""" Programming Language for 6502/6510 microprocessors, codename 'Sick' This is the parser of the IL65 code, that generates a parse tree. Written by Irmen de Jong (irmen@razorvine.net) - license: GNU GPL 3.0 """ import math import builtins import inspect import enum from collections import defaultdict from typing import Union, Generator, Tuple, List, Optional, Dict, Any, no_type_check import attr from ply.yacc import yacc from .plylex import SourceRef, tokens, lexer, find_tok_column, print_warning from .datatypes import DataType, VarType, REGISTER_SYMBOLS, REGISTER_BYTES, REGISTER_WORDS, \ char_to_bytevalue, FLOAT_MAX_NEGATIVE, FLOAT_MAX_POSITIVE class ProgramFormat(enum.Enum): RAW = "raw" PRG = "prg" BASIC = "basicprg" class ZpOptions(enum.Enum): NOCLOBBER = "noclobber" CLOBBER = "clobber" CLOBBER_RESTORE = "clobber_restore" math_functions = {name: func for name, func in vars(math).items() if inspect.isbuiltin(func) and name != "pow" and not name.startswith("_")} builtin_functions = {name: func for name, func in vars(builtins).items() if inspect.isbuiltin(func) and not name.startswith("_")} class ParseError(Exception): def __init__(self, message: str, sourceref: SourceRef) -> None: super().__init__(message) self.sourceref = sourceref # @todo chain attribute, a list of other exceptions, so we can have more than 1 error at a time. def __str__(self): return "{} {:s}".format(self.sourceref, self.args[0]) class ExpressionEvaluationError(ParseError): pass class UndefinedSymbolError(LookupError): pass start = "start" @attr.s(cmp=False, slots=True, frozen=False, repr=False) class AstNode: # all ast nodes have: sourceref, parent, and nodes (=list of zero or more sub-nodes) sourceref = attr.ib(type=SourceRef) parent = attr.ib(init=False, default=None) # will be hooked up later nodes = attr.ib(type=list, init=False, default=attr.Factory(list)) # type: List['AstNode'] @property def lineref(self) -> str: return "src l. " + str(self.sourceref.line) def my_scope(self) -> 'Scope': # returns the closest Scope in the ancestry of this node, or raises LookupError if no scope is found scope = self.parent while scope: if isinstance(scope, Scope): return scope scope = scope.parent raise LookupError("no scope found in node ancestry", self) def all_nodes(self, *nodetypes: type) -> Generator['AstNode', None, None]: nodetypes = nodetypes or (AstNode, ) child_nodes = list(self.nodes) for node in child_nodes: if isinstance(node, nodetypes): yield node for node in child_nodes: if isinstance(node, AstNode): yield from node.all_nodes(*nodetypes) def remove_node(self, node: 'AstNode') -> None: assert node.parent is self self.nodes.remove(node) node.parent = None def replace_node(self, oldnode: 'AstNode', newnode: 'AstNode') -> None: assert oldnode.parent is self assert isinstance(newnode, AstNode) idx = self.nodes.index(oldnode) self.nodes[idx] = newnode newnode.parent = self oldnode.parent = None def add_node(self, newnode: 'AstNode', index: int = None) -> None: assert isinstance(newnode, AstNode) if index is None: self.nodes.append(newnode) else: self.nodes.insert(index, newnode) newnode.parent = self @attr.s(cmp=False) class Directive(AstNode): name = attr.ib(type=str) args = attr.ib(type=list, default=attr.Factory(list)) # no subnodes. @attr.s(cmp=False, slots=True, repr=False) class Scope(AstNode): # has zero or more subnodes level = attr.ib(type=str, init=True) nodes = attr.ib(type=list, init=True) # requires nodes in __init__ symbols = attr.ib(init=False) name = attr.ib(init=False) # will be set by enclosing block, or subroutine etc. float_const_values = attr.ib(type=dict, default=attr.Factory(dict), init=False) # floatingpoint number -> float const name _save_registers = attr.ib(type=bool, default=None, init=False) @property def save_registers(self) -> bool: if self._save_registers is not None: return self._save_registers try: return self.my_scope().save_registers except LookupError: return False @save_registers.setter def save_registers(self, save: bool) -> None: self._save_registers = save @property def parent_scope(self) -> Optional['Scope']: parent_scope = self.parent while parent_scope and not isinstance(parent_scope, Scope): parent_scope = parent_scope.parent return parent_scope def __attrs_post_init__(self): # populate the symbol table for this scope for fast lookups via scope.lookup("name") or scope.lookup("dotted.name") self.symbols = {} for node in self.nodes: assert isinstance(node, AstNode) self._populate_symboltable(node) def _populate_symboltable(self, node: AstNode) -> None: if isinstance(node, (Label, VarDef)): if node.name in self.symbols: raise ParseError("symbol '{}' already defined at {}".format(node.name, self.symbols[node.name].sourceref), node.sourceref) self.symbols[node.name] = node elif isinstance(node, (Subroutine, BuiltinFunction)): if node.name in self.symbols: raise ParseError("symbol '{}' already defined at {}".format(node.name, self.symbols[node.name].sourceref), node.sourceref) self.symbols[node.name] = node elif isinstance(node, (Block, Scope)): if node.name: if node.name != "ZP" and node.name in self.symbols: raise ParseError("symbol '{}' already defined at {}" .format(node.name, self.symbols[node.name].sourceref), node.sourceref) self.symbols[node.name] = node @no_type_check def define_builtin_functions(self) -> None: for name, func in math_functions.items(): f = BuiltinFunction(name=name, func=func, sourceref=self.sourceref) self.add_node(f) for name, func in builtin_functions.items(): f = BuiltinFunction(name=name, func=func, sourceref=self.sourceref) self.add_node(f) def define_float_constant(self, value: float) -> str: if value in self.float_const_values: return self.float_const_values[value] name = "il65_float_const_" + str(1 + len(self.float_const_values)) self.float_const_values[name] = value return name def lookup(self, name: str) -> AstNode: assert isinstance(name, str) if '.' in name: # look up the dotted name starting from the topmost scope scope = self node = self while node.parent: if isinstance(node.parent, Scope): scope = node.parent node = node.parent for namepart in name.split('.'): if isinstance(scope, (Block, Subroutine)): scope = scope.scope if not isinstance(scope, Scope): raise UndefinedSymbolError("undefined symbol: " + name) scope = scope.symbols.get(namepart, None) if not scope: raise UndefinedSymbolError("undefined symbol: " + name) return scope else: # find the name in nested scope hierarchy if name in self.symbols: return self.symbols[name] parent_scope = self.parent while parent_scope and not isinstance(parent_scope, Scope): parent_scope = parent_scope.parent if parent_scope: return parent_scope.lookup(name) raise UndefinedSymbolError("undefined symbol: " + name) def remove_node(self, node: AstNode) -> None: if hasattr(node, "name"): try: del self.symbols[node.name] # type: ignore except KeyError: pass super().remove_node(node) def replace_node(self, oldnode: AstNode, newnode: AstNode) -> None: if hasattr(oldnode, "name"): del self.symbols[oldnode.name] # type: ignore super().replace_node(oldnode, newnode) def add_node(self, newnode: AstNode, index: int=None) -> None: super().add_node(newnode, index) self._populate_symboltable(newnode) def validate_address(obj: AstNode, attrib: attr.Attribute, value: Optional[int]) -> None: if value is None: return if isinstance(obj, Block) and obj.name == "ZP": raise ParseError("zeropage block cannot have custom start {:s}".format(attrib.name), obj.sourceref) if value < 0x0200 or value > 0xffff: raise ParseError("invalid {:s} (must be from $0200 to $ffff)".format(attrib.name), obj.sourceref) def dimensions_validator(obj: 'DatatypeNode', attrib: attr.Attribute, value: List[int]) -> None: if not value: return dt = obj.to_enum() if value and dt not in (DataType.MATRIX, DataType.WORDARRAY, DataType.BYTEARRAY): raise ParseError("cannot use a dimension for this datatype", obj.sourceref) if dt == DataType.WORDARRAY or dt == DataType.BYTEARRAY: if len(value) == 1: if value[0] <= 0 or value[0] > 256: raise ParseError("array length must be 1..256", obj.sourceref) else: raise ParseError("array must have only one dimension", obj.sourceref) if dt == DataType.MATRIX: if len(value) < 2 or len(value) > 3: raise ParseError("matrix must have two dimensions, with optional interleave", obj.sourceref) if len(value) == 3: if value[2] < 1 or value[2] > 256: raise ParseError("matrix interleave must be 1..256", obj.sourceref) if value[0] < 0 or value[0] > 128 or value[1] < 0 or value[1] > 128: raise ParseError("matrix rows and columns must be 1..128", obj.sourceref) @attr.s(cmp=False, repr=False) class Block(AstNode): # has one subnode: the Scope. name = attr.ib(type=str, default=None) address = attr.ib(type=int, default=None, validator=validate_address) _unnamed_block_labels = {} # type: Dict[Block, str] @property def scope(self) -> Scope: return self.nodes[0] if self.nodes else None # type: ignore @scope.setter def scope(self, scope: Scope) -> None: assert isinstance(scope, Scope) self.nodes.clear() self.nodes.append(scope) scope.name = self.name @property def label(self) -> str: if self.name: return self.name if self in self._unnamed_block_labels: return self._unnamed_block_labels[self] label = "il65_block_{:d}".format(len(self._unnamed_block_labels)) self._unnamed_block_labels[self] = label return label @attr.s(cmp=False, repr=False) class Module(AstNode): # has one subnode: the Scope. name = attr.ib(type=str) # filename subroutine_usage = attr.ib(type=defaultdict, init=False, default=attr.Factory(lambda: defaultdict(set))) # will be populated later format = attr.ib(type=ProgramFormat, init=False, default=ProgramFormat.PRG) # can be set via directive address = attr.ib(type=int, init=False, default=0xc000, validator=validate_address) # can be set via directive zp_options = attr.ib(type=ZpOptions, init=False, default=ZpOptions.NOCLOBBER) # can be set via directive @property def scope(self) -> Scope: return self.nodes[0] if self.nodes else None # type: ignore @no_type_check def zeropage(self) -> Optional[Block]: # return the zeropage block (if defined) first_block = next(self.scope.all_nodes(Block)) if first_block.name == "ZP": return first_block return None @no_type_check def main(self) -> Optional[Block]: # return the 'main' block (if defined) for block in self.scope.all_nodes(Block): if block.name == "main": return block return None @attr.s(cmp=False) class Label(AstNode): name = attr.ib(type=str) # no subnodes. @attr.s(cmp=False, slots=True, repr=False) class Expression(AstNode): # just a common base class for the nodes that are an expression themselves: # ExpressionWithOperator, AddressOf, LiteralValue, SymbolName, Register, SubCall, Dereference def is_compile_constant(self) -> bool: raise NotImplementedError("implement in subclass") def const_value(self) -> Union[int, float, bool, str]: raise NotImplementedError("implement in subclass") @attr.s(cmp=False, slots=True) class Register(Expression): name = attr.ib(type=str, validator=attr.validators.in_(REGISTER_SYMBOLS)) datatype = attr.ib(type=DataType, init=False) # no subnodes. def __attrs_post_init__(self): if self.name in REGISTER_BYTES: self.datatype = DataType.BYTE elif self.name in REGISTER_WORDS: self.datatype = DataType.WORD else: self.datatype = None # register 'SC' etc. def __hash__(self) -> int: return hash(self.name) def __eq__(self, other) -> bool: if not isinstance(other, Register): return NotImplemented return self.name == other.name def __lt__(self, other) -> bool: if not isinstance(other, Register): return NotImplemented return self.name < other.name def is_compile_constant(self) -> bool: return False def const_value(self) -> Union[int, float, bool, str]: raise TypeError("register doesn't have a constant numeric value", self) @attr.s(cmp=False) class PreserveRegs(AstNode): registers = attr.ib(type=str) # no subnodes. @attr.s(cmp=False) class TargetRegisters(AstNode): # subnodes is is a list of 1 or more registers. # In it's multiple-register form it is only used to be able to parse # the result of a subroutine call such as A,X = sub(). # It will be replaced by a regular Register node if it contains just one register. pass @attr.s(cmp=False, repr=False) class InlineAssembly(AstNode): # no subnodes. assembly = attr.ib(type=str) @attr.s(cmp=False, slots=True) class DatatypeNode(AstNode): # no subnodes. name = attr.ib(type=str) dimensions = attr.ib(type=list, default=None, validator=dimensions_validator) # if set, 1 or more dimensions (ints) def to_enum(self): return { "byte": DataType.BYTE, "word": DataType.WORD, "float": DataType.FLOAT, "text": DataType.STRING, "ptext": DataType.STRING_P, "stext": DataType.STRING_S, "pstext": DataType.STRING_PS, "matrix": DataType.MATRIX, "array": DataType.BYTEARRAY, "wordarray": DataType.WORDARRAY }[self.name] @attr.s(cmp=False, repr=False) class BuiltinFunction(AstNode): # This is a pseudo-node that will be artificially injected in the top-most scope, # to represent all supported built-in functions or math-functions. # No child nodes. name = attr.ib(type=str) func = attr.ib(type=callable) @attr.s(cmp=False, repr=False) class Subroutine(AstNode): # one subnode: the Scope. name = attr.ib(type=str) param_spec = attr.ib(type=list) result_spec = attr.ib(type=list) address = attr.ib(type=int, default=None, validator=validate_address) @property def scope(self) -> Scope: return self.nodes[0] if self.nodes else None # type: ignore @scope.setter def scope(self, scope: Scope) -> None: assert isinstance(scope, Scope) self.nodes.clear() self.nodes.append(scope) scope.name = self.name if self.address is not None: raise ValueError("subroutine must have either a scope or an address, not both") @attr.s(cmp=True, slots=True, repr=False) class LiteralValue(Expression): # no subnodes. value = attr.ib() def __repr__(self) -> str: return "".format(self.value, self.sourceref) def const_value(self) -> Union[int, float, bool, str]: return self.value def is_compile_constant(self) -> bool: return True @attr.s(cmp=False) class AddressOf(Expression): # no subnodes. name = attr.ib(type=str, validator=attr.validators._InstanceOfValidator(type=str)) def is_compile_constant(self) -> bool: return False def const_value(self) -> Union[int, float, bool, str]: symdef = self.my_scope().lookup(self.name) if isinstance(symdef, VarDef): if symdef.zp_address is not None: return symdef.zp_address if symdef.vartype == VarType.MEMORY: return symdef.value.const_value() raise TypeError("can only take constant address of a memory mapped variable", self) raise TypeError("should be a vardef to be able to take its address", self) @attr.s(cmp=False, slots=True) class SymbolName(Expression): # no subnodes. name = attr.ib(type=str) def is_compile_constant(self) -> bool: symdef = self.my_scope().lookup(self.name) return isinstance(symdef, VarDef) and symdef.vartype == VarType.CONST def const_value(self) -> Union[int, float, bool, str]: symdef = self.my_scope().lookup(self.name) if isinstance(symdef, VarDef) and symdef.vartype == VarType.CONST: return symdef.const_value() raise TypeError("should be a const vardef to be able to take its constant numeric value", self) @attr.s(cmp=False) class Dereference(Expression): # one subnode: operand (SymbolName, integer LiteralValue or Register) datatype = attr.ib() size = attr.ib(type=int, default=None) @property def operand(self) -> Union[SymbolName, LiteralValue, Register]: return self.nodes[0] # type: ignore def __attrs_post_init__(self): # convert datatype node to enum + size if self.datatype is None: assert self.size is None self.size = 1 self.datatype = DataType.BYTE elif isinstance(self.datatype, DatatypeNode): assert self.size is None self.size = self.datatype.dimensions if not self.datatype.to_enum().isnumeric(): raise ParseError("dereference target value must be byte, word, float", self.datatype.sourceref) self.datatype = self.datatype.to_enum() if self.nodes and not isinstance(self.nodes[0], (SymbolName, LiteralValue, Register)): raise TypeError("operand of dereference invalid type", self.nodes[0], self.sourceref) def is_compile_constant(self) -> bool: return False def const_value(self) -> Union[int, float, bool, str]: raise TypeError("dereference is not a constant numeric value") @attr.s(cmp=False) class IncrDecr(AstNode): # increment or decrement something by a CONSTANT value (1 or more) # one subnode: target (TargetRegisters, Register, SymbolName, or Dereference). operator = attr.ib(type=str, validator=attr.validators.in_(["++", "--"])) howmuch = attr.ib(default=1) @property def target(self) -> Union[TargetRegisters, Register, SymbolName, Dereference]: return self.nodes[0] # type: ignore @target.setter def target(self, target: Union[TargetRegisters, Register, SymbolName, Dereference]) -> None: if isinstance(target, Register): if target.name not in REGISTER_BYTES | REGISTER_WORDS: raise ParseError("cannot incr/decr that register", self.sourceref) if isinstance(target, TargetRegisters): raise ParseError("cannot incr/decr multiple registers at once", self.sourceref) assert isinstance(target, (Register, SymbolName, Dereference)) self.nodes.clear() self.nodes.append(target) def __attrs_post_init__(self): # make sure the amount is always >= 0 if self.howmuch < 0: self.howmuch = -self.howmuch self.operator = "++" if self.operator == "--" else "--" @attr.s(cmp=False, slots=True, repr=False) class ExpressionWithOperator(Expression): # 2 nodes: left (Expression), right (not present if unary, Expression if not unary) operator = attr.ib(type=str) # when evaluating the expression, does it have to be a compile-time constant value? must_be_constant = attr.ib(type=bool, init=False, default=False) @property def unary(self) -> bool: return len(self.nodes) == 1 @property def left(self) -> Expression: return self.nodes[0] # type: ignore @left.setter def left(self, newleft: Expression) -> None: if self.nodes: self.nodes[0] = newleft else: self.nodes.append(newleft) @property def right(self) -> Optional[Expression]: return self.nodes[1] if len(self.nodes) == 2 else None # type: ignore @right.setter def right(self, newright: Expression) -> None: self.nodes[1] = newright def __attrs_post_init__(self): assert self.operator not in ("++", "--"), "incr/decr should not be an expression" if self.operator == "mod": self.operator = "%" # change it back to the more common '%' def const_value(self) -> Union[int, float, bool, str]: raise TypeError("an expression is not a constant", self) def is_compile_constant(self) -> bool: return False def evaluate_primitive_constants(self, sourceref: SourceRef) -> LiteralValue: # make sure the lvalue and rvalue are primitives, and the operator is allowed assert isinstance(self.left, LiteralValue) assert isinstance(self.right, LiteralValue) if self.operator not in {'+', '-', '*', '/', '//', '~', '|', '&', '%', '<<', '>>', '<', '>', '<=', '>=', '==', '!='}: raise ValueError("operator", self.operator) estr = "{} {} {}".format(repr(self.left.value), self.operator, repr(self.right.value)) try: lv = LiteralValue(value=eval(estr, {}, {}), sourceref=sourceref) # type: ignore # safe because of checks above lv.parent = self.parent return lv except ZeroDivisionError: raise ParseError("division by zero", sourceref) except Exception as x: raise ExpressionEvaluationError("expression error: " + str(x), self.sourceref) from None @attr.s(cmp=False, repr=False) class Goto(AstNode): # one or two subnodes: target (SymbolName, integer LiteralValue, or Dereference) and optionally: condition (Expression) if_stmt = attr.ib(default=None) @property def target(self) -> Union[SymbolName, LiteralValue, Dereference]: return self.nodes[0] # type: ignore @property def condition(self) -> Optional[Expression]: return self.nodes[1] if len(self.nodes) == 2 else None # type: ignore @attr.s(cmp=False, slots=True, repr=False) class CallArgument(AstNode): # one subnode: the value (Expression) name = attr.ib(type=str, default=None) @property def value(self) -> Expression: return self.nodes[0] # type: ignore @attr.s(cmp=False) class CallArguments(AstNode): # subnodes are zero or more subroutine call arguments (CallArgument) nodes = attr.ib(type=list, init=True) # requires nodes in __init__ @attr.s(cmp=False, repr=False) class SubCall(Expression): # has three subnodes: # 0: target (Symbolname, integer LiteralValue, or Dereference), # 1: preserve_regs (PreserveRegs) # 2: arguments (CallArguments). @property def target(self) -> Union[SymbolName, LiteralValue, Dereference]: return self.nodes[0] # type: ignore @property def preserve_regs(self) -> PreserveRegs: return self.nodes[1] # type: ignore @property def arguments(self) -> CallArguments: return self.nodes[2] # type: ignore def is_compile_constant(self) -> bool: return False def const_value(self) -> Union[int, float, bool, str]: raise TypeError("subroutine call is not a constant value", self) @attr.s(cmp=False, slots=True, repr=False) class VarDef(AstNode): # zero or one subnode: value (Expression). name = attr.ib(type=str) vartype = attr.ib() datatype = attr.ib() size = attr.ib(type=list, default=None) zp_address = attr.ib(type=int, default=None, init=False) # the address in the zero page if this var is there, will be set later @property def value(self) -> Expression: return self.nodes[0] if self.nodes else None # type: ignore @value.setter def value(self, value: Expression) -> None: assert isinstance(value, Expression) if self.nodes: self.nodes[0] = value else: self.nodes.append(value) if isinstance(value, ExpressionWithOperator): # an expression in a vardef should evaluate to a compile-time constant: value.must_be_constant = True def const_value(self) -> Union[int, float, bool, str]: if self.vartype != VarType.CONST: raise TypeError("not a constant value", self) if self.nodes and isinstance(self.nodes[0], Expression): return self.nodes[0].const_value() raise ValueError("no value", self) def __attrs_post_init__(self): # convert vartype to enum if self.vartype == "const": self.vartype = VarType.CONST elif self.vartype == "var": self.vartype = VarType.VAR elif self.vartype == "memory": self.vartype = VarType.MEMORY else: raise ValueError("invalid vartype", self.vartype) # convert datatype node to enum + size if self.datatype is None: assert self.size is None self.size = [1] self.datatype = DataType.BYTE elif isinstance(self.datatype, DatatypeNode): assert self.size is None self.size = self.datatype.dimensions or [1] self.datatype = self.datatype.to_enum() if self.datatype == DataType.MATRIX and len(self.size) not in (2, 3): raise ValueError("matrix size should be 2 dimensions with optional interleave", self) if self.datatype.isarray() and sum(self.size) in (0, 1): print("warning: {}: array/matrix with size 1, use normal byte/word instead".format(self.sourceref)) if self.value is None and (self.datatype.isnumeric() or self.datatype.isarray()): self.value = LiteralValue(value=0, sourceref=self.sourceref) self.value.parent = self # if it's a matrix with interleave, it must be memory mapped if self.datatype == DataType.MATRIX and len(self.size) == 3: if self.vartype != VarType.MEMORY: raise ParseError("matrix with interleave can only be a memory-mapped variable", self.sourceref) # note: value coercion is done later, when all expressions are evaluated @attr.s(cmp=False, repr=False) class Return(AstNode): # one, two or three subnodes: value_A, value_X, value_Y (all three Expression) @property def value_A(self) -> Optional[Expression]: return self.nodes[0] if len(self.nodes) >= 1 else None # type: ignore @property def value_X(self) -> Optional[Expression]: return self.nodes[1] if len(self.nodes) >= 2 else None # type: ignore @property def value_Y(self) -> Optional[Expression]: return self.nodes[2] if len(self.nodes) >= 3 else None # type: ignore @attr.s(cmp=False, slots=True, repr=False) class AssignmentTargets(AstNode): # a list of one or more assignment targets (TargetRegisters, Register, SymbolName, or Dereference). nodes = attr.ib(type=list, init=True) # requires nodes in __init__ def has_memvalue(self) -> bool: for t in self.nodes: if isinstance(t, Dereference): return True if isinstance(t, SymbolName): symdef = self.my_scope().lookup(t.name) if isinstance(symdef, VarDef) and symdef.vartype == VarType.MEMORY: return True return False def same_targets(self, other: 'AssignmentTargets') -> bool: if len(self.nodes) != len(other.nodes): return False # @todo be able to compare targets in different order as well (sort them) for t1, t2 in zip(self.nodes, other.nodes): if type(t1) is not type(t2): return False if isinstance(t1, TargetRegisters): pass elif isinstance(t1, Register): if t1 != t2: # __eq__ is defined return False elif isinstance(t1, SymbolName): if t1.name != t2.name: return False elif isinstance(t1, Dereference): if t1.size != t2.size or t1.datatype != t2.datatype: return False op1, op2 = t1.operand, t2.operand if type(op1) is not type(op2): return False if isinstance(op1, SymbolName): if op1.name != op2.name: return False else: if op1 != op2: return False else: return False return True @attr.s(cmp=False, slots=True, repr=False) class Assignment(AstNode): # can be single- or multi-assignment # has two subnodes: left (=AssignmentTargets) and right (=Expression, # or another Assignment but those will be converted into multi assign) @property def left(self) -> AssignmentTargets: return self.nodes[0] # type: ignore @property def right(self) -> Expression: return self.nodes[1] # type: ignore @right.setter def right(self, rvalue: Expression) -> None: assert isinstance(rvalue, Expression) self.nodes[1] = rvalue @attr.s(cmp=False, slots=True, repr=False) class AugAssignment(AstNode): # has two subnodes: left (=TargetRegisters, Register, SymbolName, or Dereference) and right (=Expression) operator = attr.ib(type=str) @property def left(self) -> Union[TargetRegisters, Register, SymbolName, Dereference]: return self.nodes[0] # type: ignore @property def right(self) -> Expression: return self.nodes[1] # type: ignore @right.setter def right(self, rvalue: Expression) -> None: assert isinstance(rvalue, Expression) self.nodes[1] = rvalue def datatype_of(targetnode: AstNode, scope: Scope) -> DataType: # tries to determine the DataType of an assignment target node if isinstance(targetnode, VarDef): return DataType.WORD if targetnode.vartype == VarType.MEMORY else targetnode.datatype elif isinstance(targetnode, (Dereference, Register)): return targetnode.datatype elif isinstance(targetnode, SymbolName): symdef = scope.lookup(targetnode.name) if isinstance(symdef, VarDef): return symdef.datatype elif isinstance(targetnode, TargetRegisters): if len(targetnode.nodes) == 1: return datatype_of(targetnode.nodes[0], scope) raise TypeError("cannot determine datatype", targetnode) def coerce_constant_value(datatype: DataType, value: Expression, sourceref: SourceRef=None) -> Tuple[bool, Expression]: # if we're a BYTE type, and the value is a single character, convert it to the numeric value assert isinstance(value, Expression) def verify_bounds(pvalue: Union[int, float, str]) -> None: # if the value is out of bounds, raise an overflow exception if isinstance(pvalue, (int, float)): if datatype == DataType.BYTE and not (0 <= pvalue <= 0xff): # type: ignore raise OverflowError("value out of range for byte: " + str(pvalue)) if datatype == DataType.WORD and not (0 <= pvalue <= 0xffff): # type: ignore raise OverflowError("value out of range for word: " + str(pvalue)) if datatype == DataType.FLOAT and not (FLOAT_MAX_NEGATIVE <= pvalue <= FLOAT_MAX_POSITIVE): # type: ignore raise OverflowError("value out of range for float: " + str(pvalue)) if isinstance(value, LiteralValue): if type(value.value) is str and len(value.value) == 1 and (datatype.isnumeric() or datatype.isarray()): # convert a string of length 1 to its numeric character value lv = LiteralValue(value=char_to_bytevalue(value.value), sourceref=value.sourceref) # type: ignore lv.parent = value.parent return True, lv # if we're an integer value and the passed value is float, truncate it (and give a warning) if datatype in (DataType.BYTE, DataType.WORD, DataType.MATRIX) and isinstance(value.value, float): frac = math.modf(value.value) if frac != 0: print_warning("float value truncated ({} to datatype {})".format(value.value, datatype.name), sourceref=sourceref) v2 = int(value.value) verify_bounds(v2) lv = LiteralValue(value=v2, sourceref=value.sourceref) # type: ignore lv.parent = value.parent return True, lv if type(value.value) in (int, float): verify_bounds(value.value) if datatype == DataType.WORD: if type(value.value) not in (int, float, str): raise TypeError("cannot assign '{:s}' to {:s}".format(type(value.value).__name__, datatype.name.lower()), sourceref) elif datatype in (DataType.BYTE, DataType.WORD, DataType.FLOAT): if type(value.value) not in (int, float): raise TypeError("cannot assign '{:s}' to {:s}".format(type(value.value).__name__, datatype.name.lower()), sourceref) elif isinstance(value, (ExpressionWithOperator, SubCall)): return False, value elif isinstance(value, SymbolName): symboldef = value.my_scope().lookup(value.name) if isinstance(symboldef, VarDef) and symboldef.vartype == VarType.CONST: return True, symboldef.value elif isinstance(value, AddressOf): try: address = value.const_value() lv = LiteralValue(value=address, sourceref=value.sourceref) # type: ignore lv.parent = value.parent return True, lv except TypeError: return False, value if datatype == DataType.WORD and not isinstance(value, (LiteralValue, Dereference, Register, SymbolName, AddressOf)): raise TypeError("cannot assign '{:s}' to {:s}".format(type(value).__name__, datatype.name.lower()), sourceref) elif datatype in (DataType.BYTE, DataType.WORD, DataType.FLOAT) \ and not isinstance(value, (LiteralValue, Dereference, Register, SymbolName, AddressOf)): raise TypeError("cannot assign '{:s}' to {:s}".format(type(value).__name__, datatype.name.lower()), sourceref) return False, value def check_symbol_definition(name: str, scope: Scope, sref: SourceRef) -> Any: try: return scope.lookup(name) except UndefinedSymbolError as x: raise ParseError(str(x), sref) # ----------------- PLY parser definition follows ---------------------- def p_start(p): """ start : empty | module_elements """ if p[1]: scope = Scope(nodes=p[1], level="module", sourceref=_token_sref(p, 1)) scope.name = "<" + p.lexer.source_filename + " global scope>" p[0] = Module(name=p.lexer.source_filename, sourceref=_token_sref(p, 1)) p[0].nodes.append(scope) else: scope = Scope(nodes=[], level="module", sourceref=_token_sref(p, 1)) scope.name = "<" + p.lexer.source_filename + " global scope>" p[0] = Module(name=p.lexer.source_filename, sourceref=SourceRef(lexer.source_filename, 1, 1)) p[0].nodes.append(scope) def p_module(p): """ module_elements : module_elt | module_elements module_elt """ if len(p) == 2: if p[1] is None: p[0] = [] else: p[0] = [p[1]] else: if p[2] is None: p[0] = p[1] else: p[0] = p[1] + [p[2]] def p_module_elt(p): """ module_elt : ENDL | directive | block """ if p[1] != '\n': p[0] = p[1] def p_directive(p): """ directive : DIRECTIVE ENDL | DIRECTIVE directive_args ENDL """ if len(p) == 3: p[0] = Directive(name=p[1], sourceref=_token_sref(p, 1)) else: p[0] = Directive(name=p[1], args=p[2], sourceref=_token_sref(p, 1)) def p_directive_args(p): """ directive_args : directive_arg | directive_args ',' directive_arg """ if len(p) == 2: p[0] = [p[1]] else: p[0] = p[1] + [p[3]] def p_directive_arg(p): """ directive_arg : NAME | INTEGER | STRING | BOOLEAN """ p[0] = p[1] def p_block_name_addr(p): """ block : BITINVERT NAME INTEGER endl_opt scope """ p[0] = Block(name=p[2], address=p[3], sourceref=_token_sref(p, 2)) p[0].scope = p[5] def p_block_name(p): """ block : BITINVERT NAME endl_opt scope """ p[0] = Block(name=p[2], sourceref=_token_sref(p, 2)) p[0].scope = p[4] def p_block(p): """ block : BITINVERT endl_opt scope """ p[0] = Block(sourceref=_token_sref(p, 1)) p[0].scope = p[3] def p_endl_opt(p): """ endl_opt : empty | ENDL """ pass def p_scope(p): """ scope : '{' scope_elements_opt '}' """ p[0] = Scope(nodes=p[2] or [], level="block", sourceref=_token_sref(p, 1)) def p_scope_elements_opt(p): """ scope_elements_opt : empty | scope_elements """ p[0] = p[1] def p_scope_elements(p): """ scope_elements : scope_element | scope_elements scope_element """ if len(p) == 2: p[0] = [] if p[1] in (None, '\n') else [p[1]] else: if p[2] in (None, '\n'): p[0] = p[1] else: p[0] = p[1] + [p[2]] def p_scope_element(p): """ scope_element : ENDL | label | directive | vardef | subroutine | inlineasm | statement """ if p[1] != '\n': p[0] = p[1] else: p[0] = None def p_label(p): """ label : LABEL """ p[0] = Label(name=p[1], sourceref=_token_sref(p, 1)) def p_inlineasm(p): """ inlineasm : INLINEASM ENDL """ p[0] = InlineAssembly(assembly=p[1], sourceref=_token_sref(p, 1)) def p_vardef(p): """ vardef : VARTYPE type_opt NAME ENDL """ p[0] = VarDef(name=p[3], vartype=p[1], datatype=p[2], sourceref=_token_sref(p, 3)) def p_vardef_value(p): """ vardef : VARTYPE type_opt NAME IS expression """ p[0] = VarDef(name=p[3], vartype=p[1], datatype=p[2], sourceref=_token_sref(p, 3)) p[0].value = p[5] def p_type_opt(p): """ type_opt : DATATYPE '(' dimensions ')' | DATATYPE | empty """ if len(p) == 5: p[0] = DatatypeNode(name=p[1], dimensions=p[3], sourceref=_token_sref(p, 1)) elif len(p) == 2 and p[1]: p[0] = DatatypeNode(name=p[1], sourceref=_token_sref(p, 1)) def p_dimensions(p): """ dimensions : INTEGER | dimensions ',' INTEGER """ if len(p) == 2: p[0] = [p[1]] else: p[0] = p[1] + [p[3]] def p_literal_value(p): """literal_value : INTEGER | FLOATINGPOINT | STRING | CHARACTER | BOOLEAN""" tok = p.slice[-1] if tok.type == "CHARACTER": p[1] = char_to_bytevalue(p[1]) # character literals are converted to byte value. elif tok.type == "BOOLEAN": p[1] = int(p[1]) # boolean literals are converted to integer form (true=1, false=0). p[0] = LiteralValue(value=p[1], sourceref=_token_sref(p, 1)) def p_subroutine(p): """ subroutine : SUB NAME '(' sub_param_spec ')' RARROW '(' sub_result_spec ')' subroutine_body ENDL """ body = p[10] if isinstance(body, Scope): p[0] = Subroutine(name=p[2], param_spec=p[4] or [], result_spec=p[8] or [], sourceref=_token_sref(p, 1)) p[0].scope = body elif type(body) is int: p[0] = Subroutine(name=p[2], param_spec=p[4] or [], result_spec=p[8] or [], address=body, sourceref=_token_sref(p, 1)) else: raise TypeError("subroutine_body", p.slice) def p_sub_param_spec(p): """ sub_param_spec : empty | sub_param_list """ p[0] = p[1] def p_sub_param_list(p): """ sub_param_list : sub_param | sub_param_list ',' sub_param """ if len(p) == 2: p[0] = [p[1]] else: p[0] = p[1] + [p[3]] def p_sub_param(p): """ sub_param : LABEL REGISTER | REGISTER """ if len(p) == 3: p[0] = (p[1], p[2]) elif len(p) == 2: p[0] = (None, p[1]) def p_sub_result_spec(p): """ sub_result_spec : empty | '?' | sub_result_list """ if p[1] == '?': p[0] = ['A', 'X', 'Y'] # '?' means: all registers clobbered else: p[0] = p[1] def p_sub_result_list(p): """ sub_result_list : sub_result_reg | sub_result_list ',' sub_result_reg """ if len(p) == 2: p[0] = [p[1]] else: p[0] = p[1] + [p[3]] def p_sub_result_reg(p): """ sub_result_reg : REGISTER | CLOBBEREDREGISTER """ p[0] = p[1] def p_subroutine_body(p): """ subroutine_body : scope | IS INTEGER """ if len(p) == 2: p[0] = p[1] else: p[0] = p[2] def p_statement(p): """ statement : assignment ENDL | aug_assignment ENDL | subroutine_call ENDL | goto ENDL | conditional_goto ENDL | incrdecr ENDL | return ENDL """ p[0] = p[1] def p_incrdecr(p): """ incrdecr : assignment_target INCR | assignment_target DECR """ p[0] = IncrDecr(operator=p[2], sourceref=_token_sref(p, 2)) p[0].target = p[1] def p_call_subroutine(p): """ subroutine_call : calltarget preserveregs_opt '(' call_arguments_opt ')' """ sref = _token_sref(p, 3) p[0] = SubCall(sourceref=sref) target = p[1] if isinstance(target, int): target = LiteralValue(value=target, sourceref=sref) p[0].nodes.append(target) p[0].nodes.append(p[2] or PreserveRegs(registers="", sourceref=sref)) p[0].nodes.append(CallArguments(nodes=p[4] or [], sourceref=sref)) def p_preserveregs_opt(p): """ preserveregs_opt : empty | preserveregs """ p[0] = p[1] def p_preserveregs(p): """ preserveregs : PRESERVEREGS """ p[0] = PreserveRegs(registers=p[1], sourceref=_token_sref(p, 1)) def p_call_arguments_opt(p): """ call_arguments_opt : empty | call_arguments """ p[0] = p[1] def p_call_arguments(p): """ call_arguments : call_argument | call_arguments ',' call_argument """ if len(p) == 2: p[0] = [p[1]] else: p[0] = p[1] + [p[3]] def p_call_argument(p): """ call_argument : expression | register IS expression | NAME IS expression """ if len(p) == 2: p[0] = CallArgument(sourceref=_token_sref(p, 1)) p[0].nodes.append(p[1]) elif len(p) == 4: if isinstance(p[1], AstNode): sref = p[1].sourceref else: sref = _token_sref(p, 2) p[0] = CallArgument(name=p[1], sourceref=sref) p[0].nodes.append(p[3]) def p_return(p): """ return : RETURN | RETURN expression | RETURN expression ',' expression | RETURN expression ',' expression ',' expression """ if len(p) == 2: p[0] = Return(sourceref=_token_sref(p, 1)) elif len(p) == 3: p[0] = Return(sourceref=_token_sref(p, 1)) p[0].nodes.append(p[2]) # A elif len(p) == 5: p[0] = Return(sourceref=_token_sref(p, 1)) p[0].nodes.append(p[2]) # A p[0].nodes.append(p[4]) # X elif len(p) == 7: p[0] = Return(sourceref=_token_sref(p, 1)) p[0].nodes.append(p[2]) # A p[0].nodes.append(p[4]) # X p[0].nodes.append(p[6]) # Y def p_register(p): """ register : REGISTER """ p[0] = Register(name=p[1], sourceref=_token_sref(p, 1)) def p_goto(p): """ goto : GOTO calltarget """ p[0] = Goto(sourceref=_token_sref(p, 1)) target = p[2] if isinstance(target, int): target = LiteralValue(value=target, sourceref=p[0].sourceref) p[0].nodes.append(target) def p_conditional_goto_plain(p): """ conditional_goto : IF GOTO calltarget """ p[0] = Goto(if_stmt=p[1], sourceref=_token_sref(p, 1)) p[0].nodes.append(p[3]) def p_conditional_goto_expr(p): """ conditional_goto : IF expression GOTO calltarget """ p[0] = Goto(if_stmt=p[1], sourceref=_token_sref(p, 1)) p[0].nodes.append(p[4]) p[0].nodes.append(p[2]) def p_calltarget(p): """ calltarget : symbolname | INTEGER | dereference """ p[0] = p[1] def p_dereference(p): """ dereference : '[' dereference_operand ']' """ p[0] = Dereference(datatype=p[2][1], sourceref=_token_sref(p, 1)) operand = p[2][0] if isinstance(operand, int): p[0].nodes.append(LiteralValue(value=operand, sourceref=p[0].sourceref)) elif isinstance(operand, str): p[0].nodes.append(Register(name=operand, sourceref=p[0].sourceref)) elif isinstance(operand, SymbolName): p[0].nodes.append(operand) attr.validate(p[0]) def p_dereference_operand(p): """ dereference_operand : symbolname type_opt | REGISTER type_opt | INTEGER type_opt """ p[0] = (p[1], p[2]) def p_symbolname(p): """ symbolname : NAME | DOTTEDNAME """ p[0] = SymbolName(name=p[1], sourceref=_token_sref(p, 1)) def p_assignment(p): """ assignment : assignment_target IS expression | assignment_target IS assignment """ p[0] = Assignment(sourceref=_token_sref(p, 2)) p[0].nodes.append(AssignmentTargets(nodes=[p[1]], sourceref=p[0].sourceref)) p[0].nodes.append(p[3]) def p_aug_assignment(p): """ aug_assignment : assignment_target AUGASSIGN expression """ p[0] = AugAssignment(operator=p[2], sourceref=_token_sref(p, 2)) p[0].nodes.append(p[1]) p[0].nodes.append(p[3]) precedence = ( # following the python operator precedence rules mostly; https://docs.python.org/3/reference/expressions.html#operator-precedence ('left', 'LOGICOR'), ('left', 'LOGICAND'), ('right', 'LOGICNOT'), ('left', "LT", "GT", "LE", "GE", "EQUALS", "NOTEQUALS"), ('left', 'BITOR'), ('left', 'BITXOR'), ('left', 'BITAND'), ('left', 'SHIFTLEFT', 'SHIFTRIGHT'), ('left', '+', '-'), ('left', '*', '/', 'INTEGERDIVIDE', 'MODULO'), ('right', 'UNARY_MINUS', 'BITINVERT', "UNARY_ADDRESSOF"), ('left', 'POWER'), ('nonassoc', "COMMENT"), ) def p_expression(p): """ expression : expression '+' expression | expression '-' expression | expression '*' expression | expression '/' expression | expression MODULO expression | expression BITOR expression | expression BITXOR expression | expression BITAND expression | expression SHIFTLEFT expression | expression SHIFTRIGHT expression | expression LOGICOR expression | expression LOGICAND expression | expression POWER expression | expression INTEGERDIVIDE expression | expression LT expression | expression GT expression | expression LE expression | expression GE expression | expression EQUALS expression | expression NOTEQUALS expression """ p[0] = ExpressionWithOperator(operator=p[2], sourceref=_token_sref(p, 2)) p[0].nodes.append(p[1]) p[0].nodes.append(p[3]) def p_expression_uminus(p): """ expression : '-' expression %prec UNARY_MINUS """ p[0] = ExpressionWithOperator(operator=p[1], sourceref=_token_sref(p, 1)) p[0].nodes.append(p[2]) def p_expression_addressof(p): """ expression : BITAND symbolname %prec UNARY_ADDRESSOF """ p[0] = AddressOf(name=p[2].name, sourceref=_token_sref(p, 1)) def p_unary_expression_bitinvert(p): """ expression : BITINVERT expression """ p[0] = ExpressionWithOperator(operator=p[1], sourceref=_token_sref(p, 1)) p[0].nodes.append(p[2]) def p_unary_expression_logicnot(p): """ expression : LOGICNOT expression """ p[0] = ExpressionWithOperator(operator=p[1], sourceref=_token_sref(p, 1)) p[0].nodes.append(p[2]) def p_expression_group(p): """ expression : '(' expression ')' """ p[0] = p[2] def p_expression_expr_value(p): """expression : expression_value""" p[0] = p[1] def p_expression_value(p): """ expression_value : literal_value | symbolname | register | subroutine_call | dereference """ p[0] = p[1] def p_assignment_target(p): """ assignment_target : target_registers | symbolname | dereference """ if isinstance(p[1], TargetRegisters): # if the target registers is just a single register, use that instead if len(p[1].nodes) == 1: assert isinstance(p[1].nodes[0], Register) p[1] = p[1].nodes[0] p[0] = p[1] def p_target_registers(p): """ target_registers : register | target_registers ',' register """ if len(p) == 2: p[0] = TargetRegisters(sourceref=_token_sref(p, 1)) p[0].nodes.append(p[1]) else: p[1].nodes.append(p[3]) p[0] = p[1] def p_empty(p): """empty :""" pass def p_error(p): stack_state_str = ' '.join([symbol.type for symbol in parser.symstack][1:]) print('\n[ERROR DEBUG: parser state={:d} stack: {} . {} ]'.format(parser.state, stack_state_str, p)) if p: sref = SourceRef(p.lexer.source_filename, p.lineno, find_tok_column(p)) if p.value in ("", "\n"): p.lexer.error_function(sref, "syntax error before end of line") else: p.lexer.error_function(sref, "syntax error before or at '{:.20s}'", str(p.value).rstrip()) else: lexer.error_function(None, "syntax error at end of input", lexer.source_filename) def _token_sref(p, token_idx): """ Returns the coordinates for the YaccProduction object 'p' indexed with 'token_idx'. The coordinate includes the 'lineno' and 'column', starting from 1. """ last_cr = p.lexer.lexdata.rfind('\n', 0, p.lexpos(token_idx)) if last_cr < 0: last_cr = -1 chunk = p.lexer.lexdata[last_cr:p.lexpos(token_idx)] column = len(chunk.expandtabs()) return SourceRef(p.lexer.source_filename, p.lineno(token_idx), column) class TokenFilter: def __init__(self, lexer): self.lexer = lexer self.prev_was_EOL = False assert "ENDL" in tokens def token(self): # make sure we only ever emit ONE "ENDL" token in sequence if self.prev_was_EOL: # skip all EOLS that might follow while True: tok = self.lexer.token() if not tok or tok.type != "ENDL": break self.prev_was_EOL = False else: tok = self.lexer.token() self.prev_was_EOL = tok and tok.type == "ENDL" return tok parser = yacc(write_tables=True) def connect_parents(node: AstNode, parent: AstNode) -> None: node.parent = parent for childnode in node.nodes: if isinstance(childnode, AstNode): connect_parents(childnode, node) def parse_file(filename: str, lexer_error_func=None) -> Module: lexer.error_function = lexer_error_func lexer.lineno = 1 lexer.source_filename = filename tfilter = TokenFilter(lexer) with open(filename, "rU") as inf: sourcecode = inf.read() result = parser.parse(input=sourcecode, tokenfunc=tfilter.token) connect_parents(result, None) return result