def value_lof(ins):
""" LOF: get length of file. """
util.skip_white(ins)
num = vartypes.pass_int_unpack(parse_bracket(ins), maxint=0xffff)
util.range_check(0, 255, num)
the_file = devices.get_file(num)
return fp.pack(fp.Single.from_int(the_file.lof()))
def value_loc(ins):
""" LOC: get file pointer. """
util.skip_white(ins)
num = vartypes.pass_int_unpack(parse_bracket(ins), maxint=0xffff)
util.range_check(0, 255, num)
the_file = devices.get_file(num)
return vartypes.pack_int(the_file.loc())
def value_loc(ins):
""" LOC: get file pointer. """
util.skip_white(ins)
num = vartypes.pass_int_unpack(parse_bracket(ins), maxint=0xffff)
util.range_check(0, 255, num)
the_file = devices.get_file(num)
return vartypes.pack_int(the_file.loc())
def value_lof(ins):
""" LOF: get length of file. """
util.skip_white(ins)
num = vartypes.pass_int_unpack(parse_bracket(ins), maxint=0xffff)
util.range_check(0, 255, num)
the_file = devices.get_file(num)
return fp.pack(fp.Single.from_int(the_file.lof()))
def value_eof(ins):
""" EOF: get end-of-file. """
util.skip_white(ins)
num = vartypes.pass_int_unpack(parse_bracket(ins), maxint=0xffff)
if num == 0:
return vartypes.null('%')
util.range_check(0, 255, num)
the_file = devices.get_file(num, 'IR')
return vartypes.bool_to_integer(the_file.eof())
def value_eof(ins):
""" EOF: get end-of-file. """
util.skip_white(ins)
num = vartypes.pass_int_unpack(parse_bracket(ins), maxint=0xffff)
if num == 0:
return vartypes.null('%')
util.range_check(0, 255, num)
the_file = devices.get_file(num, 'IR')
return vartypes.bool_to_integer(the_file.eof())
def value_rnd(ins):
""" RND: get pseudorandom value. """
if util.skip_white(ins) == '(':
return rnd.get_random(
fp.unpack(vartypes.pass_single(parse_bracket(ins))))
else:
return rnd.get_random_int(1)
def parse_expression(ins, allow_empty=False, empty_err=error.MISSING_OPERAND):
""" Compute the value of the expression at the current code pointer. """
units, operators = [], []
while True:
d = util.skip_white(ins)
if d in tk.end_expression:
break
units.append(parse_expr_unit(ins))
d = util.skip_white(ins)
# string lit breaks expression, number after string lit breaks expression, + or - doesnt (could be an operator...
if d not in operator_tokens:
break
else:
ins.read(1)
if d in (tk.O_LT, tk.O_EQ, tk.O_GT):
nxt = util.skip_white(ins)
if nxt in (tk.O_LT, tk.O_EQ, tk.O_GT):
ins.read(1)
if d == nxt:
raise error.RunError(error.STX)
else:
d += nxt
if d[0] == tk.O_EQ:
# =>, =<
d = d[1] + d[0]
elif d == tk.O_GT + tk.O_LT: # ><
d = tk.O_LT + tk.O_GT
operators.append(d)
# empty expression is a syntax error (inside brackets) or Missing Operand (in an assignment) or ok (in print)
# PRINT 1+ :err 22
# Print (1+) :err 2
# print 1+) :err 2
if len(units) == 0:
if allow_empty:
return None
else:
if d in (')', ']'):
ins.read(1) # for positioning of cursor in edit gadget
# always 22 here now that the bracket is taken out?
raise error.RunError(error.STX if d in (')', ']') else empty_err)
if None in units:
raise error.RunError(error.STX)
if len(units) <= len(operators):
if d in (')', ']'):
ins.read(1)
raise error.RunError(error.STX if d in (')', ']') else error.MISSING_OPERAND)
return parse_operators(operators, units)
def parse_expression(ins, allow_empty=False, empty_err=error.MISSING_OPERAND):
""" Compute the value of the expression at the current code pointer. """
units, operators = [], []
while True:
d = util.skip_white(ins)
if d in tk.end_expression:
break
units.append(parse_expr_unit(ins))
d = util.skip_white(ins)
# string lit breaks expression, number after string lit breaks expression, + or - doesnt (could be an operator...
if d not in operator_tokens:
break
else:
ins.read(1)
if d in (tk.O_LT, tk.O_EQ, tk.O_GT):
nxt = util.skip_white(ins)
if nxt in (tk.O_LT, tk.O_EQ, tk.O_GT):
ins.read(1)
if d == nxt:
raise error.RunError(error.STX)
else:
d += nxt
if d[0] == tk.O_EQ:
# =>, =<
d = d[1] + d[0]
elif d == tk.O_GT + tk.O_LT: # ><
d = tk.O_LT + tk.O_GT
operators.append(d)
# empty expression is a syntax error (inside brackets) or Missing Operand (in an assignment) or ok (in print)
# PRINT 1+ :err 22
# Print (1+) :err 2
# print 1+) :err 2
if len(units) == 0:
if allow_empty:
return None
else:
if d in (')', ']'):
ins.read(1) # for positioning of cursor in edit gadget
# always 22 here now that the bracket is taken out?
raise error.RunError(error.STX if d in (')', ']') else empty_err)
if len(units) <= len(operators):
if d in (')', ']'):
ins.read(1)
raise error.RunError(error.STX if d in (
')', ']') else error.MISSING_OPERAND)
return parse_operators(operators, units)
def merge(g):
""" Merge program from ascii or utf8 (if utf8_files is True) stream. """
while True:
line = g.read_line()
if line is None:
break
linebuf = tokenise.tokenise_line(line)
if linebuf.read(1) == '\0':
# line starts with a number, add to program memory; store_line seeks to 1 first
store_line(linebuf)
else:
# we have read the :
if util.skip_white(linebuf) not in tk.end_line:
raise error.RunError(error.DIRECT_STATEMENT_IN_FILE)
def merge(g):
""" Merge program from ascii or utf8 (if utf8_files is True) stream. """
while True:
line = g.read_line()
if line is None:
break
linebuf = tokenise.tokenise_line(line)
if linebuf.read(1) == '\0':
# line starts with a number, add to program memory; store_line seeks to 1 first
store_line(linebuf)
else:
# we have read the :
if util.skip_white(linebuf) not in tk.end_line:
raise error.RunError(error.DIRECT_STATEMENT_IN_FILE)
def str_to_number(strval, allow_nonnum=True):
""" Convert Python str to BASIC value. """
ins = StringIO(strval)
outs = StringIO()
# skip spaces and line feeds (but not NUL).
util.skip(ins, (' ', '\n'))
tokenise_number(ins, outs)
outs.seek(0)
value = parse_value(outs)
if not allow_nonnum and util.skip_white(ins) != '':
# not everything has been parsed - error
return None
if not value:
return vartypes.null('%')
return value
def str_to_number(strval, allow_nonnum=True):
""" Convert Python str to BASIC value. """
ins = StringIO(strval)
outs = StringIO()
# skip spaces and line feeds (but not NUL).
util.skip(ins, (' ', '\n'))
tokenise_number(ins, outs)
outs.seek(0)
value = parse_value(outs)
if not allow_nonnum and util.skip_white(ins) != '':
# not everything has been parsed - error
return None
if not value:
return vartypes.null('%')
return value
def merge(g):
""" Merge program from ascii or utf8 (if utf8_files is True) stream. """
while True:
line = g.read_line()
if line is None:
break
#line, first_char = first_char + line, ''
linebuf = tokenise.tokenise_line(line)
if linebuf.read(1) == '\0':
# line starts with a number, add to program memory; store_line seeks to 1 first
store_line(linebuf)
else:
# we have read the :
if util.skip_white(linebuf) not in util.end_line:
# direct statement in file
raise error.RunError(66)
def load_ascii_file(g, first_char=''):
""" Load an ascii-codepage or utf8 (if utf8_files is True) encoded program from g. """
eof = False
lg = LineGetter(g, universal_newline, utf8=utf8_files)
while not eof:
line, eof = lg.get_line()
line, first_char = first_char + line, ''
linebuf = tokenise.tokenise_line(line)
if linebuf.read(1) == '\0':
# line starts with a number, add to program memory; store_line seeks to 1 first
store_line(linebuf)
else:
# we have read the :
if util.skip_white(linebuf) not in util.end_line:
# direct statement in file
raise error.RunError(66)
def str_to_value_keep(strval, allow_nonnum=True):
""" Convert BASIC string to BASIC value. """
if strval == ('$', ''):
return vartypes.null['%']
strval = str(vartypes.pass_string_unpack(strval))
ins = StringIO(strval)
outs = StringIO()
# skip spaces and line feeds (but not NUL).
util.skip(ins, (' ', '\n'))
tokenise_number(ins, outs)
outs.seek(0)
value = util.parse_value(outs)
if not allow_nonnum:
if util.skip_white(ins) != '':
# not everything has been parsed - error
return None
return value
def value_varptr(ins):
""" VARPTR, VARPTR$: get memory address for variable or FCB. """
dollar = util.skip_white_read_if(ins, ('$',))
util.require_read(ins, ('(',))
if (not dollar) and util.skip_white(ins) == '#':
filenum = parse_file_number_opthash(ins)
var_ptr = machine.varptr_file(filenum)
else:
name, indices = get_var_or_array_name(ins)
var_ptr = machine.varptr(name, indices)
util.require_read(ins, (')',))
if var_ptr < 0:
raise error.RunError(error.IFC)
if dollar:
return vartypes.pack_string(bytearray(chr(var.byte_size[name[-1]])) + vartypes.value_to_uint(var_ptr))
else:
return vartypes.pack_int(var_ptr)
def value_varptr(ins):
""" VARPTR, VARPTR$: get memory address for variable or FCB. """
dollar = util.skip_white_read_if(ins, ('$',))
util.require_read(ins, ('(',))
if (not dollar) and util.skip_white(ins) == '#':
filenum = parse_file_number_opthash(ins)
var_ptr = machine.varptr_file(filenum)
else:
name, indices = parse_variable(ins)
var_ptr = machine.varptr(name, indices)
util.require_read(ins, (')',))
if var_ptr < 0:
raise error.RunError(error.IFC)
var_ptr = vartypes.int_to_integer_unsigned(var_ptr)
if dollar:
return state.basic_state.strings.store(chr(vartypes.byte_size[name[-1]]) + vartypes.integer_to_bytes(var_ptr))
else:
return var_ptr
def value_varptr(ins):
""" VARPTR, VARPTR$: get memory address for variable or FCB. """
dollar = util.skip_white_read_if(ins, ('$', ))
util.require_read(ins, ('(', ))
if (not dollar) and util.skip_white(ins) == '#':
filenum = parse_file_number_opthash(ins)
var_ptr = machine.varptr_file(filenum)
else:
name, indices = get_var_or_array_name(ins)
var_ptr = machine.varptr(name, indices)
util.require_read(ins, (')', ))
if var_ptr < 0:
raise error.RunError(error.IFC)
if dollar:
return vartypes.pack_string(
bytearray(chr(var.byte_size[name[-1]])) +
vartypes.value_to_uint(var_ptr))
else:
return vartypes.pack_int(var_ptr)
def str_to_value_keep(strval, allow_nonnum=True):
""" Convert BASIC string to BASIC value (VAL). """
if strval == ('$', ''):
if allow_nonnum:
return vartypes.null['%']
else:
return None
strval = str(vartypes.pass_string_unpack(strval))
ins = StringIO(strval)
outs = StringIO()
# skip spaces and line feeds (but not NUL).
util.skip(ins, (' ', '\n'))
tokenise_number(ins, outs)
outs.seek(0)
value = parse_value(outs)
if not allow_nonnum:
if util.skip_white(ins) != '':
# not everything has been parsed - error
return None
return value
def value_varptr(ins):
""" VARPTR, VARPTR$: get memory address for variable or FCB. """
dollar = util.skip_white_read_if(ins, ('$', ))
util.require_read(ins, ('(', ))
if (not dollar) and util.skip_white(ins) == '#':
filenum = parse_file_number_opthash(ins)
var_ptr = machine.varptr_file(filenum)
else:
name, indices = parse_variable(ins)
var_ptr = machine.varptr(name, indices)
util.require_read(ins, (')', ))
if var_ptr < 0:
raise error.RunError(error.IFC)
var_ptr = vartypes.int_to_integer_unsigned(var_ptr)
if dollar:
return state.basic_state.strings.store(
chr(vartypes.byte_size[name[-1]]) +
vartypes.integer_to_bytes(var_ptr))
else:
return var_ptr
def read_entry():
""" READ a unit of DATA. """
current = state.basic_state.bytecode.tell()
state.basic_state.bytecode.seek(state.basic_state.data_pos)
if util.peek(state.basic_state.bytecode) in tk.end_statement:
# initialise - find first DATA
util.skip_to(state.basic_state.bytecode, ('\x84', )) # DATA
if state.basic_state.bytecode.read(1) not in ('\x84', ','):
raise error.RunError(error.OUT_OF_DATA)
vals, word, literal = '', '', False
while True:
# read next char; omit leading whitespace
if not literal and vals == '':
c = util.skip_white(state.basic_state.bytecode)
else:
c = util.peek(state.basic_state.bytecode)
# parse char
if c == '' or (not literal and c == ',') or (
c in tk.end_line or (not literal and c in tk.end_statement)):
break
elif c == '"':
state.basic_state.bytecode.read(1)
literal = not literal
if not literal:
util.require(state.basic_state.bytecode,
tk.end_statement + (',', ))
else:
state.basic_state.bytecode.read(1)
if literal:
vals += c
else:
word += c
# omit trailing whitespace
if c not in tk.whitespace:
vals += word
word = ''
state.basic_state.data_pos = state.basic_state.bytecode.tell()
state.basic_state.bytecode.seek(current)
return vals
def read_entry():
""" READ a unit of DATA. """
current = state.basic_state.bytecode.tell()
state.basic_state.bytecode.seek(state.basic_state.data_pos)
if util.peek(state.basic_state.bytecode) in util.end_statement:
# initialise - find first DATA
util.skip_to(state.basic_state.bytecode, ('\x84',)) # DATA
if state.basic_state.bytecode.read(1) not in ('\x84', ','):
# out of DATA
raise error.RunError(4)
vals, word, literal = '', '', False
while True:
# read next char; omit leading whitespace
if not literal and vals == '':
c = util.skip_white(state.basic_state.bytecode)
else:
c = util.peek(state.basic_state.bytecode)
# parse char
if c == '' or (not literal and c == ',') or (c in util.end_line or (not literal and c in util.end_statement)):
break
elif c == '"':
state.basic_state.bytecode.read(1)
literal = not literal
if not literal:
util.require(state.basic_state.bytecode, util.end_statement+(',',))
else:
state.basic_state.bytecode.read(1)
if literal:
vals += c
else:
word += c
# omit trailing whitespace
if c not in util.whitespace:
vals += word
word = ''
state.basic_state.data_pos = state.basic_state.bytecode.tell()
state.basic_state.bytecode.seek(current)
return vals
def parse_literal(ins):
""" Compute the value of the literal at the current code pointer. """
d = util.skip_white(ins)
# string literal
if d == '"':
ins.read(1)
if ins == state.basic_state.bytecode:
address = ins.tell() + memory.code_start
else:
address = None
output = bytearray()
# while tokenised numbers inside a string literal will be printed as tokenised numbers, they don't actually execute as such:
# a \00 character, even if inside a tokenised number, will break a string literal (and make the parser expect a
# line number afterwards, etc. We follow this.
d = ins.read(1)
while d not in tk.end_line + ('"', ):
output += d
d = ins.read(1)
if d == '\0':
ins.seek(-1, 1)
# store for easy retrieval, but don't reserve space in string memory
return state.basic_state.strings.store(output, address)
# number literals as ASCII are accepted in tokenised streams. only if they start with a figure (not & or .)
# this happens e.g. after non-keywords like AS. They are not acceptable as line numbers.
elif d in string.digits:
outs = StringIO()
representation.tokenise_number(ins, outs)
outs.seek(0)
return representation.parse_value(outs)
# number literals
elif d in tk.number:
return representation.parse_value(ins)
# gw-basic allows adding line numbers to numbers
elif d == tk.T_UINT:
return vartypes.int_to_integer_unsigned(util.parse_jumpnum(ins))
else:
raise error.RunError(error.STX)
def parse_literal(ins):
""" Compute the value of the literal at the current code pointer. """
d = util.skip_white(ins)
# string literal
if d == '"':
ins.read(1)
if ins == state.basic_state.bytecode:
address = ins.tell() + memory.code_start
else:
address = None
output = bytearray()
# while tokenised numbers inside a string literal will be printed as tokenised numbers, they don't actually execute as such:
# a \00 character, even if inside a tokenised number, will break a string literal (and make the parser expect a
# line number afterwards, etc. We follow this.
d = ins.read(1)
while d not in tk.end_line + ('"',):
output += d
d = ins.read(1)
if d == '\0':
ins.seek(-1, 1)
# store for easy retrieval, but don't reserve space in string memory
return state.basic_state.strings.store(output, address)
# number literals as ASCII are accepted in tokenised streams. only if they start with a figure (not & or .)
# this happens e.g. after non-keywords like AS. They are not acceptable as line numbers.
elif d in string.digits:
outs = StringIO()
representation.tokenise_number(ins, outs)
outs.seek(0)
return representation.parse_value(outs)
# number literals
elif d in tk.number:
return representation.parse_value(ins)
# gw-basic allows adding line numbers to numbers
elif d == tk.T_UINT:
return vartypes.int_to_integer_unsigned(util.parse_jumpnum(ins))
else:
raise error.RunError(error.STX)
def parse_expression(ins, allow_empty=False):
""" Compute the value of the expression at the current code pointer. """
stack = deque()
units = deque()
d = ''
missing_error = error.MISSING_OPERAND
# see https://en.wikipedia.org/wiki/Shunting-yard_algorithm
while True:
last = d
d = util.skip_white(ins)
# two-byte function tokens
if d in tk.twobyte:
d = util.peek(ins, n=2)
if d == tk.NOT and not (last in operators or last == ''):
# unary NOT ends expression except after another operator or at start
break
elif d in operators:
ins.read(len(d))
# get combined operators such as >=
if d in combinable:
nxt = util.skip_white(ins)
if nxt in combinable:
d += ins.read(len(nxt))
if last in operators or last == '' or d == tk.NOT:
# also if last is ( but that leads to recursive call and last == ''
nargs = 1
# zero operands for a binary operator is always syntax error
# because it will be seen as an illegal unary
if d not in unary:
raise error.RunError(error.STX)
else:
nargs = 2
_evaluate_stack(stack, units, operators[d], error.STX)
stack.append((d, nargs))
elif not (last in operators or last == ''):
# repeated unit ends expression
# repeated literals or variables or non-keywords like 'AS'
break
elif d == '(':
units.append(parse_bracket(ins))
elif d and d in string.ascii_letters:
# variable name
name, indices = parse_variable(ins)
units.append(var.get_variable(name, indices))
elif d in functions:
# apply functions
ins.read(len(d))
try:
units.append(functions[d](ins))
except (ValueError, ArithmeticError) as e:
units.append(_handle_math_error(e))
elif d in tk.end_statement:
break
elif d in tk.end_expression or d in tk.keyword:
# missing operand inside brackets or before comma is syntax error
missing_error = error.STX
break
else:
# literal
units.append(parse_literal(ins))
# empty expression is a syntax error (inside brackets)
# or Missing Operand (in an assignment)
# or not an error (in print and many functions)
if units or stack:
_evaluate_stack(stack, units, 0, missing_error)
return units[0]
elif allow_empty:
return None
else:
raise error.RunError(missing_error)
def parse_expr_unit(ins):
""" Compute the value of the expression unit at the current code pointer. """
d = util.skip_white(ins)
# string literal
if d == '"':
ins.read(1)
output = bytearray()
# while tokenised nmbers inside a string literal will be printed as tokenised numbers, they don't actually execute as such:
# a \00 character, even if inside a tokenised number, will break a string literal (and make the parser expect a
# line number afterwards, etc. We follow this.
d = ins.read(1)
while d not in tk.end_line + ('"', ):
output += d
d = ins.read(1)
if d == '\0':
ins.seek(-1, 1)
return vartypes.pack_string(output)
# variable name
elif d in string.ascii_uppercase:
name, indices = get_var_or_array_name(ins)
return var.get_var_or_array(name, indices)
# number literals as ASCII are accepted in tokenised streams. only if they start with a figure (not & or .)
# this happens e.g. after non-keywords like AS. They are not acceptable as line numbers.
elif d in string.digits:
outs = StringIO()
representation.tokenise_number(ins, outs)
outs.seek(0)
return representation.parse_value(outs)
# number literals
elif d in tk.number:
return representation.parse_value(ins)
# gw-basic allows adding line numbers to numbers
elif d == tk.T_UINT:
return vartypes.pack_int(util.parse_jumpnum(ins))
# brackets
elif d == '(':
return parse_bracket(ins)
# single-byte tokens
else:
ins.read(1)
if d == tk.INPUT: return value_input(ins)
elif d == tk.SCREEN: return value_screen(ins)
elif d == tk.USR: return value_usr(ins)
elif d == tk.FN: return value_fn(ins)
elif d == tk.NOT: return value_not(ins)
elif d == tk.ERL: return value_erl(ins)
elif d == tk.ERR: return value_err(ins)
elif d == tk.STRING: return value_string(ins)
elif d == tk.INSTR: return value_instr(ins)
elif d == tk.VARPTR: return value_varptr(ins)
elif d == tk.CSRLIN: return value_csrlin(ins)
elif d == tk.POINT: return value_point(ins)
elif d == tk.INKEY: return value_inkey(ins)
elif d == tk.O_PLUS: return parse_expr_unit(ins)
elif d == tk.O_MINUS:
return value_neg(ins)
# two-byte tokens
elif d == '\xFD':
d += ins.read(1)
if d == tk.CVI: return value_cvi(ins)
elif d == tk.CVS: return value_cvs(ins)
elif d == tk.CVD: return value_cvd(ins)
elif d == tk.MKI: return value_mki(ins)
elif d == tk.MKS: return value_mks(ins)
elif d == tk.MKD: return value_mkd(ins)
elif d == tk.EXTERR: return value_exterr(ins)
# two-byte tokens
elif d == '\xFE':
d += ins.read(1)
if d == tk.DATE: return value_date(ins)
elif d == tk.TIME: return value_time(ins)
elif d == tk.PLAY: return value_play(ins)
elif d == tk.TIMER: return value_timer(ins)
elif d == tk.ERDEV: return value_erdev(ins)
elif d == tk.IOCTL: return value_ioctl(ins)
elif d == tk.ENVIRON: return value_environ(ins)
elif d == tk.PMAP: return value_pmap(ins)
# two-byte tokens
elif d == '\xFF':
d += ins.read(1)
if d == tk.LEFT: return value_left(ins)
elif d == tk.RIGHT: return value_right(ins)
elif d == tk.MID: return value_mid(ins)
elif d == tk.SGN: return value_sgn(ins)
elif d == tk.INT: return value_int(ins)
elif d == tk.ABS: return value_abs(ins)
elif d == tk.SQR: return value_sqr(ins)
elif d == tk.RND: return value_rnd(ins)
elif d == tk.SIN: return value_sin(ins)
elif d == tk.LOG: return value_log(ins)
elif d == tk.EXP: return value_exp(ins)
elif d == tk.COS: return value_cos(ins)
elif d == tk.TAN: return value_tan(ins)
elif d == tk.ATN: return value_atn(ins)
elif d == tk.FRE: return value_fre(ins)
elif d == tk.INP: return value_inp(ins)
elif d == tk.POS: return value_pos(ins)
elif d == tk.LEN: return value_len(ins)
elif d == tk.STR: return value_str(ins)
elif d == tk.VAL: return value_val(ins)
elif d == tk.ASC: return value_asc(ins)
elif d == tk.CHR: return value_chr(ins)
elif d == tk.PEEK: return value_peek(ins)
elif d == tk.SPACE: return value_space(ins)
elif d == tk.OCT: return value_oct(ins)
elif d == tk.HEX: return value_hex(ins)
elif d == tk.LPOS: return value_lpos(ins)
elif d == tk.CINT: return value_cint(ins)
elif d == tk.CSNG: return value_csng(ins)
elif d == tk.CDBL: return value_cdbl(ins)
elif d == tk.FIX: return value_fix(ins)
elif d == tk.PEN: return value_pen(ins)
elif d == tk.STICK: return value_stick(ins)
elif d == tk.STRIG: return value_strig(ins)
elif d == tk.EOF: return value_eof(ins)
elif d == tk.LOC: return value_loc(ins)
elif d == tk.LOF: return value_lof(ins)
else:
return None
def value_rnd(ins):
""" RND: get pseudorandom value. """
if util.skip_white(ins) == '(':
return rnd.get_random(fp.unpack(vartypes.pass_single(parse_bracket(ins))))
else:
return rnd.get_random_int(1)
def parse_expr_unit(ins):
""" Compute the value of the expression unit at the current code pointer. """
d = util.skip_white(ins)
# string literal
if d == '"':
ins.read(1)
output = bytearray()
# while tokenised nmbers inside a string literal will be printed as tokenised numbers, they don't actually execute as such:
# a \00 character, even if inside a tokenised number, will break a string literal (and make the parser expect a
# line number afterwards, etc. We follow this.
d = ins.read(1)
while d not in util.end_line + ('"',):
output += d
d = ins.read(1)
if d == '\0':
ins.seek(-1, 1)
return vartypes.pack_string(output)
# variable name
elif d >= 'A' and d <= 'Z':
name, indices = get_var_or_array_name(ins)
return var.get_var_or_array(name, indices)
# number literals as ASCII are accepted in tokenised streams. only if they start with a figure (not & or .)
# this happens e.g. after non-keywords like AS. They are not acceptable as line numbers.
elif d >= '0' and d <= '9':
outs = StringIO()
representation.tokenise_number(ins, outs)
outs.seek(0)
return util.parse_value(outs)
# number literals
elif d in token.number:
return util.parse_value(ins)
# gw-basic allows adding line numbers to numbers
elif d == token.T_UINT:
return vartypes.pack_int(util.parse_jumpnum(ins))
# brackets
elif d == '(':
return parse_bracket(ins)
# single-byte tokens
else:
ins.read(1)
if d == '\x85': return value_input(ins)
elif d == '\xC8': return value_screen(ins)
elif d == '\xD0': return value_usr(ins)
elif d == '\xD1': return value_fn(ins)
elif d == '\xD3': return value_not(ins)
elif d == '\xD4': return value_erl(ins)
elif d == '\xD5': return value_err(ins)
elif d == '\xD6': return value_string(ins)
elif d == '\xD8': return value_instr(ins)
elif d == '\xDA': return value_varptr(ins)
elif d == '\xDB': return value_csrlin(ins)
elif d == '\xDC': return value_point(ins)
elif d == '\xDE': return value_inkey(ins)
elif d == '\xE9': return parse_expr_unit(ins)
elif d == '\xEA': return value_neg(ins)
# two-byte tokens
elif d == '\xFD':
d = ins.read(1)
if d == '\x81': return value_cvi(ins)
elif d =='\x82': return value_cvs(ins)
elif d =='\x83': return value_cvd(ins)
elif d =='\x84': return value_mki(ins)
elif d =='\x85': return value_mks(ins)
elif d =='\x86': return value_mkd(ins)
elif d == '\x8b': return value_exterr(ins)
# two-byte tokens
elif d == '\xFE':
d = ins.read(1)
if d == '\x8D': return value_date(ins)
elif d == '\x8E': return value_time(ins)
elif d == '\x93': return value_play(ins)
elif d == '\x94': return value_timer(ins)
elif d == '\x95': return value_erdev(ins)
elif d == '\x96': return value_ioctl(ins)
elif d == '\x9B': return value_environ(ins)
elif d == '\x9E': return value_pmap(ins)
# two-byte tokens
elif d == '\xFF':
d = ins.read(1)
if d == '\x81': return value_left(ins)
elif d == '\x82': return value_right(ins)
elif d == '\x83': return value_mid(ins)
elif d == '\x84': return value_sgn(ins)
elif d == '\x85': return value_int(ins)
elif d == '\x86': return value_abs(ins)
elif d == '\x87': return value_sqrt(ins)
elif d == '\x88': return value_rnd(ins)
elif d == '\x89': return value_sin(ins)
elif d == '\x8a': return value_log(ins)
elif d == '\x8b': return value_exp(ins)
elif d == '\x8c': return value_cos(ins)
elif d == '\x8D': return value_tan(ins)
elif d == '\x8E': return value_atn(ins)
elif d == '\x8F': return value_fre(ins)
elif d == '\x90': return value_inp(ins)
elif d == '\x91': return value_pos(ins)
elif d == '\x92': return value_len(ins)
elif d == '\x93': return value_str(ins)
elif d == '\x94': return value_val(ins)
elif d == '\x95': return value_asc(ins)
elif d == '\x96': return value_chr(ins)
elif d == '\x97': return value_peek(ins)
elif d == '\x98': return value_space(ins)
elif d == '\x99': return value_oct(ins)
elif d == '\x9A': return value_hex(ins)
elif d == '\x9B': return value_lpos(ins)
elif d == '\x9C': return value_cint(ins)
elif d == '\x9D': return value_csng(ins)
elif d == '\x9E': return value_cdbl(ins)
elif d == '\x9F': return value_fix(ins)
elif d == '\xA0': return value_pen(ins)
elif d == '\xA1': return value_stick(ins)
elif d == '\xA2': return value_strig(ins)
elif d == '\xA3': return value_eof(ins)
elif d == '\xA4': return value_loc(ins)
elif d == '\xA5': return value_lof(ins)
else:
return None
def parse_expr_unit(ins):
""" Compute the value of the expression unit at the current code pointer. """
d = util.skip_white(ins)
# string literal
if d == '"':
ins.read(1)
output = bytearray()
# while tokenised nmbers inside a string literal will be printed as tokenised numbers, they don't actually execute as such:
# a \00 character, even if inside a tokenised number, will break a string literal (and make the parser expect a
# line number afterwards, etc. We follow this.
d = ins.read(1)
while d not in tk.end_line + ('"',):
output += d
d = ins.read(1)
if d == '\0':
ins.seek(-1, 1)
return vartypes.pack_string(output)
# variable name
elif d in string.ascii_uppercase:
name, indices = get_var_or_array_name(ins)
return var.get_var_or_array(name, indices)
# number literals as ASCII are accepted in tokenised streams. only if they start with a figure (not & or .)
# this happens e.g. after non-keywords like AS. They are not acceptable as line numbers.
elif d in string.digits:
outs = StringIO()
representation.tokenise_number(ins, outs)
outs.seek(0)
return representation.parse_value(outs)
# number literals
elif d in tk.number:
return representation.parse_value(ins)
# gw-basic allows adding line numbers to numbers
elif d == tk.T_UINT:
return vartypes.pack_int(util.parse_jumpnum(ins))
# brackets
elif d == '(':
return parse_bracket(ins)
# single-byte tokens
else:
ins.read(1)
if d == tk.INPUT: return value_input(ins)
elif d == tk.SCREEN: return value_screen(ins)
elif d == tk.USR: return value_usr(ins)
elif d == tk.FN: return value_fn(ins)
elif d == tk.NOT: return value_not(ins)
elif d == tk.ERL: return value_erl(ins)
elif d == tk.ERR: return value_err(ins)
elif d == tk.STRING: return value_string(ins)
elif d == tk.INSTR: return value_instr(ins)
elif d == tk.VARPTR: return value_varptr(ins)
elif d == tk.CSRLIN: return value_csrlin(ins)
elif d == tk.POINT: return value_point(ins)
elif d == tk.INKEY: return value_inkey(ins)
elif d == tk.O_PLUS: return parse_expr_unit(ins)
elif d == tk.O_MINUS: return value_neg(ins)
# two-byte tokens
elif d == '\xFD':
d += ins.read(1)
if d == tk.CVI: return value_cvi(ins)
elif d == tk.CVS: return value_cvs(ins)
elif d == tk.CVD: return value_cvd(ins)
elif d == tk.MKI: return value_mki(ins)
elif d == tk.MKS: return value_mks(ins)
elif d == tk.MKD: return value_mkd(ins)
elif d == tk.EXTERR: return value_exterr(ins)
# two-byte tokens
elif d == '\xFE':
d += ins.read(1)
if d == tk.DATE: return value_date(ins)
elif d == tk.TIME: return value_time(ins)
elif d == tk.PLAY: return value_play(ins)
elif d == tk.TIMER: return value_timer(ins)
elif d == tk.ERDEV: return value_erdev(ins)
elif d == tk.IOCTL: return value_ioctl(ins)
elif d == tk.ENVIRON: return value_environ(ins)
elif d == tk.PMAP: return value_pmap(ins)
# two-byte tokens
elif d == '\xFF':
d += ins.read(1)
if d == tk.LEFT: return value_left(ins)
elif d == tk.RIGHT: return value_right(ins)
elif d == tk.MID: return value_mid(ins)
elif d == tk.SGN: return value_sgn(ins)
elif d == tk.INT: return value_int(ins)
elif d == tk.ABS: return value_abs(ins)
elif d == tk.SQR: return value_sqr(ins)
elif d == tk.RND: return value_rnd(ins)
elif d == tk.SIN: return value_sin(ins)
elif d == tk.LOG: return value_log(ins)
elif d == tk.EXP: return value_exp(ins)
elif d == tk.COS: return value_cos(ins)
elif d == tk.TAN: return value_tan(ins)
elif d == tk.ATN: return value_atn(ins)
elif d == tk.FRE: return value_fre(ins)
elif d == tk.INP: return value_inp(ins)
elif d == tk.POS: return value_pos(ins)
elif d == tk.LEN: return value_len(ins)
elif d == tk.STR: return value_str(ins)
elif d == tk.VAL: return value_val(ins)
elif d == tk.ASC: return value_asc(ins)
elif d == tk.CHR: return value_chr(ins)
elif d == tk.PEEK: return value_peek(ins)
elif d == tk.SPACE: return value_space(ins)
elif d == tk.OCT: return value_oct(ins)
elif d == tk.HEX: return value_hex(ins)
elif d == tk.LPOS: return value_lpos(ins)
elif d == tk.CINT: return value_cint(ins)
elif d == tk.CSNG: return value_csng(ins)
elif d == tk.CDBL: return value_cdbl(ins)
elif d == tk.FIX: return value_fix(ins)
elif d == tk.PEN: return value_pen(ins)
elif d == tk.STICK: return value_stick(ins)
elif d == tk.STRIG: return value_strig(ins)
elif d == tk.EOF: return value_eof(ins)
elif d == tk.LOC: return value_loc(ins)
elif d == tk.LOF: return value_lof(ins)
else:
return None
def parse_expression(ins, allow_empty=False):
""" Compute the value of the expression at the current code pointer. """
stack = deque()
units = deque()
d = ''
missing_error = error.MISSING_OPERAND
# see https://en.wikipedia.org/wiki/Shunting-yard_algorithm
while True:
last = d
d = util.skip_white(ins)
# two-byte function tokens
if d in tk.twobyte:
d = util.peek(ins, n=2)
if d == tk.NOT and not (last in operators or last == ''):
# unary NOT ends expression except after another operator or at start
break
elif d in operators:
ins.read(len(d))
# get combined operators such as >=
if d in combinable:
nxt = util.skip_white(ins)
if nxt in combinable:
d += ins.read(len(nxt))
if last in operators or last == '' or d == tk.NOT:
# also if last is ( but that leads to recursive call and last == ''
nargs = 1
# zero operands for a binary operator is always syntax error
# because it will be seen as an illegal unary
if d not in unary:
raise error.RunError(error.STX)
else:
nargs = 2
_evaluate_stack(stack, units, operators[d], error.STX)
stack.append((d, nargs))
elif not (last in operators or last == ''):
# repeated unit ends expression
# repeated literals or variables or non-keywords like 'AS'
break
elif d == '(':
units.append(parse_bracket(ins))
elif d and d in string.ascii_letters:
# variable name
name, indices = parse_variable(ins)
units.append(var.get_variable(name, indices))
elif d in functions:
# apply functions
ins.read(len(d))
try:
units.append(functions[d](ins))
except (ValueError, ArithmeticError) as e:
units.append(_handle_math_error(e))
elif d in tk.end_statement:
break
elif d in tk.end_expression or d in tk.keyword:
# missing operand inside brackets or before comma is syntax error
missing_error = error.STX
break
else:
# literal
units.append(parse_literal(ins))
# empty expression is a syntax error (inside brackets)
# or Missing Operand (in an assignment)
# or not an error (in print and many functions)
if units or stack:
_evaluate_stack(stack, units, 0, missing_error)
return units[0]
elif allow_empty:
return None
else:
raise error.RunError(missing_error)
