Kuroko - A bytecode-compiled scripting language

Kuroko is a bytecode-interpreted, dynamic, strongly-typed language with syntax similar to Python.

The bytecode VM / compiler is substantially based on Robert Nystrom's Crafting Interpreters.

At the moment, the intent for this project is to add a proper scripting language to Bim, to which both configuration scripts and syntax highlighting will be ported.

Kuroko, as its name should imply, will also be made available in ToaruOS as a general-purpose user language, and some utilities may end up being written in it.

Features

Kuroko inherits some core features by virtue of following Crafting Interpreters, including its basic type system, classes/methods/functions, and the design of its compiler and bytecode VM.

On top of this, Kuroko adds a number of features inspired by Python, such as:

• Indentation-based block syntax.
• Collection types: list, dict, tuple, with compiler literal syntax ([],{},(,)).
• Iterable types, with for ... in ... syntax.
• List comprehensions ([foo(x) for x in [1,2,3,4]] and similar expressions).
• Pseudo-classes for basic values (eg. everything has a __repr__, strings have format(), etc.)
• Exceptions with, with try/except/raise.
• Modules, both for native C code and managed Kuroko code.

Building Kuroko

Building as a Shared Library

Kuroko has no external dependencies beyond the system C library and support for dlopen for loading C modules.

Generally, make should suffice to build from the repository.

The compiler/VM is built as a shared object from these source files:

builtins.c  chunk.c  compiler.c  debug.c  memory.c  object.c  scanner.c  table.c  value.c  vm.c

The interpreter binary is a thin wrapper and lives in kuroko.c; rline.c provides the syntax-highlighted line editor for the REPL.

C module sources are found in src/ and provide optional added functionality. Each module source file corresponds to a resulting shared object of the same name that will be built to the modules/ directory, which itself also contains modules written in Kuroko.

The core builtins, builtins.krk are embedded in builtins.c so they are always available to the interpreter; builtins.c is provided in the repository, but can be updated by the Makefile when changes to builtins.krk are made.

Building as a Single Static Binary

An additional Makefile is provided to build a single static binary of the interpter and REPL with the additional C modules bundled as built-ins:

make clean; make -f Makefile.static

This will produce a static binary without dlopen support, so it will not be able to load additional C modules at runtime.

Code Examples

NOTE: Due to limitations with Github's markdown renderer, these snippets will be highlighted as Python code.

Hello World

print("Hello, world!")
# → Hello, world!

Multiple expressions can be supplied to print and will be concatenated with spaces:

print("Hello", 42, "!")
# → Hello 42 !

The string used to combine arguments can be changed with sep=.

print("Hello", 42, "!", sep="...")
# → Hello...42...!

Basic Types

Kuroko's basic types are integers (which use the platform long type), double-precision floats, booleans (True and False), and None.

print(1 + 2 + 3)
# → 6

When integer values are used in arithmetic operations, such as division, the result will be an integer as well:

print(1 / 2)
# → 0

To get floating-point results, one of the arguments should be explicitly typed or converted:

print(1 / 2.0)
# → 0.5

Implicit type conversion occurs late in evaluation, so be careful of integer overflows:

# Probably not what you want:
print(1000000000 * 1000000000 * 1000000000 * 3.0)
# → -2.07927e+19
# Try something like this instead:
print(1000000000.0 * 1000000000 * 1000000000 * 3.0)
# → 3e+27

Objects

Objects are values which live on the heap. Basic objects include strings, functions, classes, and instances.

Objects are passed by reference, though strings are immutable so this property is only relevant for other object types.

Strings

Strings can be concatenated, and other values can be appended to them.

print("Hello, " + 42 + "!")
# → Hello, 42!

Functions

Function syntax is essentially the same as in Python:

def greet(name):
    print("Hello, " + name + "!")
greet("user")
# → Hello, user!

Default arguments can be specified as follows:

def greet(name=None):
    if not name:
        print("Hello, world!")
    else:
        print("Hello, " + name + "!")
greet()
gree("user")
# → Hello, world!
#   Hello, user!

If a default argument value is not provided, the expression assigned to it will be evaluated as if it were at the top of the body of the function, like in Ruby (and not like in Python).

Blocks, including function def blocks and control flow structures like if and for, must be indented with spaces to a level greater than the enclosing block.

You may indent blocks to whatever level you desire, so long as ordering remains consistent, though the recommendation indentation size is 4 spaces.

It is recommended that you use an editor which provides a clear visual distinction between tabs and spaces, such as Bim.

Blocks can also accept a single inline statement:

if True: print("The first rule of Tautology Club is the first rule of Tautology Club.")
# → The first rule of Tautology Club is the first rule of Tautology Club.

Variables

In a departure from Python, Kuroko has explicit variable declaration and traditional scoping rules. Variables are declared with the let keyword and take the value None if not defined at declaration time:

let foo
print(foo)
# → None
foo = 1
print(foo)
# → 1

Closures

Functions are first-class values and may be returned from functions and stored in variables, producing closures.

When a function references local values from an outer scope, such as in the example below, the referenced variables will be captured.

def foo():
    let i = 1 # Local to this call to foo()
    def bar():
        print(i) # Reference to outer variable
        i = i + 1
    return bar # Produces a closure
let a = foo() # Each copy of `bar` gets its own `i`
let b = foo()
let c = foo()
a() # So these all print "1" as the first call,
b() # but each one also increments its own copy of i
c()
a() # So further calls will reference that copy
a()
a()
# → 1
#   1
#   1
#   2
#   3
#   4

Basic Objects and Classes

Objects and classes in Kuroko work a lot like Python or similar languages in that they have an arbitrary and mutable set of fields, which may be methods or other values.

To create a basic object without methods, the object class is provided:

let o = object()
o.foo = "bar"
print(o.foo)
# → bar

To supply methods, define a class:

class Foo():
    def printFoo():
        print(self.foo)
let o = Foo()
o.foo = "bar"
o.printFoo()
# → bar

The self keyword is implicit in all methods and does not need to be supplied in the argument list. You may optionally include it in the method declaration anyway:

class Foo():
    def printFoo(self):
        print(self.foo)
let o = Foo()
o.foo = "bar"
o.printFoo()
# → bar

When a class is instantiated, if it has an __init__ method it will be called automatically. __init__ may take arguments as well.

class Foo():
    def __init__(bar):
        self.foo = bar
    def printFoo(self):
        print(self.foo)
let o = Foo("bar")
o.printFoo()
# → bar

Some other special method names include __get__, __set__, and __str__, which will be explained later.

Note: Unlike in Python, most types are not actually instances of classes, though many of the same operations still apply to them.

Inheritance

Classes may inherit from a single super class:

class Foo():
    def __init__():
        self.type = "foo"
    def printType():
        print(self.type)

class Bar(Foo):
    def __init__():
        self.type = "bar"

let bar = Bar()
bar.printType()
# → bar

Methods can refer to the super class with the super keyword:

class Foo():
    def __init__():
        self.type = "foo"
    def printType():
        print(self.type)

class Bar(Foo):
    def __init__():
        self.type = "bar"
    def printType():
        super().printType()
        print("Also, I enjoy long walks on the beach.")

let bar = Bar()
bar.printType()
# → bar
#   Also, I enjoy long walks on the beach.

You can determine at runtime if an object is an instance of a class, either directly or through its inheritance chain, with the isinstance builtin function:

class Foo:
class Bar:
class Baz(Bar):
let b = Baz()
print(isinstance(b,Baz), isinstance(b,Bar), isinstance(b,Foo), isinstance(b,object))
# → True, True, False, True

All classes eventually inherit from the base class object, which provides default implementations of some special instance methods.

Collections

Kuroko has built-in classes for flexible arrays (list) and hashmaps (dict):

let l = list()
l.append(1)
l.append(2)
l.append("three")
l.append(False)
print(l)
# → [1, 2, three, False]
l[1] = 5
print(l)
# → [1, 5, three, False]
let d = dict()
d["foo"] = "bar"
d[1] = 2
print(d)
# → {1: 2, foo: bar}

These built-in collections can also be initialized as expressions, which act as syntactic sugar for the listOf and dictOf built-in functions:

let l = [1,2,"three",False] # or listOf(1,2,"three",False)
print(l)
# → [1, 2, three, False]
let d = {"foo": "bar", 1: 2} # or dictOf("foo","bar",1,2)
print(d)
# → {1: 2, foo: bar}

Lists can also be generated dynamically:

let fives = [x * 5 for x in [1,2,3,4,5]]
print(fives)
# → [5, 10, 15, 20, 25]

Exceptions

Kuroko provides a mechanism for handling errors at runtime. If an error is not caught, the interpreter will end and print a traceback.

def foo(bar):
    print("I expect an argument! " + bar)
foo() # I didn't provide one!
# → Traceback, most recent first, 1 call frame:
#     File "<stdin>", line 1, in <module>
#   ArgumentError: foo() takes exactly 1 argument (0 given)

To catch exceptions, use try/except:

def foo(bar):
    print("I expect an argument! " + bar)
try:
    foo() # I didn't provide one!
except:
    print("oh no!")
# → oh no!

Runtime exceptions are passed to the except block as a special variable exception. As of this writing, runtime exceptions from the VM are strings.

def foo(bar):
    print("I expect an argument! " + bar)
try:
    foo() # I didn't provide one!
except:
    print("oh no, there was an exception:", exception.arg)
# → oh no, there was an exception: foo() takes exactly 1 argument (0 given)

Exceptions can also be generated from code:

def login(password):
    if password != "supersecret":
        raise "Wrong password, try again!"
    print("[Hacker voice] I'm in.")
login("foo")
# → Traceback, most recent first, 2 call frames:
#     File "<stdin>", line 5, in <module>
#     File "<stdin>", line 3, in login
#   Wrong password, try again!

The except block is optional, and an exception may be caught and ignored.

def login(password):
    if password != "supersecret":
        raise "Wrong password, try again!"
    print("[Hacker voice] I'm in.")
try:
    login("foo")
# (no output)

Modules

Modules allow scripts to call other scripts.

# modules/demomodule.krk
foo = "bar"

# demo.krk
import demomodule
print(demomodule.foo)
# → bar

Modules are run once and then cached, so if they preform actions like printing or complex computation this will happen once when first imported. The globals table from the module is the fields table of an object. Further imports of the same module will return the same object.

Loops

Kuroku supports C-style for loops, while loops, and Python-style iterator for loops.

for i = 1, i < 5, i = i + 1:
    print(i)
# → 1
#   2
#   3
#   4

let i = 36
while i > 1:
    i = i / 2
    print(i)
# → 18
#   9
#   4
#   2
#   1

let l = [1,2,3]
for i in l:
    print(i)
# → 1
#   2
#   3

Iterators

The special method __iter__ should return an iterator. An iterator should be a function which increments an internal state and returns the next value. If there are no values remaining, return the iterator itself.

An example of an iterator is the range built-in class, which is defined like this:

class range:
    "Helpful iterable."
    def __init__(self, min, max):
        self.min = min
        self.max = max
    def __iter__(self):
        let i=self.min
        let l=self.max
        def _():
            if i>=l:
                return _
            let o=i
            i++
            return o
        return _

Objects which have an __iter__ method can then be used with the for ... in ... syntax:

for i in range(1,5):
    print(i)
# → 1
#   2
#   3
#   4

Indexing

Objects with the methods __get__ and __set__ can be used with square brackets []:

class Foo:
    def __get__(ind):
        print("You asked for ind=" + ind)
        return ind * 5
    def __set__(ind, val):
        print("You asked to set ind=" + ind + " to " + val)
let f = Foo()
print(f[4])
f[7] = "bar"
# → You asked for ind=4
#   20
#   You asked to set ind=7 to bar

String Slicing

Substrings can be extracted from strings via slicing:

print("Hello world!"[3:8])
# → lo wo
print("Hello world!"[:-1])
# → Hello world
print("Hello world!"[-1:])
# → !

NOTE: Step values are not yet supported.

String Conversion

If an object implements the __str__ method, it will be called to produce string values when concatenating or printing.

class Foo:
    def __str__():
        return "(I am a Foo!)"
let f = Foo()
print(f)
# → (I am a Foo!)

File I/O

The module fileio provides an interface for opening, reading, and writing files, including stdin/stdout/stderr.

To open and read the contents of a file:

import fileio
let f = fileio.open("README.md","r")
print(f.read())
f.close()

To write to stdout (notably, without automatic line feeds):

import fileio
fileio.stdout.write("hello, world")

To read lines from stdin:

import fileio

while True:
    fileio.stdout.write("Say something: ")
    fileio.stdout.flush()

    let data = fileio.stdin.readline()
    if data[-1] == '\n':
        data = data[:-1]
    if data == "exit":
        break
    print("You said '" + data + "'!")

Decorators

Decorators allow functions and methods to be wrapped.

def decorator(func):
    print("I take the function to be decorated as an argument:", func)
    def wrapper():
        print("And I am the wrapper.")
        func()
        print("Returned from wrapped function.")
    return wrapper

@decorator
def wrappedFunction():
    print("Hello, world")

wrappedFunction()
# → I take a function to be decorated as an argument: <function wrappedFunction>
#   And I am the wrapper.
#   Hello, world
#   Returned from wrapped function

The resulting function will have the same signature as the original function, so wrappers may take arguments to pass to the wrapped function, or may take their own arguments (or both).

Method wrappers work similarly, though be sure to explicitly provide a name (other than self) for the object instance:

def methodDecorator(method):
    def methodWrapper(instance, anExtraArgument):
        method(instance)
        print("I also required this extra argument:", anExtraArgument)
    return methodWrapper

class Foo():
    @methodDecorator
    def theMethod():
        print("I am a method, so I can obviously access", self)
        print("And I also didn't take any arguments, but my wrapper did:")

let f = Foo()
f.theMethod("the newly required argument")
# → I am a method, so I can obviously access <instance of Foo at ...>
#   And I also didn't take any arguments, but my wrapper did:
#   I also required this extra argument: the newly required argument

Decorators are expressions, just like in Python, so to make a decorator with arguments create a function that takes those arguments and returns a decorator:

def requirePassword(password):
    print("I am creating a decorator.")
    def decorator(func):
        print("I am wrapping", func, "and attaching",password)
        def wrapper(secretPassword):
            if secretPassword != password:
                print("You didn't say the magic word.")
                return
            func()
        return wrapper
    return decorator

@requirePassword("hunter2")
def superSecretFunction():
    print("Welcome!")

superSecretFunction("a wrong password")
print("Let's try again.")
superSecretFunction("hunter2")
# → I am wrapping <function superSecretFunction> and attaching hunter2
#   You didn't say the magic word.
#   Let's try again.
#   Welcome!

Keyword Arguments

Arguments may be passed to a function by specifying their name instead of using their positional location.

def aFunction(a,b,c):
    print(a,b,c)

aFunction(1,2,3)
aFunction(1,c=3,b=2)
aFunction(b=2,c=3,a=1)
# → 1 2 3
#   1 2 3
#   1 2 3

This will be slower in execution than a normal function call, as the interpreter will need to figure out where to place arguments in the requested function by examining it at runtime, but it allows for functions to take many default arguments without forcing the caller to specify the values for everything leading up to one they want to specifically set.

def aFunction(with=None,lots=None,of=None,default=None,args=None):
    print(with,lots,of,default,args)

aFunction(of="hello!")
# → None None hello! None None

*args and **kwargs

When used as parameters in a function signature, * and ** before an identifier indicate that the function will accept arbitrary additional positional arguments and keyword arguments respectively. These options are typically applied to variables named args and kwargs, and they must appear last (and in this order) in the function signature if used.

The variable marked with * will be provided as an ordered list, and ** will be an unordered dict of keyword names to values.

def takesArgs(*args):
    print(args)
takesArgs(1,2,3)
# → [1, 2, 3]
def takesKwargs(**kwargs):
    print(kwargs)
takesKwargs(a=1,b=2,c=3)
# → {'a': 1, 'b': 2, 'c': 3}
def takesEither(*args,**kwargs):
    print(args, kwargs)
takesEither(1,2,a=3,b=4)
# → [1, 2] {'a': 3, 'b': 4}
def takesARequiredAndMore(a,*args):
    print(a, args)
takesARequiredAndMore(1,2,3,4)
# → 1 [2, 3, 4]

Argument Expansion

When used in a function argument list, * and ** before a list and dict expression respectively, will expand those values into the argument list.

let l = [1,2,3]
def foo(a,b,c):
    print(a,b,c)
foo(*l)
# → 1 2 3
let d = {"foo": "a", "bar": 1}
def func(foo,bar):
    print(foo, bar)
func(**d)
# → a 1

If an expanded list provides too many, or too few, arguments, an ArgumentError will be raised.

If an expanded dict provides parameters which are not requested, an ArgumentError will be raised.

If an expanded dict provides an argument which has already been defined, either as a positional argument or through a named parameter, an error will be raised.

NOTE: Argument expansion is not currently supported for functions provided by C modules.

Ternary Expressions

Ternary expressions allow for branching conditions to be used in expression contexts:

print("true branch" if True else "false branch")
# → true branch

Ternary expressions perform short-circuit and will not evaluate the branch they do not take:

(print if True else explode)("What does this do?")
# → What does this do?

Docstrings

If the first expression in a module, function, class, or method is a string, it will be attached to the corresponding object in the field __doc__:

def foo():
    '''This is a function that does things.'''
    return 42

print(foo.__doc__)
# → This is a function that does things.

About the REPL

Kuroko's repl provides an interactive environment for executing code and seeing results.

When entering code at the repl, lines ending with colons (:) are treated specially - the repl will continue to accept input and automatically insert indentation on a new line. Please note that the repl's understanding of colons is naive: Whitespace or comments after a colon which would normally be accepted by Kuroko's parser will not be understood by the repl - if you want to place a comment after the start of a block statement, be sure that it ends in a colon so you can continue to enter statements.

Pressing backspace when the cursor is preceded by whitespace will delete up to the last column divisible by 4, which should generally delete one level of indentation automatically.

The tab key will also produce spaces when the cursor is at the beginning of the line or preceded entirely with white space.

The repl will display indentation level indicators in preceding whitespace as a helpful guide.

When a blank line or a line consisting entirely of whitespace is entered, the repl will process the full input.

Code executed in the repl runs in a global scope and reused variable names will overwrite previous definitions, allowing function and class names to be reused.

The repl will display the last value popped from the stack before returning.

Tab completion will provide the names of globals, as well as the fields and methods of objects.

What's different from Python?

You may be looking at the code examples and thinking Kuroko looks a lot more like Python than "syntax similar to Python" suggests. Still, there are some differences, and they come in two forms: Intentional differences and unintentional differences.

Unintentional differences likely represent incomplete features. Intentional differences are design decisions specifically meant to differentiate Kuroko from Python and usually are an attempt to improve upon or "fix" perceived mistakes.

Two notable intentional differences thus far are:

• Kuroko's variable scoping requires explicit declarations. This was done because Python's function-level scoping, and particularly how it interacts with globals, is often a thorn in the side of beginner and seasoned programmers alike. It's not so much seen as a mistake as it is something we don't wish to replicate.
• Default arguments to functions are evaluated at call time, not at definition time. How many times have you accidentally assigned an empty list as a default argument, only to be burned by its mutated descendent appearing in further calls? Kuroko doesn't do that - it works more like Ruby.

Interfacing C with Kuroko

There are two ways to connect Kuroko with C code: embedding and modules.

Embedding involves including the interpreter library and initializing and managing the VM state yourself.

C modules allow C code to provide functions through imported modules.

If you want to provide C functionality for Kuroko, build a module. If you want to provide Kuroko as a scripting language in a C project, embed the interpreter.

With either approach, the API provided by Kuroko is the same beyond initialization.

Embedding Kuroko

Kuroko is built as a shared libary, libkuroko.so, which can be linked against. libkuroko.so generally depends on the system dynamic linker, which may involve an additional library (eg. -ldl).

The simplest example of embedding Kuroko is to initialize the VM and interpret an embedded line of code:

#include <stdio.h>
#include <kuroko.h>

int main(int argc, char *argv[]) {
    krk_initVM();
    krk_interpret("import kuroko\nprint('Kuroko',kuroko.version)\n", 1, "<stdin>","<stdin>");
    krk_freeVM();
    return 0;
}

There is a single, shared VM state. krk_initVM() will initialize the compiler and create built-in objects.

krk_interpret compiles and executes a block of code and takes the following arguments:

KrkValue krk_interpret(const char *sourceText, int newModuleScope, char *fromName, char *fromFile);

If newModuleScope is non-zero, the interpreter will parse code in the context of a new module and the KrkValue returned will be a module object.

If newModuleScope is zero, the return value will be the last value popped from the stack during execution of sourceText. This can be used, as in the REPL, when provided interactive sessions.

The arguments fromName provide the name of the module created by when newModuleScope is non-zero, and fromFile will be used when displaying tracebacks.

Building Modules

Modules are shared objects with at least one exported symbol: krk_module_onload_{NAME} (where {NAME} is the name of your module, which should also be the name of your shared object file excluding the .so suffix).

Your module's krk_module_onload_... function should return a KrkValue representing a KrkInstance of the vm.moduleClass class.

KrkValue krk_module_onload_fileio(void) {
	KrkInstance * module = krk_newInstance(vm.moduleClass);
	/* Store it on the stack for now so we can do stuff that may trip GC
	 * and not lose it to garbage colletion... */
	krk_push(OBJECT_VAL(module));

	/* ... */

	/* Pop the module object before returning; it'll get pushed again
	 * by the VM before the GC has a chance to run, so it's safe. */
	assert(AS_INSTANCE(krk_pop()) == module);
	return OBJECT_VAL(module);
}

Defining Native Functions

Simple functions may be added to the interpreter by binding to them to vm.builtins or your own module instance.

Native functions should have a call signature as follows:

KrkNative my_native_function(int argc, KrkValue argv[], int hasKw);

If hasKw is non-zero, then the value in argv[argc-1] will represent a dictionary of keyword and value pairs. Positional arguments will provided in order in the other indexes of argv.

Functions must return a value. If you do not need to return data to callers, return NONE_VAL().

To bind the function, use krk_defineNative:

krk_defineNative(&vm.builtins->fields, "my_native_function", my_native_function);

Binding to vm.builtins->fields will make your function accessible from any scope (if its name is not shadowed by a module global or function local) and is discouraged for modules but recommended for embedded applications.

Kuroko's Object Model

For both embedding and C modules, you will likely want to create and attach functions, classes, objects, and so on.

It is recommended you read Crafting Interpreters, particularly the third section describing the implementation of clox, as a primer on the basic mechanisms of the value system that Kuroko is built upon.

Essentially, everything accessible to the VM is represented as a KrkValue, which this documentation will refer to simply as a value from here on out.

Values are small, fixed-sized items and are generally considered immutable. Simple types, such as integers, booleans, and None, are directly represented as values and do not exist in any other form.

More complex types are represented by subtypes of KrkObj known as objects, and values that represent them contain pointers to these KrkObjs. The KrkObjs themselves live on the heap and are managed by the garbage collector.

Strings, functions, closures, classes, instances, and tuples are all basic objects and carry additional data in their heap representations.

Strings (KrkString) are immutable and deduplicated - any two strings with the same text have the same object. (See Crafting Interpreters, chapter 19) Strings play a heavy role in the object model, providing the basic type for indexing into attribute tables in classes and instances.

Functions (KrkFunction) represent bytecode, argument lists, default values, local names, and constants - the underlying elements of execution for a function. Generally, functions are not relevant to either embedding or C modules and are an internal implementation detail of the VM.

Closures (KrkClosure) represent the callable objects for functions defined in user code. When embedding or building a C module, you may need to deal with closures for Kuroko code passed to your C code.

Bound methods (KrkBoundMethod) connect methods with the "self" object they belong to, allowing a single value to be passed on the stack and through fields.

Classes (KrkClass) represent collections of functions. In Kuroko, all object and value types have a corresponding KrkClass.

Instances (KrkInstance) represent user objects and store fields in a hashmap and also point to the class they are an instance of. Instances can represent many things, including collections like lists and dictionaries, modules, and so on.

Tuples (KrkTuple) represent simple fixed-sized lists and are intended to be immutable.

Finally, native functions (KrkNative) represent callable references to C code.

Most extensions to Kuroko, both in the form of embedded applications and C modules, will primarily deal with classes, instances, strings, and native functions.

Two of the high-level collection types, lists and dictionaries, are instances of classes provided by the __builtins__ module. While they are not their own type of KrkObj, some macros are provided to deal with them.

Creating Objects

Now that we've gotten the introduction out of the way, we can get to actually creating and using these things.

The C module example above demonstrates the process of creating an object in the form of an instance of the vm.moduleClass class. All C modules should create one of these instances to expose other data to user code that imports them.

Most extensions will also want to provide their own types through classes, as well as create instances of those classes.

NOTE: When creating and attaching objects, pay careful attention to the order in which you allocate new objects, including strings. If two allocations happen in sequence without the first allocated object being stored in a location reachable from the interpreter roots, the second allocation may trigger the garbage collector which will immediately free the first object. If you need to deal with complex allocation patterns, place values temporarily on the stack to prevent them from being collected.

    /* Create a class 'className_' and attach it to our module. */
    KrkClass * myNewClass = krk_newClass(krk_copyString("MyNewClass", 10));
    krk_attachNamedObject(&module->fields, "MyNewClass", (KrkObj*)myNewClass);

Here we have created a new class named MyNameClass and exposed it through the fields table of our module object under the same name. We're not done preparing our class, though:

    myNewClass->base = vm.objectClass; \
    krk_tableAddAll(&vm.objectClass->methods, &myNewClass->methods); \

We also want to make sure that our new class fits into the general inheritence hierarchy, which typically means inheriting from vm.objectClass - we do this by setting our new class's base pointer to vm.objectClass and copying vm.objectClass's method table. Now we can start customizing our class with its own methods.

Native functions are attached to class method tables in a similar manner to normal functions:

krk_defineNative(&myNewClass->methods, ".my_native_method", my_native_method);

When attaching methods, notice the . at the start of the name. This indicates to krk_defineNative that this method will take a "self" value as its first argument. This affects how the VM modifies the stack when calling native code and allows native functions to integrate with user code functions and methods.

In addition to methods, native functions may also provide classes with dynamic fields. A dynamic field works much like a method, but it is called implictly when the field is accessed. Dynamic fields are used by the native classes to provide non-instance values with field values.

krk_defineNative(&myNewClass->methods, ":my_dynamic_field", my_dynamic_field);

If your new instances of your class will be created by user code, you can provide an __init__ method, or any of the other special methods described in the Examples below.