Calling macros

ADVANCED
THIS IS AN ADVANCED ARTICLE

Calling Macros From another Macro

Sometimes you want to call a macro from another macro. As a coder, you want this often, because it enables you to split your code in small, simple chunks that can be reused over and over again. This handy technique also helps keep the stack size requirement low and mitigates the need for large code-level-nesting.

When calling a macro from another macro, you will often want to transfer data from one to the other and vice versa. In addition, it is also important to know what happens with chat output.

There are four different ways to call a macro - and they all behave a little bit differently. While reading this tutorial, please be sure to follow the wiki-links provided for any items you are not familiar with, as they will not be covered.

The Four Methods to Call a Macro

The Macro Roll Option

This is the most straight forward way to call a macro. When using the roll option [macro():], you must specify the location of the macro being called and give a single argument. Note that this argument must always be specified, even if you do not need it. In such cases, it is common practice is to use an empty string "".

This argument can be accessed inside the called macro via the special variable called macro.args.

All of the called macro's output will be inserted into the calling macro at the point in the code where the [macro():] roll option is placed. This output usually goes into the chat output. However, you can choose to instead assign the returned value to the variable macro.return.

Once the called macro is processed, the macro.return variable can be used in the calling macro in a function call.

Example

Lets say you have a macro called attackRoll that can roll any number of dice. In another macro, you want to call attackRoll, have it roll three dice and then use the result of that roll.

Calling macro:

<source lang="mtmacro" line> [h, macro("attackRoll@Lib:Token"): 3] Attack roll: [r: macro.return] </syntaxhighlight>

Called macro (named attackRoll and placed on token Lib:Token):

<source lang="mtmacro" line> [h: diceNr = macro.args] [r: roll(diceNr, 6)] </syntaxhighlight>

Note that you could write the called macro this way, too: <source lang="mtmacro" line> [h: diceNr = macro.args] [h: macro.return = roll(diceNr, 6)] </syntaxhighlight> For this simple example, it really makes no difference.

User Defined Function

Often referred to as UDF in the forum, User Defined Functions are probably the most convenient way to do complex coding. A User Defined Function can be used just like a regular function. It can have arguments and it will be replaced by its resulting value when MapTool parses the calling macro. User defined means that you can specify the macro to be called when you create the function.

Arguments are assigned to the UDF by writing them inside the parentheses (), separated by comas. The macro's complete chat output will be used as the resulting value and replace the function call, so you can easily assign it to a variableor use it as an argument for another function. It is important to note that HTML comments will be included in this output as well, even though they will not appear in the chat window. Also note that only the first macro in an execution chain will dump its output to chat.

In the called macro, the arguments are inside the macro.args variable, formatted as a JSON array. You can use argCount() and arg() for easy access.

To set up a UDF you have to call defineFunction() on every client where the function will be used. This can be done automatically by placing all the defineFunction() calls inside the special macro onCampaignLoad on a library token. This is the standard practice. It is executed whenever the campaign file is loaded by MT (whether from a server or from file).

Example

Lets rewrite the above example using a UDF.

onCampaignLoad on a token Lib:Token: <source lang="mtmacro" line> [defineFunction("attackRoll", "attackRoll@Lib:Token")]

</syntaxhighlight>

Calling macro: <source lang="mtmacro" line> Attack roll: [r: attackRoll(3)] </syntaxhighlight>

Called macro (named attackRoll and placed on token Lib:Token): <source lang="mtmacro" line> [h: assert(argCount()>0, "attackRoll() expects one argument.")] [h: diceNr = arg(0)] [r: roll(diceNr, 6)] </syntaxhighlight>

Create UDFs Automatically

You can write a macro that scans your Lib:token macros and converts them all into user defined functions. This is a nice, convenient little trick that's done here (see forum post), in really elaborated way, by aliasmask.

Macro Links

When you want to call macros on user reaction, you can send out clickable links to chat or place them into frames. Also use them if you want to work with HTML forms or the fancier form-based events.

Since the macro is not executed immediately, there is a way to use the macro's result in the calling macro. Arguments, the token in context and where the output should be sent can all be specified precisely when you create the macro link.

See also macroLink(), macroLinkText().

Example

This time, let us assume we want to send an attack roll to chat and then ask for a defense roll. We also want to send the Macro Link to everybody connected (because that is much easier) and don't care about the current token in context.

Calling macro: <source lang="mtmacro" line> [h: atk = roll(3,6)] Attack roll: [r: atk]
[r: macroLink("Do you want to defend?", "defenceRoll@Lib:Token", "all", atk)] </syntaxhighlight>

Called macro (named defenceRoll and placed on token Lib:Token): Note that this macro will be executed whenever the link is clicked. <source lang="mtmacro" line> [h: atk = macro.args] [h: def = roll(3,6)]

Defence roll: [r: def] [r, if(atk<def): "You defended successfully!"; "You are hit."] </syntaxhighlight>

Evaluate a Macro

This does not directly call a macro stored somewhere, but rather evaluates some string you feed into the function as if it were macro code. This happens in place. It is not easy to retrieve the macro code from a stored macro, thus this is not a good way to call a macro stored in the usual way. This is most often used for small code snippets created dynamically or stored on token properties.

See also evalMacro(), execMacro(), json.evaluate(), eval(). Here, evalMacro and execMacro do exactly the same thing; they are just two different names with the same functionality.

Example

Lets say a RPG has a complex weapon damage system with formulas that follow no rule. The formula for the active would be stored in a token property called damageFormula. <source lang="mtmacro" line> [h: myFormula = getProperty("damageFormula")] You made [r: evalMacro(myFormula)] damage. </syntaxhighlight>

Content of the damageFormula property: <source lang="mtmacro" line> [h: ""] [h: firstRoll = 1d3] [r, if(firstRoll==3): 2d10; 3d6] </syntaxhighlight> Yeah, such a damage system would be horrible.

One notable difference between eval and evalMacro is how you pass a parameter:
For example -- the call ... <source lang="mtmacro" line> [r:myFunction()] </syntaxhighlight>

... can also be called as follows with the two methods: <source lang="mtmacro" line> [r:evalMacro("[r:myFunction()]"] </syntaxhighlight>

Which has the same result as: <source lang="mtmacro" line> [r:eval("myFunction()"] </syntaxhighlight>

Do you see how evalMacro() works with square brackets (roll options) and eval() just the function text?

Variable Context (Scope)

Usually, a new macro creates a new context (scope) for variables, thus locally defined variables in one macro are not defined in another.

By using defineFunction(), you can call macros that operate in the same variable context (scope) as the calling macro -- when you want so.

The token context is usually transported along with the macro call. With macro links, you can specify the token context.]