomTransparentControl::Paint();
Canvas->Font->Name = "Arial";
Canvas->Font->Size = 8;
Canvas->Font->Style = TFontStyles() << fsBold;
Canvas->Font->Color = clInfoText;
Canvas->Brush->Color = clInfoBk;
Canvas->FillRect(TRect(0, 0, ClientWidth, ClientHeight));
Canvas->TextOut(0, 2, String(mValue));
Canvas->Brush->Color = clBtnShadow;
}
Pressing down the left mouse button resets mValue to 0 and paints it. And moving the mouse around the control increments mValue and paints it, whether the mouse button is held down or not.

SDL C++ Mouse Input Handling Toggle Effect

So I'm working on a tower defense game. And I need to select a tile where the user clicks the mouse (right now i'm just drawing a border around the tile). Now the code I have works sometimes and I'm not sure how to get it to work better. Right now with the code I have when I click on an area, the tile gets selected and sometimes it deselects right away and other times it doesn't. Almost like I'm holding the mouse down and it just keeps "clicking" the mouse. What I need is for it to select the tile without deselecting it no matter how long the mouse button is being pressed. My mouse input class is set up like this:
MouseInput.h
#pragma once
#include <SDL.h>
class MouseInput
{
public:
MouseInput();
~MouseInput();
void Update();
bool GetMouseButton(int index);
SDL_Event GetEvent();
bool GetPressed();
bool GetReleased();
private:
void GetButtonStates();
private:
bool mouseArray[3];
int x, y;
bool justReleased;
bool justPressed;
SDL_Event e;
};
MouseInput.cpp
#include "Input.h"
MouseInput::MouseInput()
:
x(0),
y(0),
justReleased(false),
justPressed(false)
{
SDL_PumpEvents();
for (int i = 0; i < 3; i++)
{
mouseArray[i] = false;
}
}
MouseInput::~MouseInput()
{
}
void MouseInput::Update()
{
SDL_PollEvent(&e);
justReleased = false;
justPressed = false;
if (e.type == SDL_MOUSEBUTTONDOWN && e.button.button == SDL_BUTTON_LEFT)
justPressed = true;
else if (e.type == SDL_MOUSEBUTTONUP && e.button.button == SDL_BUTTON_LEFT)
justReleased = true;
}
bool MouseInput::GetPressed()
{
return justPressed;
}
bool MouseInput::GetReleased()
{
return justReleased;
}
bool MouseInput::GetMouseButton(int index)
{
if (index <= 3 && index >= 1)
{
return mouseArray[index - 1];
}
return false;
}
SDL_Event MouseInput::GetEvent()
{
return e;
}
void MouseInput::GetButtonStates()
{
SDL_PollEvent(&e);
if (e.type == SDL_MOUSEBUTTONDOWN)
{
// 1 = Left; 2 = Center; 3 = Right
mouseArray[e.button.button - 1] = true;
}
if (e.type == SDL_MOUSEBUTTONUP)
{
mouseArray[e.button.button - 1] = false;
}
}
In my Game class I have setup the mouseinput as well as a variable to store the coordinates of the tile that was clicked. I also Have a variable pressed to check to see if the mouse was pressed. In my updating function I check if the mouse was not pressed. If the mouse was not pressed, then I check if the mouse is pressed that way the game can get the coordinates of the mouse position. I know that sounds crazy so here is some of my code:
std::pair<int, int> coords; // Will store the coordinates of the tile position that the mouse rests in
bool pressed; // If the mouse was pressed (i should call it toggle because it's true if the mouse was pressed once, then false if the mouse was pressed again, then true...ect.)
Updating function
Mouse Class
bool clicked; // In my custom Mouse.h file and set to false in constructor
MouseInput input; // In my custom Mouse.h file
Mouse class Update()
input.Update(); // Calls the MouseInput updating
if (input.GetPressed() && !clicked)
{
clicked = !clicked;
printf("OK");
}
//else if (clicked && input.GetReleased())
else if (clicked && input.GetPressed()) // Somewhat works. It works almost like before. If I click the mouse it shows up then sometimes when I click again it works, and other times I have to click a couple times
{
clicked = !clicked;
}
In my game class I just check if mouse clicked is true, then draw what i need
if (mouse.GetClicked())
{
// Draw what I need here.
// Currently is only drawn when I hold down the mouse
// And not drawn when I release the mouse.
}
Then when I'm drawing the square around the tile, I only draw it if pressed = true. Otherwise it is not drawn. This code works sometimes, and other times it doesn't. In order to get the effect I want, I have to real quickly click the mouse. If I don't, sometimes it appears as if I'm holding the mouse button down and the square just flashes.
The issue is that the "mouse-down" event is exactly that - an event. It should happen once, be handled, and then not occur until the button is physically pressed again. With your code, once pressed it'll set the internal state to "pressed" and then not change it until it's released. Which is okay, except your code polling this internal state treats it as if it's an event, not as state, so every time the update function runs while the mouse button is pressed, it'll see a "new button press".
This (simplified version of your code that uses one button and an idea of showing an image only when a button's pressed) might help you visualise it:
bool buttonPressed = false;
bool imageShown = false;
while (true)
{
SDL_Event e;
SDL_PollEvent(&e);
if (e.type == SDL_MOUSEBUTTONDOWN)
buttonPressed = true;
else if (e.type == SDL_MOUSEBUTTONUP)
buttonPressed = false;
if (!imageShown && buttonPressed)
imageShown = true;
else if (buttonPressed)
imageShown = false;
}
Do you see how each time the loop runs, even if no events have been fired this iteration, the state remains? This would cause the imageShown variable to keep toggling on and off each iteration, until a "button up" event is fired to reset the state.
You should probably rework a bit of your input controller - you need to expose a way to differentiate from "button being pressed right now" and "button that's been pressed for a while".
An idea would be to provide eg. ButtonsPressed and ButtonsJustPressed members (although I'm sure you can think of more appropriate names). The ButtonsJustPressed members should be reset to false each time you call GetButtonStates, so they're only true if on that specific iteration of the loop, they've been pressed, while the ButtonsPressed members are exactly what you're doing right now.
A tweaked version of the above to take this into account:
bool buttonJustPressed = false;
bool buttonJustReleased = false;
bool imageShown = false;
while (true)
{
// As far as we know, this iteration no events have been fired
buttonJustPressed = false;
buttonJustReleased = false
SDL_Event e;
SDL_PollEvent(&e);
if (e.type == SDL_MOUSEBUTTONDOWN)
buttonJustPressed = true;
else if (e.type == SDL_MOUSEBUTTONUP)
buttonJustReleased = true;
if (!imageShown && buttonJustPressed)
imageShown = true;
else if (imageShown && buttonJustReleased)
imageShown = false;
}
See how this version will only change the visibility of the image if a button has just been pressed/released - if the button's being held down, the events aren't fired, so the state remains false, and the image's state remains the same.

How to continuously select content in TWO RichEdit controls?

I'm currently working on project which will have two RichEdit controls one close to the other: let's say RichEdit1 is on left and RichEdit2 is on the right.
The user scenario I want to enable in the project is:
User mouse LButton down at somewhere in RichEdit1, e.g. before the 3rd char, in total 7 chars.
User drag the mouse to RichEdit2, e.g. after the 6th char, in total 11 chars.
User mouse LButton up.
I want to see both RichEdit1 3rd char to end and RichEdit2 begin to 6th char are selected.
Currently I notice that once mouse LButton down on RichEdit1, after I move the mouse to RichEdit2, the RichEdit2 could not receive mouse event before I release mouse.
Any suggestion will be appreciated. Thank you!
When the mouse button is pressed down on RichEdit1, it captures the mouse, thus subsequent mouse messages are sent to RichEdit1 until the mouse button is released. That is why RichEdit2 does not receive any mouse events while dragging over RichEdit2.
You will have to process the mouse move messages in RichEdit1 and check if their coordinates are outside of RichEdit1's client area. If so, convert them into coordinates relative to RichEdit2's client's area and then send EM_SETSEL/EM_EXSETSEL messages to RichEdit2 as needed. For example:
int RichEdit2StartIndex = -1;
...
// in RichEdit1's message handler...
case WM_MOUSEMOVE:
{
if ((wParam & MK_LBUTTON) == 0)
break;
int xPos = GET_X_LPARAM(lParam);
int yPos = GET_Y_LPARAM(lParam);
RECT r;
GetClientRect(hwndRichEdit1, &r);
if (xPos < (r.right - r.left))
{
if (RichEdit2StartIndex != -1)
{
SendMessage(hwndRichEdit2, EM_SETSEL, -1, 0);
RichEdit2StartIndex = -1;
}
}
else
{
POINT pt;
pt.x = xPos;
pt.y = yPos;
MapWindowPoints(hwndRichEdit1, hwndRichEdit2, &pt, 1);
POINTL pl;
Pl.x := pt.x;
Pl.y := pt.y;
int idx = SendMessage(hwndRichEdit2, EM_CHARFROMPOS, 0, (LPARAM)&Pl);
if (idx != -1)
{
if (RichEdit2StartIndex == -1)
RichEdit2StartIndex = idx;
SendMessage(hwndRichEdit2, EM_SETSEL, RichEdit2StartIndex, idx);
}
}
break;
}
Vice versa when dragging a selection from RichEdit2 to RichEdit1.
And make sure that both RichEdit controls have the ES_NOHIDESEL style applied so you can see the selection in both controls at the same time.

How to respond when mouse scroll event occurs?

I would like to know how to respond to a mouse scroll event in C++ only from the user's mouse wheel. I don't really care whether the user scrolls in or out as long as I catch the event that they indeed did scroll with the mouse. Can anyone show me how I can use this for an event?
You need to implement the glutMouseFunc callback like this:
void mouse(int button, int state, int x, int y)
{
// Wheel reports as button 3(scroll up) and button 4(scroll down)
if ((button == 3) || (button == 4))
{
if (state == GLUT_UP) return; // Disregard redundant GLUT_UP events
(button == 3) ? DO_ACTION : DO_SOMETHING_ELSE;
}
glutPostRedisplay();
}