Sign in to follow this  
SteveAnderson

Mapping Player Actions to one button

Recommended Posts

This is more of a question how other people would handle this issue. I am making a sim farm game. I have all logic for animating the player, objects, simple quests and things need need to do etc. I am now thinking of how I am going to interact with different objects. For example, my character when you click 'e' could do a number of things depending on what she is interacting with. If the item is a grass tile she will hoe it, if it's a dirt tile she could put down seeds, if it's a crop then she will water it, if it's a mailbox check the mail or something.

In more simple games I have just had a bit if statement in the update method. I don't want to do this because I want it to be more robust and expandable. I was thinking that when the action button is clicked an action object is sent to whatever objects are clicked and if they can handle them then they will themselves? Almost like a chain of command pattern.

like
WorldAction
HoeAction : BaseAction
WaterAction : BaseAction
CheckMailAction : BaseAction

so
phseudo

update()
if input = e then
selectedTool = selectedinventory
objects = getallselectedobjects()
for each object in objects
handletoolaction()
handlewhateverotheractions

I don't know if this all makes sense, perhaps I just need a bit of a hand organizing and focusing. I appreciate any input from you guys. If any of it is unclear or I am just not making sense I can try to elaborate more with code etc.

Share this post


Link to post
Share on other sites
I would create an abstract base class "UsableObject" with an EventHandler for click events; that way every subclass of that (HoeAction, WaterAction etc.) just needs to implement onClick with its own logic.

Share this post


Link to post
Share on other sites
I think fastest would be if you track what item you have (one item at the time), and then only ask
if(pressed(e))
{
if(item==mail)
do mail job
if(item==water)
do water job
}

Share this post


Link to post
Share on other sites
you could use a swtich statement,

you can make a enum variable,
like so

[CODE]
enum armour {mail, leather, cloth. plate, crystal};
[/CODE]

then use a switch statement

[CODE]
switch (armour) {
case mail: do this function
break;
case plate: do this function
break;
case leather: do this function
break;
case cloth: do this function
break;
case crystal: do this function
break;
default:
cout << "Invalid item"
}
[/CODE]

and then you just code the switch statement, with anything you like :)

Share this post


Link to post
Share on other sites
That was extremely quick :). THanks a lot. I am spending today and tomorrow working on this. I will post later what I ended up finishing if you guys are interested.

Share this post


Link to post
Share on other sites
I wanted to again thanks for the suggestions I kinda took the usable object approach with of course switches and ifs etc.

I have abstract class called ActionableObject which has methods

public virtual void Act(ContentObject actingObject) {}
public virtual void Act(ContentObject actingObject, Vector2 location) {}

I set my tileengine to actionableobject and same with all te crops, so I can send a ContentObject (WorldObject more or less) and each of the ActionableObjects take action toward said object.

Partial Main Update method
[CODE]
if (InputManager.IsActionTriggered(InputManager.Action.Ok))
{
player.State = Character.CharacterState.Acting;
if (Session.SelectedInventoryItem is Tool)
(Session.SelectedInventoryItem as Tool).ResetAnimation();
else if(Session.IsSelectedItemCrop)
player.State = Character.CharacterState.Idle;

_tileEngine.Act((ContentObject)Session.SelectedInventoryItem, PlayerPosition.SelectionMapVector);
/// item used
if (_tileEngine.GetCrop(_tileEngine.Map.Name, PlayerPosition.SelectionMapVector) != null &&
Session.IsSelectedItemCrop && !(Session.SelectedInventoryItem as Crop).IsGrown)
{
Session.Inventory.Remove(Session.SelectedInventoryItem);
Session.SelectedInventoryItem = null;
}
}
[/CODE]

Act Method for tileEngine
[CODE]
public override void Act(ContentObject actingObject, Vector2 location)
{
base.Act(actingObject, location);
Crop crop = GetCrop(Map.Name, location);
int tileIndex = GetTileIndexAtLocation(0, location);
int tileIndexValue = GetTileValueAtLocation(0, location);

/// IF THERE IS A CROP THERE
if (crop != null)
{
if (actingObject is Tool)
{
Tool tool = (actingObject as Tool);
switch (tool.Type)
{
case "WateringCan":
crop.IsWatered = true;
Map.BaseLayer[tileIndex] = (int)TILES.DARKDIRT;
/// remove the current handler because we want to have more
/// time for the watering :P
crop.OnNoLongerWatered -= new Crop.OnNoLongerWateredHandler(crop_OnNoLongerWatered);
crop.OnNoLongerWatered += new Crop.OnNoLongerWateredHandler(crop_OnNoLongerWatered);
break;
}
}
}
/// THERE IS NO CROP
else
{
if (actingObject is Tool)
{
Tool tool = (actingObject as Tool);
switch (tool.Type)
{
case "Hoe":
if (tileIndexValue == (int)TILES.LIGHTGRASS || tileIndexValue == (int)TILES.MEDIUMGRASS || tileIndexValue == (int)TILES.DARKGRASS)
Map.BaseLayer[tileIndex] = (int)TILES.LIGHTDIRT;
break;
}
}
else if (actingObject is Crop)
{
if (tileIndexValue == (int)TILES.LIGHTDIRT || tileIndexValue == (int)TILES.DARKDIRT)
{
Crop c = actingObject as Crop;
c.Coordinates = GetCollidingTileVector2(location);
AddCrop(Map.Name, c);
}
}
}
}
[/CODE]

I know that i will be changing this around a lot but thanks for getting me on my way.

Share this post


Link to post
Share on other sites
I'm in agreement with ndssia. Virtual functions would be the way to go with this. Each object should know what to do when its "interact()" method is invoked.

You might even like to pass in the object which is interacting with it so that you can change in different ways depending on the type of object. Example: The way a postal worker interacts with a mail box is different from how the owner interacts with it.

[code]
public class GameObject
{
string name;
public virtual void Interact(GameObject interactor);
}

public class Mailbox : GameObject
{
string address;
...
public override void Interact(GameObject actor)
{
switch(actor.name)
{
case "postal_worker":
{
((Postal_Worker)actor).place_mail_into(this);
break;
}
case "owner":
{
((Mailbox_Owner)actor).take_mail_from(this);
break;
}
default:
{
Console.write("You see a mailbox with the street address " + address);
break;
}
}
}
...
}
[/code]

If you model your classes by how each object can be interacted with, extending the functionality of objects becomes much easier (just add new functions). Hard coding specific object properties to other object properties will eventually turn your code into a maintenance nightmare and make it much more rigid.

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

Sign in to follow this