def compile_whileStatement(): # while ( expression ) { statements }
print("---COMPILING WHILE STATEMENT---")
global global_while_index
global_while_index += 1
while_index = global_while_index # local var used; ensure while_index consistent throughout this function
# global_while_index may be changed by other functions without you knowing
VMWriter.writeLabel("WHILE_EXP" + str(while_index)) # label WHILE_1
advance() # (
compile_expression() # expression
VMWriter.writeArithmetic("not") # NOT expression
advance() # )
advance() # {
VMWriter.writeIfGoto("WHILE_END" + str(while_index)) # if-goto WHILE_END_1
compile_statements()
VMWriter.writeGoto("WHILE_EXP" + str(while_index)) # goto WHILE_1
VMWriter.writeLabel("WHILE_END" + str(while_index))
advance() # }
print("---DONE COMPILING WHILE STATEMENT---")
def __init__(self, tokenizer):
"""
Ctor
_writer: a VM writer object
_cur_class: the current class
_cur_func: the current function
_cur_func_type: the current function type
_subroutine_table: a symbol table of the subroutine
_class_table: a list of the symbols in the class
_tokenizer: the tokenizer object
_result: a list that is written in the output file
counters: the kind counters
label_counters: the label counters
"""
self._writer = VMWriter()
self._cur_class = None
self._cur_func = None
self._cur_func_type = None
self._subroutine_table = None
self._class_table = []
self._tokenizer = tokenizer
self._result = []
self.counters = {'static': 0, 'field': 0, 'argument': 0, 'local': 0}
self.label_counters = {"if": 0, "while": 0}
self.compileClass()
def compileSubroutine(self):
self.SubroutineScope.reset()
ftype = self.takeWord() # (constructor|method|function)
self.advance() # type
subroutineName = self.takeWord() # SubroutineName
self.advance() # (
self.complieParameterList()
self.advance() # )
self.advance() # {
num_locals = 0
while self.tokens[self.count]["token"] in ["var"]:
num_locals += self.compileVarDec()
self.vmcode += VMWriter.writeFunction(
self.className + "." + subroutineName, num_locals)
# if function is constructor, it should allocate memory for produced object
if ftype == "constructor":
# allocate needed memory
num_field = self.ClassScope.varCount("field")
self.vmcode += VMWriter.writePush("constant", num_field)
self.vmcode += VMWriter.writeCall("Memory.alloc", 1)
# link it to This
self.vmcode += VMWriter.writePop("pointer", 0)
# change all FIELD in ClassScope to THIS
self.ClassScope.field2This()
self.compileStatements()
self.advance() # }
self.label_count["if"] = 0
self.label_count["while"] = 0
def __init__(self, file):
self.lex = Lex(file)
self.symbols = SymbolTable()
self.vm = VMWriter()
self.openout(file)
self.compile_class()
self.closeout()
def openXMLFile(self, xml_file):
self.xml_tree = ET.parse(xml_file)
self.tokens = list(self.xml_tree.getroot())
self.tokens.reverse()
self.st = SymbolTable()
self.current_class = xml_file[:-5]
self.vm = VMWriter(self.current_class + '.vm')
def compileDo(self):
self.advance() # do
caller = self.takeWord() # (className|varName)
symbol = self.findSymbol(caller)
# if object exist, push it in to stack
if symbol:
self.advance() # .
func = self.takeWord() # subroutineName
segment = 'local'
index = symbol['index']
self.vmcode += VMWriter.writePush(segment, index)
symbolType = symbol['type']
# Static method called
elif self.tokens[self.count]["token"] == ".":
symbolType = caller
self.advance() # .
func = self.takeWord() # subroutineName
else:
symbolType = self.className
func = caller
# decide subroutine name
subroutineName = symbolType + '.' + func
self.advance() # (
num_args = self.compileExpressionList()
if symbol:
# add "this"
num_args += 1
# call the function
self.vmcode += VMWriter.writeCall(subroutineName, num_args)
# since this code will not save returned value in to variable, pop it out.
self.vmcode += VMWriter.writePop('temp', '0')
self.advance() # )
self.advance() # ;
def compile_doStatement(): # do subroutineCall ;
print("---COMPILING DO STATEMENT---")
compile_subroutineCall() # subroutineCall
VMWriter.writePop("temp", 0)
advance()
print("---DONE COMPILING DO STATEMENT---")
def __init__(self, tokens, file_path):
self.tokens = tokens
self.class_symboltable = SymbolTable()
self.subroutine_symboltable = SymbolTable()
self.cur_class = None
self.cur_subroutine_type = None # for the compile return function
self.vmwriter = VMWriter(file_path)
self.labelcounter = 0
def __init__(self, inputFile, outputFile):
# self.outFile = outputFile
self.tokenizer = JackTokenizer(inputFile)
self.vmwriter = VMWriter(outputFile)
self.outFile = outputFile
self.currToken = ""
self.tabber = ""
self.argumentsCounter = 0
def __init__(self, input_file, xml_file, vm_file):
''' Initialize the xml file, the symbol table adn the VMwriter '''
super().__init__(input_file)
self.xml_file = xml_file
self.symbol_tab = SymbolTable()
self.vm_writer = VMWriter(vm_file)
self.subroutine_type = None
self.if_label, self.while_label = 0, 0
self.isarray = None
def __init__(self, tokenizer, outputFile):
self.XMLCode = []
self.CodeIndent = 0
self.tokenizer = tokenizer
self.symbolTable = SymbolTable()
self.vmWriter = VMWriter(outputFile)
self.class_name = None
self.segment_local_dict = segment_dict
self.while_count = 0
self.if_count = 0
def compileString(self, s):
# use standard library String
str_len = len(s)
self.vmcode += VMWriter.writePush("constant", str_len)
self.vmcode += VMWriter.writeCall("String.new", 1)
for c in s:
if not c == "\"":
asciic = ord(c)
self.vmcode += VMWriter.writePush("constant", asciic)
self.vmcode += VMWriter.writeCall("String.appendChar", 2)
def compileReturn(self):
self.advance() # return
if not self.tokens[self.count]["token"] == ";":
# push result to stack
self.compileExpression()
else:
# push constant 0 to stack for void
self.vmcode += VMWriter.writePush("constant", "0")
self.advance() # ;
self.vmcode += VMWriter.writeReturn()
def __init__(self, tokenizer, output):
"""Creates a new compilation engine with the given input and output.
The next routine called must be compileClass(). """
self.tokenizer = tokenizer
self.tags = []
self.scope = 0
self.writer = VMWriter(output)
self.class_table = SymbolTable()
self.method_table = SymbolTable()
self.class_name = ''
self.method_or_constructor = False
def __init__(self, file):
""" """
self.label_num = 0
self.tokenizer = JackTokenizer(file)
self.advance()
self.symbols = SymbolTable()
self.vm = VMWriter()
self.open_outfile(file)
self.compile_class()
self.close_outfile()
def __init__(self, input_path, output_path):
"""
creates a new compilation engine with the given input and output. the next routine called must be compileClass()
:param input_path: input stream/file
:param output_path: output stream/file
"""
self.labels = 0
self.jack_class = None
self.class_subroutine = None
self.tokenizer = JackTokenizer(input_path)
self._writer = VMWriter(output_path)
self.CompileClass()
def __init__(self, in_file, out_file):
"""
A compilation engine constructor
:param in_file: the file we are currently compiling
:param out_file: the file where we save the output
"""
self._tokenizer = JackTokenizer(in_file)
self._class_table = SymbolTable()
self._method_table = SymbolTable()
self._cur_class_name = ""
self._vm_writer = VMWriter(out_file)
self._label_count_while = 0
self._label_count_if = 0
def compile_returnStatement(): # return expression? ;
print("---COMPILING RETURN STATEMENT---")
if (get_current_token() !=
";"): # there must be an expression if current token is not ';'
compile_expression() # push return value of expression onto stack
else:
VMWriter.writePush("constant", 0) # if void function, push constant 0
VMWriter.writeReturn()
advance() # ;
print("---DONE COMPILING RETURN STATEMENT---")
def __init__(self, input_file, output_file):
self.tokenizer = JackTokenizer(input_file)
self.symbol_table = SymbolTable()
self.vm_writer = VMWriter(output_file)
self.current_sub_name = None
self.class_name = None
self.func_counter = 0
self.while_counter = 0
self.if_counter = 0
# starts the process
self.tokenizer.advance()
self.compile_class()
self.vm_writer.close()
def compileArrayEXP(self):
symbolName = self.takeWord() # varName
symbol = self.findSymbol(symbolName)
self.advance() # [
self.compileExpression()
# push base address to stack
self.vmcode += VMWriter.writePush("local", symbol['index'])
# add index(expression result) and base addresses
self.vmcode += VMWriter.writeArithmetic('+')
# pop address into THAT
self.vmcode += VMWriter.writePop("pointer", "1")
# push value to stack
self.vmcode += VMWriter.writePush("that", "0")
self.advance() # ]
def compileExpression(self):
# Order of operations is from front to back
self.compileTerm()
while self.tokens[self.count]["token"] in [
"+", "-", "*", "/", "&", "|", "<", ">", "="
]:
op = self.takeWord() # op
self.compileTerm()
# execute op
if op == "*":
self.vmcode += VMWriter.writeCall("Math.multiply", 2)
elif op == "/":
self.vmcode += VMWriter.writeCall("Math.divide", 2)
else:
self.vmcode += VMWriter.writeArithmetic(op)
def generateFile(foldername, filename): # (Square/mine/, Main.jack)
jack_filename = foldername.rsplit(
"/", 2)[0] + "/" + filename # "Square" + "/" + "Main.jack"
tokens = JackTokenizer.tokenize(
jack_filename) # returns tokens generated from .jack file
vm_filename = foldername + filename.split(
".")[0] + ".vm" # let VMWriter know what file to write to
VMWriter.initializeFile(
vm_filename) # open .vm file to begin writing to it
# pass tokens from Tokenizer to CompilationEngine
class_name = filename.split(".")[0]
CompilationEngine.compileTokens(tokens, class_name)
def __init__(self, inFilename, outFilename):
global f
global vmWriter
global symbolTable
global filename
global currentLine
global currentClass
global currentSubroutine
global ifEndLabelIndex
global ifFalseLabelIndex
global ifTrueLabelIndex
global whileExpLabelIndex
global whileEndLabelIndex
f = None
filename = ''
currentLine = 1
currentClass = ''
currentSubroutine = ''
labelIndex = 0
vmWriter = None
symbolTable = None
ifEndLabelIndex = 0
ifTrueLabelIndex = 0
ifFalseLabelIndex = 0
whileExpLabelIndex = 0
whileEndLabelIndex = 0
filename = inFilename
vmWriter = VMWriter.vmWriter(outFilename)
symbolTable = SymbolTable.symbolTable()
f = open(inFilename, "r")
compileClass()
def __init__(self, jackTokenizerOutput, outfileName, outVMName):
'''
In: jackTokenizerOutput which is the array of tokens output by JackTokenizer.py
Function: Creates a new compilation engine with the given input and output.
The next routine called must be compileClass().
'''
self.tokenArray = jackTokenizerOutput
self.tokenIndex = 0
self.line = self.tokenArray[0] # first item of array is class
self.curToken = self.line.split()[
1] # the second item in a line is the token
self.curTokenType = self.line.split()[
0] # gets the first tag of the token
# set buffersize to 0
self.f1 = open(outfileName, 'a', 0)
# project 11 part: add Symbol table and VMWriter output
self.sTable = st.SymbolTable()
self.vm = vmw.VMWriter(outVMName)
# project 11 class variables
self.curClass = '' # name of the current class
self.whileCount = 0
self.ifCount = 0
def __init__(self, jack_lines):
self._jack_lines = jack_lines
self._token = tk.JackTokenizer(jack_lines)
self._writer = vmw.VMWriter()
self._table = st.SymbolTable()
self._class = None
self._cur_subroutine_ret_type = None
def __init__(self, file):
self.lex = Lex(file)
self.symbols = SymbolTable()
self.vm = VMWriter()
self.openout(file)
self.compile_class()
self.closeout()
def parseOneFile(fileName):
listOfTokens = PassingTokenArray(fileName)
outFile = fileName[:-5] + ".vm"
writerObj = VMWriter.Translator(outFile)
classRoot = clssNode()
parser = JackParser(listOfTokens, writerObj, classRoot)
parser.initProcess()
def __init__(self, tokenizer, ostream, file_name_no_ext):
'''Initialize the compilation engine
@tokenizer the tokenizer from the input code file
@ostream the output stream to write the code to'''
self.tokenizer = tokenizer
self.vm_writer = VMWriter.VMWriter(ostream)
self.className = file_name_no_ext
def __init__(self, in_filename, in_file, out_xml, out_vm):
"""
Creates a new compilation engine with the given input and output.
The next routine called must be compileClass().
:param in_file: Open source Jack file.
:param out_xml: Open XML file.
:param out_vm: Open VM file.
"""
self.__in_filename = in_filename
self.__in_file, self.__out_xml = in_file, out_xml
self.__tokenizer = JackTokenizer(in_file)
self.__symbolTable = SymbolTable()
self.__vmWriter = VMWriter(in_filename, out_vm)
self.__stack = list()
self.__tokenizer.advance()
self.__resetUniqueLabels()
def __init__(self, input_filename, output_file=None):
self.input_filename = input_filename
self.output_file = output_file
self.symbol_table = SymbolTable.SymbolTable()
self.subtoutine_name = ''
self.subroutine_type = ''
self.class_name = ''
self.if_index = 0
self.while_index = 0
self.var_number = 0
self.in_xml = xml.dom.minidom.parse(self.input_filename)
self.input_child_node_idx = 1
self.vm_writer = VMWriter.VMWriter(input_filename[:-5])
self.current_token = self.in_xml.documentElement.childNodes[
self.input_child_node_idx]
self.doc = xml.dom.minidom.Document()
while self.current_token.nodeName == 'keyword' and self.current_token.childNodes[
0].nodeValue == 'class':
self.compile_class()
self.__save_xml()
pass
def __init__(self, inFile, outFile):
""" Creates a new CompilationEngine with given
input and output. The next routine called must
be compileClass() """
# Create an object of JackTokenizer with the input file
self.tokenizer = JackTokenizer.JackTokenizer(inFile)
# Create an object of SymbolTable
self.symbolTable = SymbolTable.SymbolTable()
# Create an object of VMWriter
self.vmWriter = VMWriter.VMWriter(outFile[:-3] + 'vm')
# Open a output file to write to
self.out = open(outFile, 'w')
self.currentToken = ''
self.currentTokenType = ''
self.tabs = 0
self.i = 0 # for creating unique labels
self.nfield = 0
self.nstatic = 0
self.__advance()
def __init__(self, jackTokenizer):
self.jackTokenizer = jackTokenizer
self.symbolTabel = SymbolTabel()
self.vmWriter = VMWriter()
self.result = []
self.labelCount = 0 #用来解决不同if等等语句中相同label的问题
self.fieldCount = 0
self.compileClass()
class Parser(object):
def __init__(self, file):
self.lex = Lex(file)
self.symbols = SymbolTable()
self.vm = VMWriter()
self.openout(file)
self.compile_class()
self.closeout()
# VMWriter support
def openout(self, path):
outdir = os.path.join(os.path.dirname(path), "output")
file = os.path.join(outdir, os.path.basename(path))
try:
os.mkdir(outdir)
except OSError as e:
pass
self.vm.openout(file)
def closeout(self):
self.vm.closeout()
def vm_function_name(self):
return self._cur_class + "." + self._cur_subroutine
def vm_push_variable(self, name):
(type, kind, index) = self.symbols.lookup(name)
self.vm.write_push(segments[kind], index)
def vm_pop_variable(self, name):
(type, kind, index) = self.symbols.lookup(name)
self.vm.write_pop(segments[kind], index)
# Routines to advance the token
def _require(self, tok, val=None):
lextok, lexval = self._advance()
if tok != lextok or tok in (T_KEYWORD, T_SYM) and val != lexval:
raise ParserError(self._require_failed_msg(tok, val))
else:
return lexval
def _require_failed_msg(self, tok, val):
if val == None:
val = tokens[tok]
return "Expected " + val
def _advance(self):
return self.lex.advance()
def _is_token(self, tok, val=None):
lextok, lexval = self.lex.peek()
return val == None and lextok == tok or (lextok, lexval) == (tok, val)
def _is_keyword(self, *keywords):
lextok, lexval = self.lex.peek()
return lextok == T_KEYWORD and lexval in keywords
def _is_sym(self, symbols):
lextok, lexval = self.lex.peek()
return lextok == T_SYM and lexval in symbols
# Parser and compile Jack code
# class: 'class' className '{' classVarDec* subroutineDec* '}'
def compile_class(self):
self._require(T_KEYWORD, KW_CLASS)
self.compile_class_name()
self._require(T_SYM, "{")
while self._is_class_var_dec():
self.compile_class_var_dec()
while self._is_subroutine():
self.compile_subroutine()
self._require(T_SYM, "}")
# className: identifier
def compile_class_name(self):
self._cur_class = self.compile_var_name() # Class names don't have to go into the symbol table
# Variable declarations
def _is_class_var_dec(self):
return self._is_keyword(KW_STATIC, KW_FIELD)
# classVarDec: {'static'|'field'} type varName (',' varName)* ';'
def compile_class_var_dec(self):
tok, kwd = self._advance() # static | field
self._compile_dec(kwd_to_kind[kwd])
# type varName (',' varName)* ';'
def _compile_dec(self, kind):
type = self.compile_type()
name = self.compile_var_name()
self.symbols.define(name, type, kind)
while self._is_sym(","):
self._advance()
name = self.compile_var_name()
self.symbols.define(name, type, kind)
self._require(T_SYM, ";")
def _is_type(self):
return self._is_token(T_ID) or self._is_keyword(KW_INT, KW_CHAR, KW_BOOLEAN)
# 'void' | type
def compile_void_or_type(self):
if self._is_keyword(KW_VOID):
return self._advance()[1]
else:
return self.compile_type()
# type: 'int' | 'char' | 'boolean' | className
def compile_type(self):
if self._is_type():
return self._advance()[1]
else:
raise ParserError(self._require_failed_msg(*self.lex.peek()))
def _is_var_name(self):
return self._is_token(T_ID)
# varName: identifier
def compile_var_name(self):
return self._require(T_ID)
# Subroutine declarations
def _is_subroutine(self):
return self._is_keyword(KW_CONSTRUCTOR, KW_FUNCTION, KW_METHOD)
# subroutineDec: ('constructor'|'function'|'method') ('void'|type)
# subroutineName '(' parameterList ')' subroutineBody
def compile_subroutine(self):
tok, kwd = self._advance()
type = self.compile_void_or_type()
self.compile_subroutine_name()
self.symbols.start_subroutine()
if kwd == KW_METHOD:
self.symbols.define("this", self._cur_class, SK_ARG)
self._require(T_SYM, "(")
self.compile_parameter_list()
self._require(T_SYM, ")")
self.compile_subroutine_body(kwd)
# subroutineName: identifier
def compile_subroutine_name(self):
self._cur_subroutine = self.compile_var_name() # subroutine names don't have to go in the symbol table
# parameterList: (parameter (',' parameter)*)?
def compile_parameter_list(self):
if self._is_type():
self.compile_parameter()
while self._is_sym(","):
self._advance()
self.compile_parameter()
# parameter: type varName
def compile_parameter(self):
if self._is_type():
type = self.compile_type()
name = self.compile_var_name()
self.symbols.define(name, type, SK_ARG)
# subroutineBody: '{' varDec* statements '}'
def compile_subroutine_body(self, kwd):
self._require(T_SYM, "{")
while self._is_var_dec():
self.compile_var_dec()
self.write_func_decl(kwd)
self.compile_statements()
self._require(T_SYM, "}")
def write_func_decl(self, kwd):
self.vm.write_function(self.vm_function_name(), self.symbols.var_count(SK_VAR))
self.load_this_ptr(kwd)
def load_this_ptr(self, kwd):
if kwd == KW_METHOD:
self.vm.push_arg(0)
self.vm.pop_this_ptr() # set up 'this' pointer to point to new object
elif kwd == KW_CONSTRUCTOR:
self.vm.push_const(self.symbols.var_count(SK_FIELD)) # object size
self.vm.write_call("Memory.alloc", 1)
self.vm.pop_this_ptr() # set up 'this' pointer to point to new object
def _is_var_dec(self):
return self._is_keyword(KW_VAR)
# varDec: 'var' type varName (',' varName)* ';'
def compile_var_dec(self):
self._require(T_KEYWORD, KW_VAR)
return self._compile_dec(SK_VAR)
# Statements
# statement: statement*
def compile_statements(self):
while self._is_statement():
self._compile_statement()
def _is_statement(self):
return self._is_let() or self._is_if() or self._is_while() or self._is_do() or self._is_return()
# statement: letStatement | ifStatement | whileStatement | doStatement | returnStatement
def _compile_statement(self):
if self._is_let():
self.compile_let()
elif self._is_if():
self.compile_if()
elif self._is_while():
self.compile_while()
elif self._is_do():
self.compile_do()
elif self._is_return():
self.compile_return()
def _is_let(self):
return self._is_keyword(KW_LET)
# letStatement: 'let' varName ('[' expression ']')? '=' expression ';'
def compile_let(self):
self._require(T_KEYWORD, KW_LET)
name = self.compile_var_name()
subscript = self._is_sym("[")
if subscript:
self.compile_base_plus_index(name) # calculate base+index
self._require(T_SYM, "=")
self.compile_expression() # calculate expression to assign
self._require(T_SYM, ";")
if subscript:
self.pop_array_element() # *(base+index) = expr
else:
self.vm_pop_variable(name) # pop value directly into variable
# ('[' expression ']')?
def compile_base_plus_index(self, name):
self.vm_push_variable(name) # push array ptr onto stack
self._advance()
self.compile_expression() # push index onto stack
self._require(T_SYM, "]")
self.vm.write_vm_cmd("add") # base+index - leave on the stack for later
def pop_array_element(self):
self.vm.pop_temp(TEMP_ARRAY) # Pop expr value to temp register
self.vm.pop_that_ptr() # Pop base+index into 'that' register
self.vm.push_temp(TEMP_ARRAY) # Push expr back onto stack
self.vm.pop_that() # Pop value into *(base+index)
def _is_if(self):
return self._is_keyword(KW_IF)
# ifStatement: 'if' '(' expression ')' '{' statements '}'
# ('else' '{' statements '}')?
def compile_if(self):
self._require(T_KEYWORD, KW_IF)
end_label = self.new_label()
self._compile_cond_expression_statements(end_label) # VM code for condition and if statements
if self._is_keyword(KW_ELSE):
self._advance()
self._require(T_SYM, "{")
self.compile_statements() # VM code for else statements
self._require(T_SYM, "}")
self.vm.write_label(end_label) # label end_label
def _is_while(self):
return self._is_keyword(KW_WHILE)
# whileStatement: 'while' '(' expression ')' '{' statements '}'
def compile_while(self):
self._require(T_KEYWORD, KW_WHILE)
top_label = self.new_label()
self.vm.write_label(top_label) # label top_label
self._compile_cond_expression_statements(top_label) # VM code for condition and while statements
# '(' expression ')' '{' statements '}'
def _compile_cond_expression_statements(self, label):
self._require(T_SYM, "(")
self.compile_expression()
self._require(T_SYM, ")")
self.vm.write_vm_cmd("not") # ~(cond)
notif_label = self.new_label()
self.vm.write_if(notif_label) # if-goto notif_label
self._require(T_SYM, "{")
self.compile_statements() # VM code for if statements
self._require(T_SYM, "}")
self.vm.write_goto(label) # goto label
self.vm.write_label(notif_label) # label notif_label
label_num = 0
def new_label(self):
self.label_num += 1
return "label" + str(self.label_num)
def _is_do(self):
return self._is_keyword(KW_DO)
# do_statement: 'do' subroutineCall ';'
def compile_do(self):
self._require(T_KEYWORD, KW_DO)
name = self._require(T_ID)
self.compile_subroutine_call(name) # VM code for subroutine call
self.vm.pop_temp(TEMP_RETURN) # Pop return value and discard
self._require(T_SYM, ";")
def _is_return(self):
return self._is_keyword(KW_RETURN)
# returnStatement: 'return' expression? ';'
def compile_return(self):
self._require(T_KEYWORD, KW_RETURN)
if not self._is_sym(";"):
self.compile_expression() # VM code for return expression if any
else:
self.vm.push_const(0) # push 0 if not returning a value
self._require(T_SYM, ";")
self.vm.write_return() # return
# Expressions
# expression: term (op term)*
def compile_expression(self):
self.compile_term()
# Doesn't handle normal order of operations - just left to right for now
while self._is_op():
op = self._advance()
self.compile_term()
self.vm.write_vm_cmd(vm_cmds[op[1]]) # op
def _is_term(self):
return self._is_const() or self._is_var_name() or self._is_sym("(") or self._is_unary_op()
# term: integerConstant | stringConstant | keywordConstant | varName
# | varName '[' expression ']' | subroutineCall | '(' expression ')'
# | unaryOp term
def compile_term(self):
if self._is_const():
self.compile_const()
elif self._is_sym("("):
self._advance()
self.compile_expression() # VM code to evaluate expression
self._require(T_SYM, ")")
elif self._is_unary_op():
tok, op = self._advance()
self.compile_term()
self.vm.write_vm_cmd(vm_unary_cmds[op]) # op
elif self._is_var_name():
tok, name = self._advance()
if self._is_sym("["):
self.compile_array_subscript(name) # VM code for array subscript
elif self._is_sym("(."):
self.compile_subroutine_call(name) # VM code for subroutine call
else:
self.vm_push_variable(name) # push variable on stack
def _is_const(self):
return self._is_token(T_NUM) or self._is_token(T_STR) or self._is_keyword_constant()
def _is_keyword_constant(self):
return self._is_keyword(KW_TRUE, KW_FALSE, KW_NULL, KW_THIS)
def _is_op(self):
return self._is_sym("+-*/&|<>=")
def _is_unary_op(self):
return self._is_sym("-~")
# integerConstant | stringConstant | keywordConstant
def compile_const(self):
tok, val = self._advance()
if tok == T_NUM:
self.vm.push_const(val) # push constant val
elif tok == T_STR:
self.write_string_const_init(val) # initialize string & push str addr
elif tok == T_KEYWORD:
self.compile_kwd_const(val) # push TRUE, FALSE, NULL etc.
def write_string_const_init(self, val):
self.vm.push_const(len(val))
self.vm.write_call("String.new", 1) # String.new(len(str))
for c in val:
self.vm.push_const(ord(c))
self.vm.write_call("String.appendChar", 2) # String.appendChar(nextchar)
# keywordConstant: 'true' | 'false' | 'null' | 'this'
def compile_kwd_const(self, kwd):
if kwd == KW_THIS:
self.vm.push_this_ptr()
elif kwd == KW_TRUE:
self.vm.push_const(1)
self.vm.write_vm_cmd("neg")
else: # KW_FALSE or KW_NULL
self.vm.push_const(0)
# '[' expression ']'
def compile_array_subscript(self, name):
self.vm_push_variable(name) # push array ptr onto stack
self._require(T_SYM, "[")
self.compile_expression() # push index onto stack
self._require(T_SYM, "]")
self.vm.write_vm_cmd("add") # base+index
self.vm.pop_that_ptr() # pop into 'that' ptr
self.vm.push_that() # push *(base+index) onto stack
# subroutineCall: subroutineName '(' expressionList ')'
# | (className | varName) '.' subroutineName '(' expressionList ')'
def compile_subroutine_call(self, name):
(type, kind, index) = self.symbols.lookup(name)
if self._is_sym("."):
num_args, name = self.compile_dotted_subroutine_call(name, type)
else:
num_args = 1
self.vm.push_this_ptr()
name = self._cur_class + "." + name
self._require(T_SYM, "(")
num_args += self.compile_expr_list() # VM code to push arguments
self._require(T_SYM, ")")
self.vm.write_call(name, num_args) # call name num_args
def compile_dotted_subroutine_call(self, name, type):
num_args = 0
obj_name = name
self._advance()
name = self.compile_var_name()
if self._is_builtin_type(type): # e.g. int.func(123) not allowed
ParserError('Cannot use "." operator on builtin type')
elif type == None: # Calling using class name
name = obj_name + "." + name
else: # Calling using object variable name
num_args = 1
self.vm_push_variable(obj_name) # push object ptr onto stack
name = self.symbols.type_of(obj_name) + "." + name
return num_args, name
def _is_builtin_type(self, type):
return type in [KW_INT, KW_CHAR, KW_BOOLEAN, KW_VOID]
# expressionList: (expression (',' expression)*)?
def compile_expr_list(self):
num_args = 0
if self._is_term():
self.compile_expression()
num_args = 1
while self._is_sym(","):
self._advance()
self.compile_expression()
num_args += 1
return num_args
class CompilationEngine():
def __init__(self, jackTokenizer):
self.jackTokenizer = jackTokenizer
self.symbolTabel = SymbolTabel()
self.vmWriter = VMWriter()
self.result = []
self.labelCount = 0 #用来解决不同if等等语句中相同label的问题
self.fieldCount = 0
self.compileClass()
def __getFieldNums(self): #获取一个对象中的成员变量的数量, 以便构造函数编译时分配空间, 增量代码在compileClassVarDec中
return self.fieldCount
def __getLocalNums(self): #用来解决function name n 的参数问题
pointer = self.jackTokenizer.pointer+1
n = 0
while(self.jackTokenizer.context[pointer]=="var"):
n += 1
pointer += 3
while(self.jackTokenizer.context[pointer]!=";"):
pointer += 2
n += 1
pointer += 1
return n
def __match(self, label):
self.jackTokenizer.advance()
def __getAndAdvanceLabel(self):
self.labelCount += 1
return str(self.labelCount)
def __getToken(self):
return self.jackTokenizer.getToken()
def __getTokenType(self):
return self.jackTokenizer.tokenType()
def __getFurtureToken(self, n):
return self.jackTokenizer.context[self.jackTokenizer.pointer+n]
def __getPastToken(self, n):
return self.jackTokenizer.context[self.jackTokenizer.pointer-n]
def compileClass(self):
self.__match("class")
self.__match("className")
self.__match("{")
while(self.__getToken()=="static" or self.__getToken()=="field"):
self.compileClassVarDec()
while(self.__getToken()=="constructor" or self.__getToken()=="function" or self.__getToken()=="method"):
self.symbolTabel.startSubroutine() #子程序的符号表重启
self.compileSubroutineDec()
self.__match("}")
def compileClassVarDec(self):
kind = self.__getToken()
self.__match("static|field")
typE = self.__getToken()
self.__match("type")
name = []
name.append(self.__getToken())
self.__match("varName")
while(self.__getToken()!=";"):
if(self.__getToken()==","):
self.__match(",")
else:
name.append(self.__getToken())
self.__match("varName")
self.__match(";")
if(kind=="field"):
self.fieldCount += len(name)
for i in name:
self.symbolTabel.define(i, typE, kind)
def compileSubroutineDec(self):
kind = self.__getToken()
self.__match("constructor|function|method")
typE = self.__getToken()
self.__match("void|type")
functionName = self.__getToken()
self.__match("subroutineName")
self.__match("(")
self.compileParameterList()
self.__match(")")
n = self.__getLocalNums()
self.vmWriter.writeFunction(self.jackTokenizer.context[1]+"."+functionName, str(n))
if(kind=="constructor"):
self.vmWriter.writePush("constant", str(self.__getFieldNums()))
self.vmWriter.writeCall("Memory.alloc", "1")
self.vmWriter.writePop("pointer", "0")
elif(kind=="method"):
#将this指向传入的对象
self.vmWriter.writePush("argument", "0")
self.vmWriter.writePop("pointer", "0")
self.compileSubroutineBody()
def compileSubroutineBody(self):
self.__match("{")
while(self.__getToken()=="var"):
self.compileVarDec()
self.compileStatements()
self.__match("}")
def compileStatements(self):
tmp = ["let", "if", "while", "do", "return"]
while(self.__getToken() in tmp):
if(self.__getToken() == "while"):
self.compileWhileStatement()
elif(self.__getToken() == "if"):
self.compileIfStatement()
elif(self.__getToken() == "return"):
self.compileReturnStatement()
elif(self.__getToken() == "let"):
self.compileLetStatement()
else:
self.compileDoStatement()
def compileDoStatement(self): #do后面跟的必定是void类型的函数或方法
self.__match("do")
name = ""
n = 0
if(self.__getFurtureToken(1)=="."):
prefix = self.__getToken()
self.__match("className|varName")
self.__match(".")
name = self.__getToken()
self.__match("subroutineName")
if(self.symbolTabel.kindOf(prefix)!="none"): #varName.method
n += 1
#调用method, 所以要把对象的指针push进去
if(self.symbolTabel.kindOf(prefix)=="field"):
self.vmWriter.writePush("this", self.symbolTabel.indexOf(prefix))
elif(self.symbolTabel.kindOf(prefix)=="var"):
self.vmWriter.writePush("local", self.symbolTabel.indexOf(prefix))
elif(self.symbolTabel.kindOf(prefix)=="argument"):
self.vmWriter.writePush("argument", self.symbolTabel.indexOf(prefix))
elif(self.symbolTabel.kindOf(prefix)=="static"):
self.vmWriter.writePush("static", self.symbolTabel.indexOf(prefix))
name = self.symbolTabel.typeOf(prefix)+"."+name
else: #className.function
name = prefix+"."+name
self.__match("(")
n += self.compileExpressionList()
self.__match(")")
else: #do method()
name = self.jackTokenizer.context[1]+"."+self.__getToken()
self.__match("subroutineName")
self.vmWriter.writePush("pointer", "0")
self.__match("(")
n = self.compileExpressionList()
n += 1 #对象的方法
self.__match(")")
self.__match(";")
self.vmWriter.writeCall(name, str(n))
self.vmWriter.writePop("temp", "0")
def compileLetStatement(self):
isArray = False
self.__match("let")
name = self.__getToken()
self.__match("varName")
if(self.__getToken()=="["):
isArray = True
self.__match("[")
self.compileExpression()
self.__match("]")
self.__match("=")
self.compileExpression()
self.__match(";")
#以下为代码生成
if(isArray==True):
self.vmWriter.writePop("temp", "0") # "="右边的值
kind = self.symbolTabel.kindOf(name)
index = self.symbolTabel.indexOf(name)
if(kind=="field"): #把数组的指针push进去, 现在栈顶是指针和偏移值
self.vmWriter.writePush("this", str(index))
elif(kind=="argument"):
self.vmWriter.writePush("argument", str(index))
elif(kind=="var"):
self.vmWriter.writePush("local", str(index))
elif(kind=="static"):
self.vmWriter.writePush("static", str(index))
self.vmWriter.writeArithmetic("add")
self.vmWriter.writePop("pointer", "1")
self.vmWriter.writePush("temp", "0")
self.vmWriter.writePop("that", "0")
else:
kind = self.symbolTabel.kindOf(name)
index = self.symbolTabel.indexOf(name)
if(kind=="field"):
self.vmWriter.writePop("this", index)
elif(kind=="argument"):
self.vmWriter.writePop("argument", index)
elif(kind=="var"):
self.vmWriter.writePop("local", index)
elif(kind=="static"):
self.vmWriter.writePop("static", index)
def compileWhileStatement(self):
label = self.__getAndAdvanceLabel()
self.__match("while")
self.vmWriter.writeLabel("exp-examination"+label)
self.__match("(")
self.compileExpression()
self.__match(")")
self.vmWriter.writeIfGoto("loop"+label)
self.vmWriter.writeGoto("end"+label)
self.vmWriter.writeLabel("loop"+label)
self.__match("{")
self.compileStatements()
self.__match("}")
self.vmWriter.writeGoto("exp-examination"+label)
self.vmWriter.writeLabel("end"+label)
def compileIfStatement(self):
label = self.__getAndAdvanceLabel()
self.__match("if")
self.__match("(")
self.compileExpression()
self.__match(")")
self.vmWriter.writeIfGoto("true"+label)
self.vmWriter.writeGoto("false"+label)
self.vmWriter.writeLabel("true"+label)
self.__match("{")
self.compileStatements()
self.__match("}")
self.vmWriter.writeGoto("end"+label)
self.vmWriter.writeLabel("false"+label)
if(self.__getToken()=="else"):
self.__match("else")
self.__match("{")
self.compileStatements()
self.__match("}")
self.vmWriter.writeLabel("end"+label)
def compileReturnStatement(self):
self.__match("return")
if(self.__getToken()!=";"):
self.compileExpression()
else: #return;
self.vmWriter.writePush("constant", "0")
self.vmWriter.writeReturn()
self.__match(";")
def compileVarDec(self):
kind = "var"
self.__match("var")
typE = self.__getToken()
self.__match("type")
name = []
name.append(self.__getToken())
self.__match("name")
while(self.__getToken()!=";"):
if(self.__getToken()==","):
self.__match(",")
name.append(self.__getToken())
self.__match("name")
self.__match(";")
for i in name:
self.symbolTabel.define(i, typE, kind)
def compileParameterList(self):
kind = "argument"
typE = []
name = []
if(self.__getPastToken(4)=="method"): #对于方法, 第一个参数是对象的指针
typE.append(self.jackTokenizer.context[1]) #className
name.append("this")
while(self.__getToken()!=")"):
typE.append(self.__getToken())
self.__match("type")
name.append(self.__getToken())
self.__match("varName")
if(self.__getToken()==","):
self.__match(",")
for i,j in zip(name,typE):
self.symbolTabel.define(i, j, kind)
def compileExpression(self): #遇到op就放到term后面压栈, term求值之后压栈, expression求值之后压栈
op = ["+", "-", "*", "/", "&", "|", "<", ">", "="]
self.compileTerm()
while(self.__getToken() in op):
tmp_op = self.__getToken()
self.__match("op")
self.compileTerm()
#压入运算符
if(tmp_op=="*"):
self.vmWriter.writeCall("Math.multiply", "2")
elif(tmp_op=="+"):
self.vmWriter.writeArithmetic("add")
elif(tmp_op=="-"):
self.vmWriter.writeArithmetic("sub")
elif(tmp_op=="/"):
self.vmWriter.writeCall("Math.divide", "2")
elif(tmp_op=="&"):
self.vmWriter.writeArithmetic("and")
elif(tmp_op=="|"):
self.vmWriter.writeArithmetic("or")
elif(tmp_op=="<"):
self.vmWriter.writeArithmetic("lt")
elif(tmp_op==">"):
self.vmWriter.writeArithmetic("gt")
elif(tmp_op=="="):
self.vmWriter.writeArithmetic("eq")
def compileTerm(self):
keywordConstant = ["true", "false", "null", "this"]
unaryOp = ["-", "~"]
if(self.__getTokenType()=="integerConstant"):
self.vmWriter.writePush("constant", self.__getToken())
self.__match("integerConstant")
elif(self.__getTokenType()=="stringConstant"):
string = self.__getToken()
string = string.replace('"', "")
self.__match("string")
#申请空间
self.vmWriter.writePush("constant", str(len(string)))
self.vmWriter.writeCall("String.new", "1")
#调用String.appendChar <- method, 接收一个char, 返回一个字符串的地址)
for i in string:
self.vmWriter.writePush("constant", str(ord(i)))
self.vmWriter.writeCall("String.appendChar", "2")
elif(self.__getToken() in keywordConstant):
tmp = self.__getToken()
if(tmp=="null"):
self.vmWriter.writePush("constant", "0")
elif(tmp=="false"):
self.vmWriter.writePush("constant", "0")
elif(tmp=="true"):
self.vmWriter.writePush("constant", "1")
self.vmWriter.writeArithmetic("neg")
elif(tmp=="this"):
self.vmWriter.writePush("pointer", "0")
self.__match("keywordConstant")
elif(self.__getToken()=="("): #(expression)
self.__match("(")
self.compileExpression()
self.__match(")")
elif(self.__getToken() in unaryOp): #unaryOp term
tmp_uop = self.__getToken()
self.__match("unaryOp")
self.compileTerm()
if(tmp_uop=="-"):
self.vmWriter.writeArithmetic("neg")
elif(tmp_uop=="~"):
label = self.__getAndAdvanceLabel()
self.vmWriter.writeIfGoto("true"+label)
self.vmWriter.writeGoto("false"+label)
self.vmWriter.writeLabel("true"+label)
self.vmWriter.writePush("constant", "0")
self.vmWriter.writeGoto("end"+label)
self.vmWriter.writeLabel("false"+label)
self.vmWriter.writePush("constant", "1")
self.vmWriter.writeArithmetic("neg")
self.vmWriter.writeLabel("end"+label)
elif(self.__getFurtureToken(1)=="["): #varName[expression]
name = self.__getToken()
typE = self.symbolTabel.typeOf(name)
kind = self.symbolTabel.kindOf(name)
index = self.symbolTabel.indexOf(name)
if(kind=="field"):
self.vmWriter.writePush("this", str(index))
elif(kind=="var"):
self.vmWriter.writePush("local", str(index))
else:
self.vmWriter.writePush(kind, str(index))
self.__match("varName")
self.__match("[")
self.compileExpression()
self.__match("]")
self.vmWriter.writeArithmetic("add")
self.vmWriter.writePop("pointer", "1")
self.vmWriter.writePush("that", "0")
elif(self.__getFurtureToken(1)=="("): #subroutineName(expressionList)
name = self.__getToken()
name = self.jackTokenizer.context[1]+"."+self.__getToken()
self.__match("subroutineCall")
self.__match("(")
n = self.compileExpressionList()
n += 1
self.__match(")")
self.vmWriter.writePush("pointer", "0")
self.vmWriter.writeCall(name, str(n))
elif(self.__getFurtureToken(1)=="."): #(className|varName).subroutineName(expressionList)
prefix = self.__getToken()
self.__match("className|varName")
self.__match(".")
name = self.__getToken()
self.__match("subroutineName")
self.__match("(")
n = self.compileExpressionList()
self.__match(")")
if(self.symbolTabel.kindOf(prefix)=="none"): #className
name = prefix+"."+name
self.vmWriter.writeCall(name, str(n))
else: #varName
name = self.symbolTabel.typeOf(prefix)+"."+name
n += 1
#调用method, 所以要把对象的指针push进去
if(self.symbolTabel.kindOf(prefix)=="field"):
self.vmWriter.writePush("this", self.symbolTabel.indexOf(prefix))
elif(self.symbolTabel.kindOf(prefix)=="var"):
self.vmWriter.writePush("local", self.symbolTabel.indexOf(prefix))
elif(self.symbolTabel.kindOf(prefix)=="argument"):
self.vmWriter.writePush("argument", self.symbolTabel.indexOf(prefix))
elif(self.symbolTabel.kindOf(prefix)=="static"):
self.vmWriter.writePush("static", self.symbolTabel.indexOf(prefix))
self.vmWriter.writeCall(name, str(n))
else: #varName
name = self.__getToken()
kind = self.symbolTabel.kindOf(name)
index = self.symbolTabel.indexOf(name)
if(kind=="field"):
self.vmWriter.writePush("this", index)
elif(kind=="var"):
self.vmWriter.writePush("local", index)
else:
self.vmWriter.writePush(kind, index)
self.__match("varName")
def compileExpressionList(self): #返回expression的数量
n = 0
if(self.__getToken()!=")"):
self.compileExpression()
n += 1
while(self.__getToken()==","):
self.__match(",")
self.compileExpression()
n += 1
return n
def write(self, path):
for i in range(len(self.result)):
self.result[i] += "\n"
with open(path, "w") as f:
f.writelines(self.result)
