# HG changeset patch # User windel # Date 1356353820 -3600 # Node ID 3f772feb12efc901d8da1961991f1c41e9bb9a00 # Parent 5a965e9664f2e6a7b3dae3315d4af015442d5b6c movage diff -r 5a965e9664f2 -r 3f772feb12ef python/ide.py --- a/python/ide.py Mon Dec 24 13:32:54 2012 +0100 +++ b/python/ide.py Mon Dec 24 13:57:00 2012 +0100 @@ -7,9 +7,9 @@ from PyQt4.QtGui import * # Compiler imports: -sys.path.insert(0, os.path.join('..','libs')) +sys.path.insert(0, '.') # Add current path to the path string from project import Project -from compiler import Compiler +from ppci import Compiler from widgets import CodeEdit, AstViewer lcfospng = base64.decodestring(b'iVBORw0KGgoAAAANSUhEUgAAACAAAAAgCAYAAABzenr0AAAAAXNSR0IArs4c6QAAAAZiS0dEAP8A\n/wD/oL2nkwAAAAlwSFlzAAALEwAACxMBAJqcGAAAAAd0SU1FB9sJEhMKBk7B678AAAA/SURBVFjD\n7dbBCQAgDATBi9h/y7EFA4Kf2QLCwH1S6XQu6sqoujublc8BAAAAAAAAAAB8B+zXT6YJAAAAAKYd\nWSgFQNUyijIAAAAASUVORK5CYII=\n') diff -r 5a965e9664f2 -r 3f772feb12ef python/ppci/frontends/ks/lexer.py --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/python/ppci/frontends/ks/lexer.py Mon Dec 24 13:57:00 2012 +0100 @@ -0,0 +1,71 @@ +import collections +import re +from .errors import CompilerException + +""" + Lexical analyzer part. Splits the input character stream into tokens. +""" + +# Token is used in the lexical analyzer: +Token = collections.namedtuple('Token', 'typ val row col') + +keywords = ['and', 'array', 'begin', 'by', 'case', 'const', 'div', 'do', \ + 'else', 'elsif', 'end', 'false', 'for', 'if', 'import', 'in', 'is', \ + 'mod', 'module', 'nil', 'not', 'of', 'or', 'pointer', 'procedure', \ + 'record', 'repeat', 'return', 'then', 'to', 'true', 'type', 'until', 'var', \ + 'while', 'asm' ] + +def tokenize(s): + """ + Tokenizer, generates an iterator that + returns tokens! + + This GREAT example was taken from python re doc page! + """ + tok_spec = [ + ('REAL', r'\d+\.\d+'), + ('HEXNUMBER', r'0x[\da-fA-F]+'), + ('NUMBER', r'\d+'), + ('ID', r'[A-Za-z][A-Za-z\d_]*'), + ('NEWLINE', r'\n'), + ('SKIP', r'[ \t]'), + ('COMMENTS', r'{.*}'), + ('LEESTEKEN', r':=|[\.,=:;\-+*\[\]/\(\)]|>=|<=|<>|>|<'), + ('STRING', r"'.*?'") + ] + tok_re = '|'.join('(?P<%s>%s)' % pair for pair in tok_spec) + gettok = re.compile(tok_re).match + line = 1 + pos = line_start = 0 + mo = gettok(s) + while mo is not None: + typ = mo.lastgroup + val = mo.group(typ) + if typ == 'NEWLINE': + line_start = pos + line += 1 + elif typ == 'COMMENTS': + pass + elif typ != 'SKIP': + if typ == 'ID': + if val in keywords: + typ = val + elif typ == 'LEESTEKEN': + typ = val + elif typ == 'NUMBER': + val = int(val) + elif typ == 'HEXNUMBER': + val = int(val[2:], 16) + typ = 'NUMBER' + elif typ == 'REAL': + val = float(val) + elif typ == 'STRING': + val = val[1:-1] + yield Token(typ, val, line, mo.start()-line_start) + pos = mo.end() + mo = gettok(s, pos) + if pos != len(s): + col = pos - line_start + raise CompilerException('Unexpected character {0}'.format(s[pos]), line, col) + yield Token('END', '', line, 0) + diff -r 5a965e9664f2 -r 3f772feb12ef python/ppci/frontends/ks/parser.py --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/python/ppci/frontends/ks/parser.py Mon Dec 24 13:57:00 2012 +0100 @@ -0,0 +1,183 @@ +""" + This module define a grammar for the 'K#' language. +""" + +from .nodes import * +from .errors import CompilerException, Error +from .modules import loadModule +from .display import printNode +from .builtin import * +from . import assembler + +class Grammar: + # TODO: implement some base class? + pass + +class Parser: + #TODO + pass + +class KsParser(Parser): + def __init__(self): + self.loadGrammar(KsGrammar) + +# For now, try to parse an expression as test case: +class KsGrammer(Grammar): + + def __init__(self): + pass + + # Parsing expressions: + """ + grammar of expressions: + expression = term { addoperator term } + addoperator = '+' | '-' + term = factor { muloperator factor } + muloperator = '*' | '/' + factor = number | "(" expression ")" + """ + + @rule(Term) + def Expression1(self, term): + return Expression(term) + + @rule(Term, AddOperator, Term) + def Expression2(self, term1, op, term2): + return Expression(term1, op, term2) + + # Parsing arithmatic expressions: + def parseTerm(self): + a = self.parseFactor() + while self.token.typ in ['*', '/', 'mod', 'div', 'and']: + loc = self.getLocation() + op = self.Consume() + b = self.parseTerm() + # Type determination and checking: + if op in ['mod', 'div']: + if not isType(a.typ, integer): + self.Error('First operand should be integer, not {0}'.format(a.typ)) + if not isType(b.typ, integer): + self.Error('Second operand should be integer, not {0}'.format(b.typ)) + typ = integer + elif op == '*': + if isType(a.typ, integer) and isType(b.typ, integer): + typ = integer + elif isType(a.typ, real) or isType(b.typ, real): + if isType(a.typ, integer): + # Automatic type cast + a = Unop(a, 'INTTOREAL', real) + if isType(b.typ, integer): + b = Unop(b, 'INTTOREAL', real) + if not isType(a.typ, real): + self.Error('first operand must be a real!') + if not isType(b.typ, real): + self.Error('second operand must be a real!') + typ = real + else: + self.Error('Unknown operands for multiply: {0}, {1}'.format(a, b)) + elif op == '/': + # Division always yields a real result, for integer division use div + if isType(a.typ, integer): + # Automatic type cast + a = Unop(a, 'INTTOREAL', real) + if isType(b.typ, integer): + b = Unop(b, 'INTTOREAL', real) + if not isType(a.typ, real): + self.Error('first operand must be a real!') + if not isType(b.typ, real): + self.Error('second operand must be a real!') + typ = real + elif op == 'and': + if not isType(a.typ, boolean): + self.Error('First operand of and must be boolean') + if not isType(b.typ, boolean): + self.Error('Second operand of and must be boolean') + typ = boolean + else: + self.Error('Unknown operand {0}'.format(op)) + + a = self.setLocation(Binop(a, op, b, typ), loc) + return a + + @rule( + def parseFactor(self): + if self.hasConsumed('('): + e = self.parseExpression() + self.Consume(')') + return e + elif self.token.typ == 'NUMBER': + loc = self.getLocation() + val = self.Consume('NUMBER') + return self.setLocation(Constant(val, integer), loc) + elif self.token.typ == 'REAL': + loc = self.getLocation() + val = self.Consume('REAL') + return self.setLocation(Constant(val, real), loc) + elif self.token.typ == 'CHAR': + val = self.Consume('CHAR') + return Constant(val, char) + elif self.token.typ in ['true', 'false']: + val = self.Consume() + val = True if val == 'true' else False + return Constant(val, boolean) + elif self.hasConsumed('nil'): + return Constant(0, NilType()) + elif self.hasConsumed('not'): + f = self.parseFactor() + if not isType(f.typ, boolean): + self.Error('argument of boolean negation must be boolean type') + return Unop(f, 'not', boolean) + elif self.token.typ == 'ID': + designator = self.parseDesignator() + # TODO: handle functions different here? + if self.token.typ == '(' and type(designator.typ) is ProcedureType: + return self.parseProcedureCall(designator) + else: + return designator + else: + self.Error('Expected NUMBER, ID or ( expr ), got'+str(self.token)) + + def parseSimpleExpression(self): + """ Arithmatic expression """ + if self.token.typ in ['+', '-']: + # Handle the unary minus + op = self.Consume() + a = self.parseTerm() + typ = a.typ + if not isType(typ,real) and not isType(typ, integer): + self.Error('Unary minus or plus can be only applied to real or integers') + if op == '-': + a = Unop(a, op, typ) + else: + a = self.parseTerm() + while self.token.typ in ['+', '-', 'or']: + loc = self.getLocation() + op = self.Consume() + b = self.parseTerm() + if op in ['+', '-']: + if isType(a.typ, real) or isType(b.typ, real): + typ = real + if isType(a.typ, integer): + # Automatic type cast + a = Unop(a, 'INTTOREAL', real) + if not isType(a.typ, real): + self.Error('first operand must be a real!') + if isType(b.typ, integer): + b = Unop(b, 'INTTOREAL', real) + if not isType(b.typ, real): + self.Error('second operand must be a real!') + elif isType(a.typ, integer) and isType(b.typ, integer): + typ = integer + else: + self.Error('Invalid types {0} and {1}'.format(a.typ, b.typ)) + elif op == 'or': + if not isType(a.typ, boolean): + self.Error('first operand must be boolean for or operation') + if not isType(b.typ, boolean): + self.Error('second operand must be boolean for or operation') + typ = boolean + else: + self.Error('Unknown operand {0}'.format(op)) + a = self.setLocation(Binop(a, op, b, typ), loc) + return a + diff -r 5a965e9664f2 -r 3f772feb12ef python/ppci/frontends/ksparser.py --- a/python/ppci/frontends/ksparser.py Mon Dec 24 13:32:54 2012 +0100 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,183 +0,0 @@ -""" - This module define a grammar for the 'K#' language. -""" - -from .nodes import * -from .errors import CompilerException, Error -from .modules import loadModule -from .display import printNode -from .builtin import * -from . import assembler - -class Grammar: - # TODO: implement some base class? - pass - -class Parser: - #TODO - pass - -class KsParser(Parser): - def __init__(self): - self.loadGrammar(KsGrammar) - -# For now, try to parse an expression as test case: -class KsGrammer(Grammar): - - def __init__(self): - pass - - # Parsing expressions: - """ - grammar of expressions: - expression = term { addoperator term } - addoperator = '+' | '-' - term = factor { muloperator factor } - muloperator = '*' | '/' - factor = number | "(" expression ")" - """ - - @rule(Term) - def Expression1(self, term): - return Expression(term) - - @rule(Term, AddOperator, Term) - def Expression2(self, term1, op, term2): - return Expression(term1, op, term2) - - # Parsing arithmatic expressions: - def parseTerm(self): - a = self.parseFactor() - while self.token.typ in ['*', '/', 'mod', 'div', 'and']: - loc = self.getLocation() - op = self.Consume() - b = self.parseTerm() - # Type determination and checking: - if op in ['mod', 'div']: - if not isType(a.typ, integer): - self.Error('First operand should be integer, not {0}'.format(a.typ)) - if not isType(b.typ, integer): - self.Error('Second operand should be integer, not {0}'.format(b.typ)) - typ = integer - elif op == '*': - if isType(a.typ, integer) and isType(b.typ, integer): - typ = integer - elif isType(a.typ, real) or isType(b.typ, real): - if isType(a.typ, integer): - # Automatic type cast - a = Unop(a, 'INTTOREAL', real) - if isType(b.typ, integer): - b = Unop(b, 'INTTOREAL', real) - if not isType(a.typ, real): - self.Error('first operand must be a real!') - if not isType(b.typ, real): - self.Error('second operand must be a real!') - typ = real - else: - self.Error('Unknown operands for multiply: {0}, {1}'.format(a, b)) - elif op == '/': - # Division always yields a real result, for integer division use div - if isType(a.typ, integer): - # Automatic type cast - a = Unop(a, 'INTTOREAL', real) - if isType(b.typ, integer): - b = Unop(b, 'INTTOREAL', real) - if not isType(a.typ, real): - self.Error('first operand must be a real!') - if not isType(b.typ, real): - self.Error('second operand must be a real!') - typ = real - elif op == 'and': - if not isType(a.typ, boolean): - self.Error('First operand of and must be boolean') - if not isType(b.typ, boolean): - self.Error('Second operand of and must be boolean') - typ = boolean - else: - self.Error('Unknown operand {0}'.format(op)) - - a = self.setLocation(Binop(a, op, b, typ), loc) - return a - - @rule( - def parseFactor(self): - if self.hasConsumed('('): - e = self.parseExpression() - self.Consume(')') - return e - elif self.token.typ == 'NUMBER': - loc = self.getLocation() - val = self.Consume('NUMBER') - return self.setLocation(Constant(val, integer), loc) - elif self.token.typ == 'REAL': - loc = self.getLocation() - val = self.Consume('REAL') - return self.setLocation(Constant(val, real), loc) - elif self.token.typ == 'CHAR': - val = self.Consume('CHAR') - return Constant(val, char) - elif self.token.typ in ['true', 'false']: - val = self.Consume() - val = True if val == 'true' else False - return Constant(val, boolean) - elif self.hasConsumed('nil'): - return Constant(0, NilType()) - elif self.hasConsumed('not'): - f = self.parseFactor() - if not isType(f.typ, boolean): - self.Error('argument of boolean negation must be boolean type') - return Unop(f, 'not', boolean) - elif self.token.typ == 'ID': - designator = self.parseDesignator() - # TODO: handle functions different here? - if self.token.typ == '(' and type(designator.typ) is ProcedureType: - return self.parseProcedureCall(designator) - else: - return designator - else: - self.Error('Expected NUMBER, ID or ( expr ), got'+str(self.token)) - - def parseSimpleExpression(self): - """ Arithmatic expression """ - if self.token.typ in ['+', '-']: - # Handle the unary minus - op = self.Consume() - a = self.parseTerm() - typ = a.typ - if not isType(typ,real) and not isType(typ, integer): - self.Error('Unary minus or plus can be only applied to real or integers') - if op == '-': - a = Unop(a, op, typ) - else: - a = self.parseTerm() - while self.token.typ in ['+', '-', 'or']: - loc = self.getLocation() - op = self.Consume() - b = self.parseTerm() - if op in ['+', '-']: - if isType(a.typ, real) or isType(b.typ, real): - typ = real - if isType(a.typ, integer): - # Automatic type cast - a = Unop(a, 'INTTOREAL', real) - if not isType(a.typ, real): - self.Error('first operand must be a real!') - if isType(b.typ, integer): - b = Unop(b, 'INTTOREAL', real) - if not isType(b.typ, real): - self.Error('second operand must be a real!') - elif isType(a.typ, integer) and isType(b.typ, integer): - typ = integer - else: - self.Error('Invalid types {0} and {1}'.format(a.typ, b.typ)) - elif op == 'or': - if not isType(a.typ, boolean): - self.Error('first operand must be boolean for or operation') - if not isType(b.typ, boolean): - self.Error('second operand must be boolean for or operation') - typ = boolean - else: - self.Error('Unknown operand {0}'.format(op)) - a = self.setLocation(Binop(a, op, b, typ), loc) - return a - diff -r 5a965e9664f2 -r 3f772feb12ef python/ppci/frontends/lexer.py --- a/python/ppci/frontends/lexer.py Mon Dec 24 13:32:54 2012 +0100 +++ /dev/null Thu Jan 01 00:00:00 1970 +0000 @@ -1,71 +0,0 @@ -import collections -import re -from .errors import CompilerException - -""" - Lexical analyzer part. Splits the input character stream into tokens. -""" - -# Token is used in the lexical analyzer: -Token = collections.namedtuple('Token', 'typ val row col') - -keywords = ['and', 'array', 'begin', 'by', 'case', 'const', 'div', 'do', \ - 'else', 'elsif', 'end', 'false', 'for', 'if', 'import', 'in', 'is', \ - 'mod', 'module', 'nil', 'not', 'of', 'or', 'pointer', 'procedure', \ - 'record', 'repeat', 'return', 'then', 'to', 'true', 'type', 'until', 'var', \ - 'while', 'asm' ] - -def tokenize(s): - """ - Tokenizer, generates an iterator that - returns tokens! - - This GREAT example was taken from python re doc page! - """ - tok_spec = [ - ('REAL', r'\d+\.\d+'), - ('HEXNUMBER', r'0x[\da-fA-F]+'), - ('NUMBER', r'\d+'), - ('ID', r'[A-Za-z][A-Za-z\d_]*'), - ('NEWLINE', r'\n'), - ('SKIP', r'[ \t]'), - ('COMMENTS', r'{.*}'), - ('LEESTEKEN', r':=|[\.,=:;\-+*\[\]/\(\)]|>=|<=|<>|>|<'), - ('STRING', r"'.*?'") - ] - tok_re = '|'.join('(?P<%s>%s)' % pair for pair in tok_spec) - gettok = re.compile(tok_re).match - line = 1 - pos = line_start = 0 - mo = gettok(s) - while mo is not None: - typ = mo.lastgroup - val = mo.group(typ) - if typ == 'NEWLINE': - line_start = pos - line += 1 - elif typ == 'COMMENTS': - pass - elif typ != 'SKIP': - if typ == 'ID': - if val in keywords: - typ = val - elif typ == 'LEESTEKEN': - typ = val - elif typ == 'NUMBER': - val = int(val) - elif typ == 'HEXNUMBER': - val = int(val[2:], 16) - typ = 'NUMBER' - elif typ == 'REAL': - val = float(val) - elif typ == 'STRING': - val = val[1:-1] - yield Token(typ, val, line, mo.start()-line_start) - pos = mo.end() - mo = gettok(s, pos) - if pos != len(s): - col = pos - line_start - raise CompilerException('Unexpected character {0}'.format(s[pos]), line, col) - yield Token('END', '', line, 0) -