prog8/compiler/antlr/prog8.g4
Irmen de Jong 32d82e8e56 asmsubs can now specify params and return values to be on the stack.
implemented fast integer fixed-point sin and cosine functions.
2019-01-04 22:11:46 +01:00

262 lines
6.4 KiB
ANTLR

/*
Prog8 combined lexer and parser grammar
NOTES:
- whitespace is ignored. (tabs/spaces)
- every position can be empty, be a comment, or contain ONE statement.
*/
grammar prog8;
LINECOMMENT : [\r\n][ \t]* COMMENT -> channel(HIDDEN);
COMMENT : ';' ~[\r\n]* -> channel(HIDDEN) ;
WS : [ \t] -> skip ;
EOL : [\r\n]+ ;
NAME : [a-zA-Z_][a-zA-Z0-9_]* ;
DEC_INTEGER : ('0'..'9') | (('1'..'9')('0'..'9')+);
HEX_INTEGER : '$' (('a'..'f') | ('A'..'F') | ('0'..'9'))+ ;
BIN_INTEGER : '%' ('0' | '1')+ ;
FLOAT_NUMBER : FNUMBER (('E'|'e') ('+' | '-')? FNUMBER)? ; // sign comes later from unary expression
fragment FNUMBER : ('0' .. '9') + ('.' ('0' .. '9') +)? ;
fragment STRING_ESCAPE_SEQ : '\\' . | '\\' EOL;
STRING :
'"' ( STRING_ESCAPE_SEQ | ~[\\\r\n\f"] )* '"'
{
// get rid of the enclosing quotes
String s = getText();
setText(s.substring(1, s.length() - 1));
}
;
INLINEASMBLOCK :
'{{' .+? '}}'
{
// get rid of the enclosing double braces
String s = getText();
setText(s.substring(2, s.length() - 2));
}
;
SINGLECHAR :
'\'' ( STRING_ESCAPE_SEQ | ~[\\\r\n\f"] ) '\''
{
// get rid of the enclosing quotes
String s = getText();
setText(s.substring(1, s.length() - 1));
}
;
module : (modulestatement | EOL)* EOF ;
modulestatement: directive | block ;
block: '~' identifier integerliteral? statement_block EOL ;
statement :
directive
| varinitializer
| vardecl
| constdecl
| memoryvardecl
| assignment
| augassignment
| unconditionaljump
| postincrdecr
| functioncall_stmt
| if_stmt
| branch_stmt
| subroutine
| asmsubroutine
| inlineasm
| returnstmt
| forloop
| whileloop
| repeatloop
| breakstmt
| continuestmt
| labeldef
;
labeldef : identifier ':' ;
unconditionaljump : 'goto' (integerliteral | scoped_identifier) ;
directive :
directivename=('%output' | '%launcher' | '%zeropage' | '%zpreserved' | '%address' | '%import' |
'%breakpoint' | '%asminclude' | '%asmbinary' | '%option')
(directivearg? | directivearg (',' directivearg)*)
;
directivearg : stringliteral | identifier | integerliteral ;
vardecl: datatype arrayspec? identifier ;
varinitializer : datatype arrayspec? identifier '=' expression ;
constdecl: 'const' varinitializer ;
memoryvardecl: 'memory' varinitializer;
datatype: 'ubyte' | 'byte' | 'uword' | 'word' | 'float' | 'str' | 'str_p' | 'str_s' | 'str_ps' ;
arrayspec: '[' expression ']' ;
assignment : assign_targets '=' expression ;
assign_targets : assign_target (',' assign_target)* ;
augassignment :
assign_target operator=('+=' | '-=' | '/=' | '//=' | '*=' | '**=' | '&=' | '|=' | '^=' | '%=' ) expression
;
assign_target:
register
| scoped_identifier
| arrayindexed
| directmemory
;
postincrdecr : assign_target operator = ('++' | '--') ;
expression :
functioncall
| <assoc=right> prefix = ('+'|'-'|'~') expression
| left = expression bop = '**' right = expression
| left = expression bop = ('*' | '/' | '//' | '%' ) right = expression
| left = expression bop = ('+' | '-' ) right = expression
| left = expression bop = ('<' | '>' | '<=' | '>=') right = expression
| left = expression bop = ('==' | '!=') right = expression
| left = expression bop = '&' right = expression
| left = expression bop = '^' right = expression
| left = expression bop = '|' right = expression
| rangefrom = expression 'to' rangeto = expression ('step' rangestep = expression)? // can't create separate rule due to mutual left-recursion
| left = expression bop = 'and' right = expression
| left = expression bop = 'or' right = expression
| left = expression bop = 'xor' right = expression
| prefix = 'not' expression
| literalvalue
| register
| scoped_identifier
| arrayindexed
| directmemory
| expression typecast
| '(' expression ')'
;
typecast : 'as' datatype;
arrayindexed : scoped_identifier arrayspec ;
directmemory : '@' '(' expression ')';
functioncall : scoped_identifier '(' expression_list? ')' ;
functioncall_stmt : scoped_identifier '(' expression_list? ')' ;
expression_list :
expression (',' EOL? expression)* // you can split the expression list over several lines
;
returnstmt : 'return' expression_list? ;
breakstmt : 'break';
continuestmt: 'continue';
identifier : NAME ;
scoped_identifier : NAME ('.' NAME)* ;
register : 'A' | 'X' | 'Y' ;
registerorpair : 'A' | 'X' | 'Y' | 'AX' | 'AY' | 'XY' ; // only used in subroutine params and returnvalues
statusregister : 'Pc' | 'Pz' | 'Pn' | 'Pv' ;
integerliteral : intpart=(DEC_INTEGER | HEX_INTEGER | BIN_INTEGER) wordsuffix? ;
wordsuffix : '.w' ;
booleanliteral : 'true' | 'false' ;
arrayliteral : '[' EOL? expression (',' EOL? expression)* EOL? ']' ; // you can split the array list over several lines
stringliteral : STRING ;
charliteral : SINGLECHAR ;
floatliteral : FLOAT_NUMBER ;
literalvalue :
integerliteral
| booleanliteral
| arrayliteral
| stringliteral
| charliteral
| floatliteral
;
inlineasm : '%asm' INLINEASMBLOCK;
subroutine :
'sub' identifier '(' sub_params? ')' sub_return_part? (statement_block EOL)
;
sub_return_part : '->' sub_returns ;
statement_block :
'{' EOL
(statement | EOL) *
'}'
;
sub_params : vardecl (',' EOL? vardecl)* ;
sub_returns : datatype (',' EOL? datatype)* ;
asmsubroutine :
'asmsub' identifier '(' asmsub_params? ')'
'->' 'clobbers' '(' clobber? ')' '->' '(' asmsub_returns? ')' (asmsub_address | statement_block )
;
asmsub_address : '=' address=integerliteral ;
asmsub_params : asmsub_param (',' EOL? asmsub_param)* ;
asmsub_param : vardecl '@' (registerorpair | statusregister | stack='stack');
clobber : register (',' register)* ;
asmsub_returns : asmsub_return (',' EOL? asmsub_return)* ;
asmsub_return : datatype '@' (registerorpair | statusregister | stack='stack') ;
if_stmt : 'if' expression EOL? (statement | statement_block) EOL? else_part? ; // statement is constrained later
else_part : 'else' EOL? (statement | statement_block) ; // statement is constrained later
branch_stmt : branchcondition EOL? (statement | statement_block) EOL? else_part? EOL ;
branchcondition: 'if_cs' | 'if_cc' | 'if_eq' | 'if_z' | 'if_ne' | 'if_nz' | 'if_pl' | 'if_pos' | 'if_mi' | 'if_neg' | 'if_vs' | 'if_vc' ;
forloop : 'for' datatype? (register | identifier) 'in' expression EOL? statement_block ;
whileloop: 'while' expression EOL? (statement | statement_block) ;
repeatloop: 'repeat' (statement | statement_block) EOL? 'until' expression ;