def compileClass(self):
self._process_token('class')
self.classname = self._process_ttype(IDENTIFIER)
self._process_token('{')
# open symbol table
self.st = VMSymbolTable(self.classname)
token = self._get_token()
while token != '}':
if token in ('field', 'static'):
self.compileClassVarDec()
elif token in ('constructor', 'method', 'function'):
self.compileSubroutine()
else:
raise Exception("wtf?", token)
token = self._get_token()
self._process_token("}")
class CompilationEngine(object):
""" Effects the actual compilation output. Gets its input from a
JackTokenizer and emits its parsed structure into an output file/stream. The
output is generated by a series of compilexxx() routines, one for every
syntactic element xxx of the Jack grammar. The contract between these
routines is that each compilexxx() routine should read the syntactic
construct xxx from the input, advance() the tokenizer exactly beyond xxx,
and output the parsing of xxx. Thus, compilexxx() may only be called if
indeed xxx is the next syntactic element of the input. In the first version
of the compiler, described in Chapter 10, this module emits a structured
printout of the code, wrapped in XML tags. In the final version of the
compiler, described in Chapter 11, this module generates executable VM code.
In both cases, the parsing logic and module API are exactly the same.
"""
_get_token = lambda self: self.tknz.current_token
_get_ttype = lambda self: self.tknz.token_type
_get_nttype = lambda self: TYPES[self._get_ttype()]
def _process_ttype(self, *ttype_options):
""" processes a token based on its type. returns token. """
saved = self._get_token()
print(GREENON+saved+RESETCOLOR)
assert self._get_ttype() in ttype_options, \
r"unexpected token type: " \
r"%s should be in %s" % (TYPES[self._get_ttype()],
[TYPES[e] for e in ttype_options])
self.tknz.advance()
return saved
def _process_token(self, *token_options):
""" processes a token based on its format. returns token. """
saved = self._get_token()
print(GREENON+saved+RESETCOLOR)
if token_options != (True,):
assert self._get_token() in token_options, \
r"unexpected token: " \
r"'%s' should be in '%s'" % (self._get_token(), token_options)
self.tknz.advance()
return saved
def __init__(self, tokenizer, vmwriter, classname=None):
self.tknz = tokenizer
self.vmw = vmwriter
self.st = None
self.classname = None
self.tknz.advance()
self.compileClass()
self.vmw.close()
def compileClass(self):
self._process_token('class')
self.classname = self._process_ttype(IDENTIFIER)
self._process_token('{')
# open symbol table
self.st = VMSymbolTable(self.classname)
token = self._get_token()
while token != '}':
if token in ('field', 'static'):
self.compileClassVarDec()
elif token in ('constructor', 'method', 'function'):
self.compileSubroutine()
else:
raise Exception("wtf?", token)
token = self._get_token()
self._process_token("}")
def compileClassVarDec(self):
var_kind = self._process_token("static", "field")
var_type = self._process_ttype(KEYWORD, IDENTIFIER)
variables = []
variables.append(self._process_ttype(IDENTIFIER))
while self._get_token() == ',':
self._process_token(',')
variables.append(self._process_ttype(IDENTIFIER))
self._process_token(";")
scope = self.st.parent
for var_name in variables:
scope.addVar(var_name, var_type, var_kind)
print(scope)
def compileSubroutine(self):
subr_type = self._process_token("method", "constructor", "function")
retr_type = self._process_ttype(KEYWORD, IDENTIFIER)
subr_name = self._process_ttype(IDENTIFIER)
self.whilecount = 0
self.ifcount = 0
self.st.createScope(subr_name, subr_type)
self.current_subr = subr_name
self._process_token("(")
self.compileParameterList()
self._process_token(")")
self._process_token('{')
while self._get_token() == 'var':
self.compileVarDec()
self.vmw.write("function %s.%s %s" % (self.classname, self.current_subr,
self.st.getScope(self.current_subr).countKind('local')))
if subr_type == 'constructor':
class_vars_count = self.st.parent.countKind('field')
self.vmw.write("push constant %s" % class_vars_count)
self.vmw.write("call Memory.alloc 1")
self.vmw.write("pop pointer 0")
if subr_type == 'method':
self.vmw.write("push argument 0")
self.vmw.write("pop pointer 0")
if self._get_token() != '}':
self.compileStatements()
self._process_token("}")
def compileStatements(self):
while self._get_token() != '}':
if self._get_token() == 'do':
self.compileDo()
elif self._get_token() == 'if':
self.compileIf()
elif self._get_token() == 'let':
self.compileLet()
elif self._get_token() == 'while':
self.compileWhile()
elif self._get_token() == 'return':
self.compileReturn()
else:
raise Exception("invalid statement:", self._get_token())
def compileVarDec(self):
self._process_token("var")
var_type = self._process_ttype(KEYWORD, IDENTIFIER)
variables = []
while True:
variables.append(self._process_ttype(IDENTIFIER))
if self._get_token() != ',':
break
self._process_token(",")
self._process_token(";")
scope = self.st.getScope(self.current_subr)
for var_name in variables:
scope.addVar(var_name, var_type, 'local')
print(scope)
def compileParameterList(self):
# it might be empty
if self._get_token() == ')':
return
variables = []
while True:
var_type = self._process_ttype(KEYWORD, IDENTIFIER)
var_name = self._process_ttype(IDENTIFIER)
variables.append((var_name, var_type))
if self._get_token() != ',':
break
self._process_token(',')
scope = self.st.getScope(self.current_subr)
for (var_name, var_type) in variables:
scope.addVar(var_name, var_type, 'argument')
print(scope)
def compileDo(self):
# TRANSFORM THIS INTO A COMPILE CALL FUNCTION????
"""
possible calls:
do a.b(); // a is a var => b is a method
do c.b(); // c is a class => b isn't a method
do m(); // m is a method to this class
"""
self._process_token('do')
a = b = None
a = self._process_ttype(IDENTIFIER)
# it's an a.b() call
if self._get_token() == '.':
self._process_token('.')
b = self._process_ttype(IDENTIFIER)
else: pass # it's an m() call
args_count = 0
# check if it's a method call
if not b: # it's a call to an inner method
self.vmw.write("push pointer 0")
args_count += 1
b = a
a = self.classname
call_class = a
elif self.st.isAccessible(self.current_subr, a): # a is a variable
var_add = self.st.getAddress(self.current_subr, a)
self.vmw.write("push %s %s" % var_add) # push object
args_count += 1
call_class = self.st.getVar(self.current_subr, a)['type']
else: # not a method call
call_class = a
self._process_token("(")
args_count += self.compileExpressionList()
self._process_token(")")
self._process_token(";")
self.vmw.write("call %s.%s %s" % (call_class, b, args_count))
self.vmw.write("pop temp 0") # throw out return value
def compileExpression(self):
operators = {'+': 'add', '-': 'sub',
'*': 'call Math.multiply 2', '/': 'call Math.divide 2',
'&': 'and', '|': 'or', '<': 'lt', '>': 'gt', '=': 'eq'}
self.compileTerm()
while self._get_token() in operators:
token = self._get_token()
self._process_token(token)
self.compileTerm()
self.vmw.write(operators[token])
def compileTerm(self):
ttype = self._get_ttype()
token = self._get_token()
if ttype == INT_CONST:
self.vmw.write("push constant %s" % token)
self._process_ttype(INT_CONST)
elif ttype == STRING_CONST:
str_size = len(token)-2
self.vmw.write("push constant %s" % str_size)
self.vmw.write("call String.new 1")
for char in token[1:-1]:
self.vmw.write("push constant %s" % ord(char))
self.vmw.write("call String.appendChar 2")
self._process_ttype(STRING_CONST)
elif token in ('true', 'false', 'null'):
self.vmw.write("push constant 0")
if token == 'true':
self.vmw.write("not")
self._process_token("true", "false", "null")
elif token == 'this':
self.vmw.write("push pointer 0")
self._process_token("this")
elif ttype == IDENTIFIER:
a = self._process_ttype(IDENTIFIER)
# it's a function call
# this section follows pattern of compileDo
if self._get_token() in ('.', '('):
if self._get_token() == '.':
# its an a.b() call
self._process_token('.')
b = self._process_ttype(IDENTIFIER)
else: pass # it's an a() call
args_count = 0
# check if it's a method call
if not b: # it's a call to an inner method
self.vmw.write("push pointer 0")
args_count += 1
b = a
a = self.classname
call_class = a
elif self.st.isAccessible(self.current_subr, a): # a is a var
var_add = self.st.getAddress(self.current_subr, a)
self.vmw.write("push %s %s" % var_add) # push object
args_count += 1
call_class = self.st.getVar(self.current_subr, a)['type']
else: # not a method call
call_class = a
self._process_token("(")
args_count += self.compileExpressionList()
self._process_token(")")
self.vmw.write("call %s.%s %s" % (call_class, b, args_count))
elif self._get_token() == '[':
# process a[]
assert self.st.isAccessible(self.current_subr, a)
var_add = self.st.getAddress(self.current_subr, a)
self._process_token("[")
self.compileExpression()
self._process_token("]")
self.vmw.write("push %s %s" % var_add)
self.vmw.write("add")
self.vmw.write("pop pointer 1")
self.vmw.write("push that 0")
else:
assert self.st.isAccessible(self.current_subr, a)
var_add = self.st.getAddress(self.current_subr, a)
self.vmw.write("push %s %s" % var_add)
elif token == '(': # '(' expression ')'
self._process_token('(')
self.compileExpression()
self._process_token(')') # already done above
elif token in ('-', '~'): # unary opeartion term
if token == '-': op = 'neg'
elif token == '~': op = 'not'
self._process_token("-", "~")
self.compileTerm()
self.vmw.write(op)
else:
raise Exception("aqui! termo nao identificado", token)
def compileExpressionList(self):
exps = 0
while self._get_token() != ')':
self.compileExpression()
exps += 1
if self._get_token() == ',':
self._process_token(',')
return exps
def compileReturn(self):
self._process_token("return")
if self._get_token() != ';':
self.compileExpression()
else:
self.vmw.write("push constant 0")
self._process_token(";")
self.vmw.write("return")
def compileLet(self):
self._process_token("let")
var_name = self._process_ttype(IDENTIFIER)
assert self.st.isAccessible(self.current_subr, var_name)
var_add = self.st.getAddress(self.current_subr, var_name)
if self._get_token() == '[':
self._process_token("[")
self.compileExpression() # push k
self._process_token("]")
self.vmw.write("push %s %s" % var_add)
self.vmw.write("add")
self._process_token("=")
self.compileExpression()
self.vmw.write("pop temp 0")
self.vmw.write("pop pointer 1")
self.vmw.write("push temp 0")
self.vmw.write("pop that 0")
else:
self._process_token("=")
self.compileExpression()
self.vmw.write("pop %s %s" % var_add)
self._process_token(";")
def compileWhile(self):
thiswhilecount = self.whilecount
self.whilecount += 1
self._process_token('while')
self.vmw.write("label WHILE_EXP%s" % thiswhilecount)
self._process_token("(")
self.compileExpression()
self._process_token(")")
self._process_token("{")
self.vmw.write("not")
self.vmw.write("if-goto WHILE_END%s" % thiswhilecount)
self.compileStatements()
self.vmw.write("goto WHILE_EXP%s" % thiswhilecount)
self.vmw.write("label WHILE_END%s" % thiswhilecount)
self._process_token("}")
def compileIf(self):
thisifcount = self.ifcount
self.ifcount += 1
self._process_token('if')
self._process_token("(")
self.compileExpression()
self._process_token(")")
self.vmw.write("if-goto IF_TRUE%s" % thisifcount)
self.vmw.write("goto IF_FALSE%s" % thisifcount)
self.vmw.write("label IF_TRUE%s" % thisifcount)
self._process_token("{")
self.compileStatements()
self._process_token("}")
if self._get_token() == 'else':
self.vmw.write("goto IF_END%s" % thisifcount)
self.vmw.write("label IF_FALSE%s" % thisifcount)
self._process_token('else')
self._process_token("{")
self.compileStatements()
self._process_token("}")
self.vmw.write("label IF_END%s" % thisifcount)
else:
self.vmw.write("label IF_FALSE%s" % thisifcount)