-
A simple programming language aimed to be used for educational purposes, and to teach how programming language interpreters can be made.
-
Since it's made for educational purposes, all the parts that make a lexer are made from scratch
lexer
,parser
andevaluator
.
- The language is a
C
derived language, with the ability to createstructs
as complex types, and declare functions as First class values. - The language is dynamically typed, to make it easier to implement
// variable declarations
var a = 10
var c = "some string"
var arrayValue = [1, 2, 3]
func add(a, b) {
return a + b
}
func isBigger(a, b) {
return a > b
}
struct Data {
func init(a, b, c) {
this.a = a
this.c = a + b
}
func doSomething() {
return a + c
}
}
// STRUCT_DECL: struct Identifier BLOCK
// BLOCK: { functions* }
//
// var data = new Data(a, b, c)
// data.doSomething()
// Object declaration
var object = new()
object.field = "field value"
object.method = func(a) {
return a * 2
}
var value = object.call()
var loopConstruct = func(a, b) {
var sum = 0
for i = 1; i < 10; i += 1 {
sum += i
}
return sum
}
var conditional = func(a, b) {
}
// a program entrypoint
func main() {
var c = a + add(1, 4)
// builtin functions
println(c)
}
- Below are the grammar rules for the language (it can be left recursive, we will deal with it at some point)
// Parser rules
grammar Comet;
comet:
function_declaration+ EOF
;
function_declaration:
'func' IDENTIFIER'('parameterless')' '{' statements '}' # functionDeclaration
;
parameterless:
parameters*
;
parameters:
IDENTIFIER(','IDENTIFIER)* | expression (','expression)*
;
statements:
statement*
;
statement:
assignment_statement # assignment
| if_statement # ifStatement
| return_statement # returnStatement
| for_loop # forLoop
;
assignment_statement:
IDENTIFIER '=' expression
;
if_statement:
'if' '('expression')' '{' statements '}' ('else' '{' statements '}')?
;
return_statement:
'return' expression
;
for_loop:
'for' '('expression')' '{' statements '}'
;
# struct and object declaration to be defined later
# struct_statement:
# 'struct' IDENTIFIER '{' FIELD* '}'
expression:
function_call
| IDENTIFIER
| STRING
| NUMBER
| BOOL
| string_concat
| math_expression
| array_declaration
;
array_declaration:
'['(expression (','expression)*)?']'
function_call:
IDENTIFIER '('parameterless')'
;
string_concat:
STRING ('+' STRING)*
;
math_expression:
term (('+' | '-') term)*
;
term:
factor (('*' | '/') factor)*
;
factor:
IDENTIFIER
| string
| number
| '('expression')'
;
// Lexer rules
WS : [ \t\r\n\u000C]+ -> skip;
COMMENT : '/*' .*? '*/' -> skip;
LINE_COMMENT : '//' ~[\r\n]* -> skip;
IDENTIFIER: ('_'|LETTER)(LETTER|DIGIT|'_')*;
NUMBER: ('-' | '+')? (([1-9][0-9]*) | ([0-9]));
OPERATOR: '+' | '-' | '*' | '/' | '%' | '!';
fragment DIGIT: [0-9];
BOOL: 'true' | 'false';
LETTER: [a-zA-Z];
STRING: '"' ~('\r' | '\n' | '"')* '"' ;
These are the list of the tasks that I think should be done before having the first release:
- Add variable declarations
- Add conditionals
- Add Proper scoping
- Add support for loops
- Add Arrays support
- Add Comments support
- Add Hash support
- Add import modules support
- Add testing framework
- Add build system
- Add a standard library
- Feel free to submit new issues, or pull requests.