def __init__(self, opt):
self.opt = opt
self.sym = SymbolTable(opt)
self.out_str = None
self.lex = Lexer()
def __init__(self, opt):
self.opt = opt
self.sym = SymbolTable(opt)
self.out_str = None
self.lex = Lexer()
class MacroParser(object):
"""
The macro language parser and code generator class.
"""
# pylint: disable=too-few-public-methods
re_is_int = re.compile("^[-+]?[0-9]+$")
re_is_hex = re.compile("^0[xX][0-9a-fA-F]+$")
# Class constructor
###############################################################################
def __init__(self, opt):
self.opt = opt
self.sym = SymbolTable(opt)
self.out_str = None
self.lex = Lexer()
# function parse
#
# The used grammar is:
# data: /* empty */
# | data GIBBERISH
# | data IDENTIFIER
# | data '{:' data ':}'
# | data if_block
# ;
#
# if_block: IF '(' exp_or ')' '{:' data ':}' elif_blocks else_block
# ;
#
# elif_blocks: /* empty */
# | elif_blocks ELIF '(' exp_or ')' '{:' data ':}'
# ;
#
# else_block: /* empty */
# | ELSE '{:' data ':}'
# ;
#
# exp_or: exp_and
# | exp_or TOK_OR exp_and
# ;
#
# exp_and: term
# | exp_and TOK_AND exp_comparison
# ;
#
# exp_comparison: term TOK_COMPARISON term
# ;
#
# term: LITERAL
# | IDENTIFIER
# | '(' exp_or ')'
# ;
###############################################################################
def parse(self, in_str):
"""
Parse a macro string.
"""
self.lex.set_str(in_str)
self.out_str = ""
self._parse_data(do_print=True)
tok = self.lex.peek()
if tok != self.lex.tok_eof:
raise ParseError(
"{0:s}: error: misaligned closing block '{1:s}'"
.format(sys.argv[0], self.lex.text))
# function _parse_data
###############################################################################
def _parse_data(self, do_print):
"""
Private top-level parsing function.
"""
tok = self.lex.peek()
while tok != self.lex.tok_eof:
if tok == self.lex.tok_gibberish:
self._parse_gibberish(do_print)
elif tok == self.lex.tok_block_open:
self._parse_data_block(do_print)
elif tok == self.lex.tok_identifier and self.lex.text == "if":
self._parse_if_block(do_print)
elif tok == self.lex.tok_identifier:
self._parse_identifier(do_print)
elif tok == self.lex.tok_block_close:
return
else:
raise ParseError(
"{0:s}: error: wrong token '{1:s}'".format(sys.argv[0], self.lex.text))
tok = self.lex.peek()
# function _parse_gibberish
###############################################################################
def _parse_gibberish(self, do_print):
"""
Parse gibberish.
Actually, just print the characters in 'text' if do_print is True.
"""
if do_print:
self.out_str = self.out_str + self.lex.text
self.lex.advance()
# function _parse_identifier
###############################################################################
def _parse_identifier(self, do_print):
"""
Parse an identifier.
"""
try:
sym_value = self.sym.get_terminal(self.lex.text)
except SymbolLookupError:
raise ParseError(
"{0:s}: error: unknown terminal '{1:s}'"
.format(sys.argv[0], self.lex.text))
self.lex.advance()
if do_print:
self.lex.prepend(sym_value)
# function _parse_if_block
###############################################################################
def _parse_if_block(self, do_print):
"""
Parse an if block.
"""
# parse the expression following the 'if' and the associated block.
exp_res = self._parse_conditional_block(do_print)
do_print = do_print and not exp_res
# try $elif
tok = self.lex.peek()
while tok == self.lex.tok_identifier and self.lex.text == "elif":
exp_res = self._parse_conditional_block(do_print)
do_print = do_print and not exp_res
tok = self.lex.peek()
# try $else
if tok == self.lex.tok_identifier and self.lex.text == "else":
# get rid of the tok_identifier, 'else' and following spaces
self.lex.advance()
self.lex.delete_spaces()
# expect a data block
self._parse_data_block(do_print)
# function _parse_conditional_block
###############################################################################
def _parse_conditional_block(self, do_print):
"""
Parse a conditional block (such as $if or $elif).
Return the truth value of the expression.
"""
# get rid of the tok_identifier, 'if' or 'elif'
self.lex.advance()
self.lex.set_state(self.lex.state_expr)
# expect an open parenthesis
tok = self.lex.peek()
if tok != self.lex.tok_par_open:
raise ParseError(
"{0:s}: error: open parenthesis expected: '{1:s}'"
.format(sys.argv[0], self.lex.text))
self.lex.advance()
# parse the boolean expression
exp_res = self._parse_exp_or()
# expect a closed parenthesis
tok = self.lex.peek()
if tok != self.lex.tok_par_close:
raise ParseError(
"{0:s}: error: closed parenthesis expected: '{1:s}'"
.format(sys.argv[0], self.lex.text))
self.lex.advance()
# get rid of eventual spaces, and switch back to gibberish.
self.lex.delete_spaces()
self.lex.set_state(self.lex.state_gibberish)
# expect a data block
self._parse_data_block(do_print and exp_res)
# get rid of eventual spaces
# but only if followed by $if, $else or $elif
self.lex.delete_spaces(skip_unconditional=False)
return exp_res
# function _parse_data_block
###############################################################################
def _parse_data_block(self, do_print):
"""
Parse a data block.
"""
# expect an open block
tok = self.lex.peek()
if tok != self.lex.tok_block_open:
raise ParseError(
"{0:s}: error: open block expected: '{1:s}'"
.format(sys.argv[0], self.lex.text))
self.lex.advance(skip_nl=True)
# more data follows...
self._parse_data(do_print)
# expect a closed block
tok = self.lex.peek()
if tok != self.lex.tok_block_close:
raise ParseError(
"{0:s}: error: closed block expected: '{1:s}'"
.format(sys.argv[0], self.lex.text))
self.lex.advance(skip_nl=True)
# function _parse_exp_or
###############################################################################
def _parse_exp_or(self):
"""
Parse a boolean 'or' expression.
"""
ret = False
while True:
ret = self._parse_exp_and() or ret
# is the expression terminated?
tok = self.lex.peek()
if tok == self.lex.tok_par_close:
return ret
# expect an 'or' token.
elif tok == self.lex.tok_or:
self.lex.advance()
# everything else is the end of the expression.
# Let the caling function worry about error reporting.
else:
return ret
return False
# function _parse_exp_and
###############################################################################
def _parse_exp_and(self):
"""
Parse a boolean 'and' expression.
"""
ret = True
while True:
ret = self._parse_exp_comparison() and ret
# is the expression terminated?
tok = self.lex.peek()
if tok == self.lex.tok_par_close:
return ret
# expect an 'and' token.
elif tok == self.lex.tok_and:
self.lex.advance()
# everything else is a parse error.
else:
return ret
return False
# function _parse_exp_comparison
###############################################################################
def _parse_exp_comparison(self):
"""
Parse a boolean comparison.
"""
# left hand side of the comparison
lhs = self._parse_exp_term()
# expect a comparison
tok = self.lex.peek()
if tok != self.lex.tok_op:
raise ParseError(
"{0:s}: error: operator expected: '{1:s}'"
.format(sys.argv[0], self.lex.text))
operator = self.lex.text
self.lex.advance()
# right hand side of the comparison
rhs = self._parse_exp_term()
# if both operands ar numbers, convert them
num_l = self._get_num(lhs)
num_r = self._get_num(rhs)
if num_l != None and num_r != None:
lhs = num_l
rhs = num_r
# now calculate the result of the comparison, whatever that means
if operator == "<=":
ret = lhs <= rhs
elif operator == "<":
ret = lhs < rhs
elif operator == "==":
ret = lhs == rhs
elif operator == "!=":
ret = lhs != rhs
elif operator == ">=":
ret = lhs >= rhs
elif operator == ">":
ret = lhs > rhs
else:
raise ParseError(
"{0:s}: error: unknow operator: '{1:s}'"
.format(sys.argv[0], self.lex.text))
return ret
# function _parse_exp_term
###############################################################################
def _parse_exp_term(self):
"""
Parse a terminal.
"""
tok = self.lex.peek()
# identifier
if tok == self.lex.tok_identifier:
try:
ret = self.sym.get_terminal(self.lex.text)
except SymbolLookupError:
raise ParseError(
"{0:s}: error: unknown terminal '{1:s}'"
.format(sys.argv[0], self.lex.text))
if ret == None:
ret = "Undefined"
# string
elif tok == self.lex.tok_str:
ret = self.lex.text
# number
elif tok == self.lex.tok_num:
ret = self.lex.text
# parenthesised expression
elif tok == self.lex.tok_par_open:
self.lex.advance()
ret = self._parse_exp_or()
tok = self.lex.peek()
if tok != self.lex.tok_par_close:
raise ParseError(
"{0:s}: error: closed parenthesis expected: '{1:s}'"
.format(sys.argv[0], self.lex.text))
self.lex.advance()
return ret
# function _get_num
###############################################################################
def _get_num(self, in_str):
"""
Check if in_str is a number and return the numeric value.
"""
ret = None
if in_str != None:
match = self.re_is_int.match(in_str)
if match != None:
ret = int(in_str)
match = self.re_is_hex.match(in_str)
if match != None:
ret = int(in_str, 16)
return ret
class MacroParser(object):
"""
The macro language parser and code generator class.
"""
# pylint: disable=too-few-public-methods
re_is_int = re.compile("^[-+]?[0-9]+$")
re_is_hex = re.compile("^0[xX][0-9a-fA-F]+$")
# Class constructor
###############################################################################
def __init__(self, opt):
self.opt = opt
self.sym = SymbolTable(opt)
self.out_str = None
self.lex = Lexer()
# function parse
#
# The used grammar is:
# data: /* empty */
# | data GIBBERISH
# | data IDENTIFIER
# | data '{:' data ':}'
# | data if_block
# ;
#
# if_block: IF '(' exp_or ')' '{:' data ':}' elif_blocks else_block
# ;
#
# elif_blocks: /* empty */
# | elif_blocks ELIF '(' exp_or ')' '{:' data ':}'
# ;
#
# else_block: /* empty */
# | ELSE '{:' data ':}'
# ;
#
# exp_or: exp_and
# | exp_or TOK_OR exp_and
# ;
#
# exp_and: term
# | exp_and TOK_AND exp_comparison
# ;
#
# exp_comparison: term TOK_COMPARISON term
# ;
#
# term: LITERAL
# | IDENTIFIER
# | '(' exp_or ')'
# ;
###############################################################################
def parse(self, in_str):
"""
Parse a macro string.
"""
self.lex.set_str(in_str)
self.out_str = ""
self._parse_data(do_print=True)
tok = self.lex.peek()
if tok != self.lex.tok_eof:
raise ParseError(
"{0:s}: error: misaligned closing block '{1:s}'".format(
sys.argv[0], self.lex.text))
# function _parse_data
###############################################################################
def _parse_data(self, do_print):
"""
Private top-level parsing function.
"""
tok = self.lex.peek()
while tok != self.lex.tok_eof:
if tok == self.lex.tok_gibberish:
self._parse_gibberish(do_print)
elif tok == self.lex.tok_block_open:
self._parse_data_block(do_print)
elif tok == self.lex.tok_identifier and self.lex.text == "if":
self._parse_if_block(do_print)
elif tok == self.lex.tok_identifier:
self._parse_identifier(do_print)
elif tok == self.lex.tok_block_close:
return
else:
raise ParseError("{0:s}: error: wrong token '{1:s}'".format(
sys.argv[0], self.lex.text))
tok = self.lex.peek()
# function _parse_gibberish
###############################################################################
def _parse_gibberish(self, do_print):
"""
Parse gibberish.
Actually, just print the characters in 'text' if do_print is True.
"""
if do_print:
self.out_str = self.out_str + self.lex.text
self.lex.advance()
# function _parse_identifier
###############################################################################
def _parse_identifier(self, do_print):
"""
Parse an identifier.
"""
try:
sym_value = self.sym.get_terminal(self.lex.text)
except SymbolLookupError:
raise ParseError("{0:s}: error: unknown terminal '{1:s}'".format(
sys.argv[0], self.lex.text))
self.lex.advance()
if do_print:
self.lex.prepend(sym_value)
# function _parse_if_block
###############################################################################
def _parse_if_block(self, do_print):
"""
Parse an if block.
"""
# parse the expression following the 'if' and the associated block.
exp_res = self._parse_conditional_block(do_print)
do_print = do_print and not exp_res
# try $elif
tok = self.lex.peek()
while tok == self.lex.tok_identifier and self.lex.text == "elif":
exp_res = self._parse_conditional_block(do_print)
do_print = do_print and not exp_res
tok = self.lex.peek()
# try $else
if tok == self.lex.tok_identifier and self.lex.text == "else":
# get rid of the tok_identifier, 'else' and following spaces
self.lex.advance()
self.lex.delete_spaces()
# expect a data block
self._parse_data_block(do_print)
# function _parse_conditional_block
###############################################################################
def _parse_conditional_block(self, do_print):
"""
Parse a conditional block (such as $if or $elif).
Return the truth value of the expression.
"""
# get rid of the tok_identifier, 'if' or 'elif'
self.lex.advance()
self.lex.set_state(self.lex.state_expr)
# expect an open parenthesis
tok = self.lex.peek()
if tok != self.lex.tok_par_open:
raise ParseError(
"{0:s}: error: open parenthesis expected: '{1:s}'".format(
sys.argv[0], self.lex.text))
self.lex.advance()
# parse the boolean expression
exp_res = self._parse_exp_or()
# expect a closed parenthesis
tok = self.lex.peek()
if tok != self.lex.tok_par_close:
raise ParseError(
"{0:s}: error: closed parenthesis expected: '{1:s}'".format(
sys.argv[0], self.lex.text))
self.lex.advance()
# get rid of eventual spaces, and switch back to gibberish.
self.lex.delete_spaces()
self.lex.set_state(self.lex.state_gibberish)
# expect a data block
self._parse_data_block(do_print and exp_res)
# get rid of eventual spaces
# but only if followed by $if, $else or $elif
self.lex.delete_spaces(skip_unconditional=False)
return exp_res
# function _parse_data_block
###############################################################################
def _parse_data_block(self, do_print):
"""
Parse a data block.
"""
# expect an open block
tok = self.lex.peek()
if tok != self.lex.tok_block_open:
raise ParseError(
"{0:s}: error: open block expected: '{1:s}'".format(
sys.argv[0], self.lex.text))
self.lex.advance(skip_nl=True)
# more data follows...
self._parse_data(do_print)
# expect a closed block
tok = self.lex.peek()
if tok != self.lex.tok_block_close:
raise ParseError(
"{0:s}: error: closed block expected: '{1:s}'".format(
sys.argv[0], self.lex.text))
self.lex.advance(skip_nl=True)
# function _parse_exp_or
###############################################################################
def _parse_exp_or(self):
"""
Parse a boolean 'or' expression.
"""
ret = False
while True:
ret = self._parse_exp_and() or ret
# is the expression terminated?
tok = self.lex.peek()
if tok == self.lex.tok_par_close:
return ret
# expect an 'or' token.
elif tok == self.lex.tok_or:
self.lex.advance()
# everything else is the end of the expression.
# Let the caling function worry about error reporting.
else:
return ret
return False
# function _parse_exp_and
###############################################################################
def _parse_exp_and(self):
"""
Parse a boolean 'and' expression.
"""
ret = True
while True:
ret = self._parse_exp_comparison() and ret
# is the expression terminated?
tok = self.lex.peek()
if tok == self.lex.tok_par_close:
return ret
# expect an 'and' token.
elif tok == self.lex.tok_and:
self.lex.advance()
# everything else is a parse error.
else:
return ret
return False
# function _parse_exp_comparison
###############################################################################
def _parse_exp_comparison(self):
"""
Parse a boolean comparison.
"""
# left hand side of the comparison
lhs = self._parse_exp_term()
# expect a comparison
tok = self.lex.peek()
if tok != self.lex.tok_op:
raise ParseError("{0:s}: error: operator expected: '{1:s}'".format(
sys.argv[0], self.lex.text))
operator = self.lex.text
self.lex.advance()
# right hand side of the comparison
rhs = self._parse_exp_term()
# if both operands ar numbers, convert them
num_l = self._get_num(lhs)
num_r = self._get_num(rhs)
if num_l != None and num_r != None:
lhs = num_l
rhs = num_r
# now calculate the result of the comparison, whatever that means
if operator == "<=":
ret = lhs <= rhs
elif operator == "<":
ret = lhs < rhs
elif operator == "==":
ret = lhs == rhs
elif operator == "!=":
ret = lhs != rhs
elif operator == ">=":
ret = lhs >= rhs
elif operator == ">":
ret = lhs > rhs
else:
raise ParseError("{0:s}: error: unknow operator: '{1:s}'".format(
sys.argv[0], self.lex.text))
return ret
# function _parse_exp_term
###############################################################################
def _parse_exp_term(self):
"""
Parse a terminal.
"""
tok = self.lex.peek()
# identifier
if tok == self.lex.tok_identifier:
try:
ret = self.sym.get_terminal(self.lex.text)
except SymbolLookupError:
raise ParseError(
"{0:s}: error: unknown terminal '{1:s}'".format(
sys.argv[0], self.lex.text))
if ret == None:
ret = "Undefined"
# string
elif tok == self.lex.tok_str:
ret = self.lex.text
# number
elif tok == self.lex.tok_num:
ret = self.lex.text
# parenthesised expression
elif tok == self.lex.tok_par_open:
self.lex.advance()
ret = self._parse_exp_or()
tok = self.lex.peek()
if tok != self.lex.tok_par_close:
raise ParseError(
"{0:s}: error: closed parenthesis expected: '{1:s}'".
format(sys.argv[0], self.lex.text))
self.lex.advance()
return ret
# function _get_num
###############################################################################
def _get_num(self, in_str):
"""
Check if in_str is a number and return the numeric value.
"""
ret = None
if in_str != None:
match = self.re_is_int.match(in_str)
if match != None:
ret = int(in_str)
match = self.re_is_hex.match(in_str)
if match != None:
ret = int(in_str, 16)
return ret