Data types

You will normally only need to use the int and String data types. The smaller types are only useful for conserving memory if you are creating a very large number of variables.

To declare a variable, write the type followed by the variable name, then a semicolon. For example:

int my_variable;

declares a new 32-bit integer called my_variable

WARNING: When using the float data type, you may find that the == and != operators don't seem to work properly. For example:

float result = 2.0 * 3.0;
if (result == 6.0) {
    Display("Result is 6!");
}

may not always work. This is due to the nature of floating point variables, and the solution is to code like this:

float result = 2.0 * 3.0;
if ((result > 5.99) && (result < 6.01)) {
    Display("Result is 6!");
}

The way floating point numbers are stored means that 6 might actually be stored as 6.000001 or 5.999999; this is a common gotcha to all programming languages so just be aware of it if you use any floating point arithmetic.

Arrays

data_type name [ size ];

Arrays allow you to easily create several variables of the same type. For example, suppose you wanted to store a health variable for all the different characters in the game. One way would be to declare several different variables like this:

int egoHealth;
int badGuyHealth;
int swordsmanHealth;

but that quickly gets messy and difficult to keep up to date, since you need to use different script code to update each one. So instead, you can do this at the top of your script:

int health[50];

This example declares 50 int variables, all called health.
You access each separate variable via its index (the number in the brackets). Indexes start from 0, so in this case the health array can be accessed by indexes 0 to 49. If you attempt to access an invalid index, your game will exit with an error.

Here's an example of using the array:

health[3] = 50;
health[4] = 100;
health[player.ID] = 10;

this sets Health 3 to 50, Health 4 to 100, and the Health index that corresponds to the player character's ID number to 10.
You need to do this inside a function.

NOTE: AGS currently doesn't support initialization of a range of elements in a single assignment, like many programming languages do.

Oftentimes you may want to do certain action over each element of array, or particular range of elements, such as each element from 5th to 10th, for example. Because array's index may be taken from a variable too (like health[index]), the standard way of doing so is using loops:

int i;
while (i < 50) {
    health[i] = 0;
    i++;
}

or

for (int i = 0; i < 50; i++) {
    health[i] = 0;
}

Both examples reset every element of "health" array to a zero.

See also: Dynamic arrays, Structs, while, for

Multidimensional arrays

AGS Script currently DOESN'T natively support multi-dimensional arrays (2D and higher).

There's, however, a classic method of emulating 2D array using a 1D array, if you create 1D array of size equal to "ARR_WIDTH * ARR_HEIGHT". You may then access its elements using following formula:

value = array[row * ARR_WIDTH + col];

where "row" and "col" are indexes of row and column respectively. You may think of these as Y and X coordinates, if you prefer.

This method works with both regular arrays and dynamic arrays.

For example:

#define MAP_WIDTH   50
#define MAP_HEIGHT  20
#define MAP_SIZE    1000 // this is MAP_WIDTH * MAP_HEIGHT

int map[MAP_SIZE];

// Fill whole map with random values from min_value to max_value (inclusive)
void InitRandomMap(int min_value, int max_value) {
    // here we do not care about rows and columns, so iterate this array as a "linear" (1D) array
    for (int i = 0; i < MAP_SIZE; i++) {
        map[i] = Random(max_value - min_value) + min_value;
    }
}

// Fill particular row with identical values
void SetMapRow(int row, int value) {
    for (int col = 0; col < ARR_WIDTH; col++) {
        map[row * ARR_WIDTH + col] = value;
    }
}

// Fill particular column with identical values
void SetMapColumn(int col, int value) {
    for (int row = 0; row < ARR_HEIGHT; row++) {
        map[row * ARR_WIDTH + col] = value;
    }
}

// Get a map value at particular row and column
int GetMapValue(int col, int row)  {
    return map[row * ARR_WIDTH + col];
}

Operators

The AGS scripting engine supports the following operators in expressions. They are listed in order of precedence, with the most tightly bound at the top of the list.

This order of precedence allows expressions such as the following to evaluate as expected:

if (!a && b < 4)

which will execute the 'if' block if a is 0 and b is less than 4.

However, it is always good practice to use parenthesis to group expressions. It's much more readable to script the above expression like this:

if ((!a) && (b < 4))

NOTE: By default, when using operators of equal precedence AGS evaluates them left-to-right. For example, the expression a = 5 - 4 - 2; evaluates as a = (5 - 4) - 2;, which is what you'd commonly expect from a scripting language today.
For historical reasons AGS also supports right-to-left precedence mode which may be only useful if you import very old game and do not want to fix its scripts. To enable this mode find "Left-to-right operator precedence" option in "Backwards compatibility" section of the General Settings and set it to "false".

Version checking

If you are writing a script module, you may need to check which version of AGS the user of your module is using.

For this purpose there are two directives:

#ifver 2.72
// do stuff for 2.72 and above
#endif
#ifnver 2.72
// do stuff for 2.71 and below
#endif

Note that this ability was only added in 2.72, so you cannot use the #ifver checks if you want your module to work with earlier versions than this.

NOTE: We recommend to not rely on editor version in most cases, and use Script API version macros instead whenever possible.

See also: Preprocessor

if, else statements

if ( expression ) {
statements1
}
[ else {
statements2
} ]

If expression is true, then statements1 are run.

If expression is not true, and there is an else clause present, then statements2 are run instead.

For example:

if (player.Room == 10) {
    Display("Player is in the room 10.");
}
else {
    Display("Player is NOT in the room 10.");
}

In this example, the first message will be displayed if the return value from player.Room is 10, and the second message will be displayed if it is not.

if statements can be nested inside else statements to produce an "else if" effect. For example:

if (player.Room == 1) {
    Display("Player is in the room 1.");
}
else if (player.Room == 2) {
    Display("Player is in the room 2.");
}
else {
    Display("Player in neither in room 1 nor room 2.");
}

switch, case statements

switch ( control_expression ) {
[ case match_expression:
statements
[ break; ] ]
[ default:
statements
[ break; ] ]
}

Compares the result of control_expression against the result of match_expression for each case label in order. If a match is found, statements following that label are executed. If there is no matching label and a default: label is present, statements following the default: label are executed.

If a break statement is encountered, any statements following it are skipped and execution continues after the switch block.

Unlike many programming languages, AGS allows expression results of any type (integer, boolean, string, pointers). It also does not require that match_expressions be constant or literal values.

A switch statement is useful if you need to compare one value or expression against a series of values. The control_expression represents the value you want to compare and each case label is one value in a series to compare it against.

Example:

switch (player)
{
    case cEgo:
        Display("Hello, my name is Ego.");
        break;
    case cJohn:
        Display("Greetings, I am John.");
        break;
    case cMary:
        Display("Hi there, I am Mary.");
        break;
    default:
        Display("This might be a bug!");
        break;
}

In the above example, if the player is cEgo, the game will display "Hello, my name is Ego." If the player is cJohn, the game will display "Greetings, I am John." If the player is cMary, the game will display "Hi there, I am Mary." If the player is none of these characters, the message "This might be a bug!" will be displayed.

One of the features of a switch statement is fall-through. Labels are ignored once a match is found and indeed execution will continue until the end of the switch block or a break statement is encountered.

A switch statement that demonstrates this:

switch (player)
{
    case cJohn:
    case cMary:
        player.Say("I like oranges.")
        break;
    case cEgo:
        player.Say("I like apples.");
    default:
        player.Say("I would like some berries.");
}

In the above example, if the player is either cJohn or cMary, s/he will say "I like oranges.". If the player is cEgo, he will say "I like apples." and then also "I would like some berries." If the player is any other character, only the default "I would like some berries." will be displayed.

A match_expression can be any valid AGS expression, including a function call. The following construction can be useful when implementing responses to parser values:

switch (true)
{
    case Parser.Said("take ball"):
        player.AddInventory(iBall);
        break;
    case Parser.Said("drop ball"):
        player.LoseInventory(iBall);
        break;
}

In this situation, the match_expressions are the results of Parser.Said(). If Player.Said("take ball") is true, the ball is added to the player's inventory. If Player.Said("drop ball") is true, the ball is removed from the player's inventory.

while

while ( condition ) {
statements
}

Runs statements continuously, while condition is true.

For example:

while (cEgo.Moving) {
    Wait(1);
}

will run the script Wait(1); repeatedly, as long as cEgo.Moving is not zero. Once it is zero, the while statement will exit at the end of the loop.

do..while

do {
statements
} while ( condition );

Similarly to while runs statements continuously, so long as condition is true, but unlike while it checks the condition AFTER executing statements, not before. This also means that the statements will be executed at least once.

For example:

do
{
    cEgo.Move(cEgo.x + 1, cEgo.y);
}
while (IsKeyPressed(eKeyRightArrow));

will run the script cEgo.Move(cEgo.x + 1, cEgo.y); once, and then continue run it repeatedly, as long as the right arrow key is pressed by player.

for

for ( [initialization]; [condition]; [iteration] ) {
statements
}

This loop command first performs initialization statements, then runs statements inside curved brackets continuously. Each time before executing these statements it checks whether condition is true, and if not - ends the loop. Each time after statements were executed it additionally runs iteration statements.

Initialization is commonly used to declare variables or setting up existing variable values. If a new variable is declared in initialization - such variable will exist and may be used only inside the loop. Iteration step is usually meant to "move" to the next step, by changing some variable value. Every part of the command header - initialization, condition and iteration - is optional: there may be for command without initialization, or without iteration, or even without conditional expression (in which case loop should be ended with either break or return.

For example:

for (int i = 0; i < Game.CharacterCount; i++)
{
    Display("My name is %s", character[i].Name);
}

will look over every character in game and display their names.

Another example (note missing initialization and iteration):

for (; cEgo.x < 100;)
{
    Wait(1);
}

This will repeat Wait(1); until cEgo character does not move beyond coordinate x = 100.

break

break;

break statement ends the execution of most inner loop or switch immediately. After this script continues running from the next line after loop or switch.

For example:

while (cEgo.Moving) {
    if (IsKeyPressed(eKeyEscape))
        break;

    Wait(1);
}

will run the script Wait(1); repeatedly, as long as cEgo.Moving is not zero. If player presses Escape key, the loop is terminated immediately.

continue

continue;

continue statement makes the loop skip remaining statements in current iteration and proceed to the next end-condition check, followed by the loop restart, if condition is still met, or loop end. If in for kind of loop, the iteration statement is executed right before that.

For example:

for (int i = 0; i < 100; i++)
{
    // multiple statements here

    if (i > 50)
        continue;

    // more statements following
}

will run first part of the loop statements always, and second part only when i <= 50.

function

function name ( [type1 param1, type2 param2, ... ] )

Declares a custom function in your script. A function is a way in which you can separate out commonly used code into its own place, and thus avoid duplicating code.

For example, suppose that you quite often want to play a sound and add an inventory item at the same time. You could write both commands each time, or you could define a custom function:

function AddInvAndPlaySound(InventoryItem* item) {
    player.AddInventory(item);
    aInventorySound.Play();
}

then, elsewhere in your code you can simply call:

AddInvAndPlaySound(iKey);

to add inventory item iKey and play the sound.

Generally, you place your functions in your global script. You then need to add an import line to your script header to allow the function to be called from room scripts.

Optional parameters

You can make int parameters optional if there is a default value that the user doesn't need to supply. To do this, change the script header import declaration like this:

import function TestFunction(int stuff, int things = 5);

that declares a function with a mandatory stuff parameter, and an optional things parameter. If the caller does not supply the second parameter, it will default to 5.

NOTE: To use optional parameters, you need to have an "import" declaration for the function in the script header. The default values cannot be specified in the actual function declaration itself.

return

return;

Immediately quits currently run function and returns to the previous script function current one was called from, if there was any, otherwise passes execution to engine. return can be put in any place in the the function, no matter if it is inside the if/else statement group, loop or switch - it will still work as immediate function exit.

If the function is declared with return type other than void (or simply like function), then the return statement has to specify return value.

int GetHowManyTradeGoodsShopkeeperHas() {
    return 2;
}

Alternatively, when function is not supposed to have any return value, sometimes you may want to break out of current function before it ends naturally:

function DoThisAndOptionallyThat(bool do_all) {
    // multiple statements here

    if (!do_all)
        return; // quit the function prematurely

    // more statements following
}

struct

struct name {

Declares a custom struct type in your script.
Structs allow you to group together related variables in order to make your script more structured and readable. For example, suppose that wanted to store some information on weapons that the player could carry. You could declare the variables like this:

int swordDamage;
int swordPrice;
String swordName;

but that quickly gets out of hand and leaves you with tons of variables to keep track of. This is where structs come in:

struct Weapon {
    int damage;
    int price;
    String name;
};

Now, you can declare a struct in one go, like so:

Weapon sword;
sword.damage = 10;
sword.price = 50;
sword.name = "Fine sword";

Much neater and better organized. You can also combine structs with arrays:

// at top of script
Weapon weapons[10];
// inside script function
weapons[0].damage = 10;
weapons[0].price = 50;
weapons[0].name = "Fine sword";
weapons[1].damage = 20;
weapons[1].price = 80;
weapons[1].name = "Poison dagger";

structs are essential if you have complex data that you need to store in your scripts.

managed

managed struct name {

Managed is a modifier that can be applied to struct declaration to make them managed structs.

Managed structs are special in the way that objects of their types are created in dynamic pool as opposed to global variables, that exist from the game start to when the game is shut down, and local variables, that exist only when their function is run. You cannot declare a variable of managed struct, but only a pointer variable.

The advantage of such managed (or dynamic) objects is that they are created only when needed and disposed of when no longer needed. Also, since you work with pointer to object instead of object itself, you may assign them to another variable without copying object itself, pass them to function as parameter, or return from the function.

IMPORTANT: there is a big limitation for user-defined managed structs now, it is that they themselves cannot have members of pointer types (or dynamic arrays).

Example:

managed struct Apple {
    int color;
    int freshness;
};

This declares managed struct. To declare a pointer to such struct you do:

Apple* my_apple;

This creates a pointer variable my_apple of managed type Apple.

However, this does not create an object itself yet, and my_apple is assigned null value now. If you try to access struct members using my_apple now, you will get errors. To create an actual object you need to use a new keyword:

my_apple = new Apple;

The object is now created in the dynamic memory pool, and variable my_apple points to it. This lets us access object contents:

my_apple.color = Game.GetColorFromRGB(255, 0, 0);
my_apple.freshness = 100;

You may copy pointer to another variable of same type:

Apple* apple2 = my_apple;

This does not copy object itself, only its address in dynamic pool, meaning both variables - my_apple and apple2 - point to same object!

You may write a function that take such pointer as parameter:

function DisplayAppleDescription(Apple* apple) {
    String s = String.Format("Apple has color %d and freshness %d", apple.color, apple.freshness);
    Display(s);
}

and then call it like:

DisplayAppleDescription(my_apple);

You may write a function that returns pointer to apple:

Apple* CreateYellowApple(int fresh) {
    Apple* apple = new Apple;
    apple.color = Game.GetColorFromRGB(255, 0, 255);
    apple.freshness = fresh;
    return apple;
}

and then use such function just like:

Apple *my_apple = CreateYellowApple(50);

When does the dynamic object gets destroyed? After you created dynamic object as described above, it will exist in memory as long as there is at least one pointer variable pointing to it. As soon as the last pointer gets destroyed itself (for example, if it was local function variable, and function ended), or is assigned another object, or simply assigned null, then the dynamic object is removed from your game forever.

See also: new, Pointers in AGS

new

pointer_variable = new managed_type;

Creates a new dynamic (managed) object of managed_type and assigns it to pointer_variable.

Example:

// Here we declare a managed struct for Apple
managed struct Apple {
    int color;
    int freshness;
};

// ...and declare a global pointer to Apple
Apple* SomeApple;

// At the game start we create a new dynamic object of Apple type
// and assign its address to the pointer variable
function game_start()
{
    SomeApple = new Apple;
}

See also: managed, Pointers in AGS

enum

enum name {
option1 [ = value1 ],
option2 [ = value2 ],
...
};

Declares an enumeration type. An enumeration allows you to group together a set of related options, where only one will be true at any one time -- a bit like the contents of a list box.

For example, if you have a script function, doStuff, that can perform 3 different operations, you could do this:

function doStuff(int param) {
    if (param == 1) {
        // do something
    }
    else if (param == 2) {
        // do something else
    }
    // etc
}

but it's hard to read, and when calling the function from elsewhere in your script, it's not clear what 1 or 2 means. That's where enums come in:

enum DoStuffOption {
    BakeCake,
    DoLaundry
};

function doStuff(DoStuffOption param) {
    if (param == BakeCake) {
        // do something
    }
    else if (param == DoLaundry) {
        // do something else
    }
    // etc
}

and then the calling code looks like:
doStuff(BakeCake);
thus making it perfectly clear what the command will do.

Normally, you would put the enum definition into the script header.

If you don't assign a value for the first enum item, it is assigned the value 1. Other enum items have the value of the enum before plus 1. Ex: enum ABC { eA,eB, eC}; get's eA=1 eB=eA+1=2 eC=eB+1=3, by default.

In summary, enums are not an essential part of scripting and you can get away perfectly well without using them, but in some specific situations they're very handy.

this

There are two uses for the this keyword.

1. Accessing members of the current struct

When you are creating custom structs, you use the "this" keyword inside member functions to refer to the current struct. For example:

Suppose you had this in your script header:

struct MyStruct {
    int myValue;

    import function MyMethod();
};

Then, in your main script, you could put this:

function MyStruct::MyMethod()
{
    this.myValue = 5;
}

The MyStruct::MyMethod tells AGS that you are defining the function MyMethod which belongs to the struct MyStruct (the :: operator means "belongs to").

The code above will mean that when the MyMethod function is called, it sets the myValue variable to 5.

2. Declaring extender functions

Please see the Extender functions page for details.

import

import [attribute] declaration ;

Declares declaration as a variable or function which is external to the current script, but that the script needs access to it. You use this to provide your room scripts with access to parts of your global script.

For example:

import int counter;
import function add_numbers (int, int);

This imports an integer variable counter and the function add_numbers from the global script to enable the current script to call them. You normally place import statements into the script header so that all rooms can benefit from them.

In order to import the variable, it must have been exported from the global script with the export keyword.

NOTE: You MUST import external variables with the correct type. If counter was declared as a short in the global script, you MUST import it as a short, otherwise your game may crash.

NOTE: You cannot import old-style string variables (this does not apply to new-style String variables).

As a special case, specifying the optional keyword attribute within a struct definition will define a virtual property, which will also require matching getter and setter functions.

For example:

struct Weapon
{
    protected int damage;
    import attribute int Damage;
};

Please see the Object Oriented programming page for more details.

export

export variable [, variable ... ] ;

Declares that variable is exported and is available to access in other scripts, if declared using the import keyword in those scripts.

For example:

export my_variable;
export counter, strength;

This exports three variables (my_variable, counter, and strength) to be available for import.

readonly

readonly data_type variable [ = value ];

The readonly keyword is used when declaring a variable, to indicate that its value cannot be changed.

For example:

readonly int zero;
readonly int months = 12;

writeprotected

writeprotected data_type struct_member;

The writeprotected keyword is used to define a property within a struct, which can only be modified by struct members (including extender functions) using the this keyword. Reading the value is not restricted.

For example:

struct Weapon
{
    writeprotected int Damage;
    import bool SetDamage(int damage);
};

protected

protected data_type struct_member;

The protected keyword is used to define a property within a struct, which can only be accessed or modified by struct members (including extender functions) using the this keyword.

For example:

struct Weapon
{
    protected int Damage;
    import bool SetDamage(int damage);
};

noloopcheck

function noloopcheck function_name ( parameters ... ) {

The noloopcheck keyword disables the script loop checking for the current function.

Normally, if a while loop runs for more than 150,000 loops, AGS will assume that the script has hung and abort the game. This is to assist scripting since otherwise the game would lock up if you scripted a loop wrongly.

However, there are some rare situations in which you need a loop to run several thousand times (for example, when initializing a very large array). In this case, the noloopcheck keyword can be used to stop AGS aborting your script.

NOTE: The noloopcheck keyword must be placed between "function" and the function's name.
If you import the function into the script header, you do not include the "noloopcheck" keyword in the import declaration -- it is only included in the actual function body.

NOTE: If AGS gives you a script iterations error, DO NOT just automatically add this keyword as a way to fix the problem -- more often than not, it is a fault in your scripting and using this keyword will mean that the game will hang rather than abort.

For example:

function noloopcheck initialize_array() {
    char bigarray[200000];
    int a = 0;
    while (a < 200000) {
        bigarray[a] = 1;
        a++;
    }
}

without the "noloopcheck" keyword here, AGS would abort that script.