Compilerobjectdynfunc.t[40]

Compiler: This object provides a simplified interface to the dynamic compiler. The methods here can be used instead of manually creating DynamicFunc instances.

The main advantage of using this object to compile code is that it automatically provides access to the global symbol table that was used to compile the current program, for use in dynamic code. Without the global symbol table, dynamic code won't have access to object names, property names, function names, and so on. That doesn't stop you from compiling code that only depends upon its own function parameters and local variables, but for most purposes the global symbols are useful to have around.

Note that including this object in a project will automatically save the global symbol table in the compiled .t3 file. This increases the size of the .t3 file, as well as memory usage during execution. If you're concerned about minimizing the .t3 file size or the run-time memory footprint, *and* you don't need global symbols for dynamic code (or you don't use the dynamic compiler at all), you can save some space by omitting this whole module from the build.

Compiler :   PreinitObject

Superclass Tree   (in declaration order)

Compiler
        PreinitObject
                ModuleExecObject
                        object

Summary of Properties  

macros_  symtab_ 

Inherited from PreinitObject :
execBeforeMe  reverseGlobalSymbols 

Inherited from ModuleExecObject :
execAfterMe  hasInitialized_  isDoingExec_  isExecuted_ 

Summary of Methods  

compile  defineFunc  eval  execute 

Inherited from ModuleExecObject :
_execute  classExec 

Properties  

macros_dynfunc.t[149]

a saved referenced to the preprocessor macro table

symtab_dynfunc.t[146]
a saved reference to the global symbol table

Methods  

compile (str, locals?)dynfunc.t[73]

Compile an expression or function. 'str' is a string giving the code to compile. This can be a simple value expression, such as 'Me.location' or 'new BigNumber(12345).sqrt()'. Or, it can be a complete unnamed function definition, using this syntax:


'function(x, y, z) { ...body of function... }'

The body of the function can contain any executable code that you could write in a regular function in static code: if, while, switch, return, etc.

The return value is a DynamicFunc containing the compiled expression or function. You call it by using the return value as though it were a function:


local f = Compiler.compile('Me.location');
local loc = f();

If the source string was just an expression, it acts like a function that takes zero arguments, and returns the computed value of the expression. The expression is evaluated anew each time you invoke it, so you'll get the "live" value of an expression that refers to object properties or other external data. In the example above, we'd get the current value of Me.location every time we call f().

The source string is actually compiled immediately when you call this function. This means it's checked for errors, such as syntax errors and unknown symbol names. If the code contains any errors, this method throws a CompilerException describing the problem.

defineFunc (name, str, locals?)dynfunc.t[85]
Compile a dynamic function string, and add it to the global symbol table as a function with the given name. This effectively creates a new named function that you can call from other dynamic code objects.

eval (str, locals?)dynfunc.t[118]
Evaluate an expression. 'str' is a string giving code to compile. In most cases, this is simply a simple value expression, although it's also acceptable to use the 'function()' syntax to create a function that takes no arguments.

This method compiles the source string and immediately calls the resulting compiled code. The return value is the value returned from the compiled code itself. This method thus provides a quick way to evaluate an expression.

If the string contains any syntax errors or other compilation errors, the method throws a CompilerException. In addition, it's possible for the compiled code to throw exceptions of its own; this method doesn't catch those, leaving it up to the caller to handle them.

If you expect to evaluate the same expression repeatedly, you're better off using compile() to get the compiled representation of the expression, and then call that compiled code each time the value is needed. That's more efficient than using eval() each time, since eval() to recompile the expression on every call, which is a fairly complex process.

execute ( )OVERRIDDENdynfunc.t[138]
During preinit, save a reference to the program's global symbol table in a property of self. The VM always makes the global symbols available during preinit, but by default it discards the table after that because most programs don't need it. That means that the symbols aren't available by default during normal execution. However, saving a reference here prevents the garbage collector from discarding the table when preinit finishes, which forces it to be saved in the final .t3 file and thus makes it available permanently.

Adv3Lite Library Reference Manual
Generated on 03/07/2024 from adv3Lite version 2.1