"""The builtin int type (W_AbstractInt) and the base impl (W_IntObject) based on rpython ints. In order to have the same behavior running on CPython, and after RPython translation this module uses rarithmetic.ovfcheck to explicitly check for overflows, something CPython does not do anymore. """ import operator import sys from rpython.rlib import jit from rpython.rlib.objectmodel import instantiate, enforceargs from rpython.rlib.rarithmetic import ( LONG_BIT, intmask, is_valid_int, ovfcheck, r_longlong, r_uint, string_to_int) from rpython.rlib.rbigint import ( InvalidEndiannessError, InvalidSignednessError, rbigint) from rpython.rlib.rfloat import DBL_MANT_DIG from rpython.rlib.rstring import ( ParseStringError, ParseStringOverflowError, MaxDigitsError) from rpython.tool.sourcetools import func_renamer, func_with_new_name from pypy.interpreter import typedef from pypy.interpreter.baseobjspace import W_Root from pypy.interpreter.error import OperationError, oefmt from pypy.interpreter.gateway import ( WrappedDefault, applevel, interp2app, interpindirect2app, unwrap_spec) from pypy.interpreter.typedef import TypeDef from pypy.objspace.std import newformat from pypy.objspace.std.util import ( BINARY_OPS, CMP_OPS, COMMUTATIVE_OPS, IDTAG_INT, IDTAG_SHIFT, wrap_parsestringerror) SENTINEL = object() HASH_BITS = 61 if sys.maxsize > 2 ** 31 - 1 else 31 HASH_MODULUS = 2 ** HASH_BITS - 1 MININT = -sys.maxint-1 class W_AbstractIntObject(W_Root): __slots__ = () def is_w(self, space, w_other): from pypy.objspace.std.boolobject import W_BoolObject if (not isinstance(w_other, W_AbstractIntObject) or isinstance(w_other, W_BoolObject)): return False if self.user_overridden_class or w_other.user_overridden_class: return self is w_other x = space.bigint_w(self, allow_conversion=False) y = space.bigint_w(w_other, allow_conversion=False) return x.eq(y) def immutable_unique_id(self, space): if self.user_overridden_class: return None b = space.bigint_w(self) b = b.lshift(IDTAG_SHIFT).int_or_(IDTAG_INT) return space.newlong_from_rbigint(b) @staticmethod @unwrap_spec(byteorder='text', signed=bool) def descr_from_bytes(space, w_inttype, w_obj, byteorder='big', signed=False): """int.from_bytes(bytes, byteorder='big', *, signed=False) -> int Return the integer represented by the given array of bytes. The bytes argument must either support the buffer protocol or be an iterable object producing bytes. Bytes and bytearray are examples of built-in objects that support the buffer protocol. The byteorder argument determines the byte order used to represent the integer. If byteorder is 'big', the most significant byte is at the beginning of the byte array. If byteorder is 'little', the most significant byte is at the end of the byte array. To request the native byte order of the host system, use `sys.byteorder' as the byte order value. The signed keyword-only argument indicates whether two's complement is used to represent the integer. """ from pypy.objspace.std.bytesobject import makebytesdata_w bytes = makebytesdata_w(space, w_obj) try: bigint = rbigint.frombytes(bytes, byteorder=byteorder, signed=signed) except InvalidEndiannessError: raise oefmt(space.w_ValueError, "byteorder must be either 'little' or 'big'") try: as_int = bigint.toint() except OverflowError: w_obj = space.newlong_from_rbigint(bigint) else: w_obj = space.newint(as_int) if not space.is_w(w_inttype, space.w_int): # That's what from_bytes() does in CPython 3.5.2 too w_obj = space.call_function(w_inttype, w_obj) return w_obj @unwrap_spec(length=int, byteorder='text', signed=bool) def descr_to_bytes(self, space, length=1, byteorder='big', signed=False): """to_bytes(...) int.to_bytes(length=1, byteorder='big', *, signed=False) -> bytes Return an array of bytes representing an integer. The integer is represented using length bytes. An OverflowError is raised if the integer is not representable with the given number of bytes. The byteorder argument determines the byte order used to represent the integer. If byteorder is 'big', the most significant byte is at the beginning of the byte array. If byteorder is 'little', the most significant byte is at the end of the byte array. To request the native byte order of the host system, use `sys.byteorder' as the byte order value. The signed keyword-only argument determines whether two's complement is used to represent the integer. If signed is False and a negative integer is given, an OverflowError is raised. """ bigint = space.bigint_w(self) try: byte_string = bigint.tobytes(length, byteorder=byteorder, signed=signed) except InvalidEndiannessError: raise oefmt(space.w_ValueError, "byteorder must be either 'little' or 'big'") except InvalidSignednessError: raise oefmt(space.w_OverflowError, "can't convert negative int to unsigned") except OverflowError: raise oefmt(space.w_OverflowError, "int too big to convert") return space.newbytes(byte_string) def descr_round(self, space, w_ndigits=None): """Rounding an Integral returns itself. Rounding with an ndigits argument also returns an integer. """ # To round an integer m to the nearest 10**n (n positive), we # make use of the divmod_near operation, defined by: # # divmod_near(a, b) = (q, r) # # where q is the nearest integer to the quotient a / b (the # nearest even integer in the case of a tie) and r == a - q * b. # Hence q * b = a - r is the nearest multiple of b to a, # preferring even multiples in the case of a tie. # # So the nearest multiple of 10**n to m is: # # m - divmod_near(m, 10**n)[1] # XXX: since divmod_near is pure python we can probably remove # the longs used here. or this could at least likely be more # efficient for W_IntObject from pypy.objspace.std.longobject import newlong if space.is_none(w_ndigits): return self.int(space) ndigits = space.bigint_w(space.index(w_ndigits)) # if ndigits >= 0 then no rounding is necessary; return self # unchanged if ndigits.ge(rbigint.fromint(0)): return self.int(space) # result = self - divmod_near(self, 10 ** -ndigits)[1] right = rbigint.fromint(10).pow(ndigits.neg()) w_tuple = divmod_near(space, self, newlong(space, right)) _, w_r = space.fixedview(w_tuple, 2) return space.sub(self, w_r) def _self_unaryop(opname, doc=None): @func_renamer('descr_' + opname) def descr_unaryop(self, space): return self.int(space) descr_unaryop.__doc__ = doc return descr_unaryop descr_conjugate = _self_unaryop( 'conjugate', "Returns self, the complex conjugate of any int.") descr_pos = _self_unaryop('pos', "x.__pos__() <==> +x") descr_index = _self_unaryop('index', "x[y:z] <==> x[y.__index__():z.__index__()]") descr_trunc = _self_unaryop('trunc', "Truncating an Integral returns itself.") descr_floor = _self_unaryop('floor', "Flooring an Integral returns itself.") descr_ceil = _self_unaryop('ceil', "Ceiling of an Integral returns itself.") descr_get_numerator = _self_unaryop('get_numerator') descr_get_real = _self_unaryop('get_real') def descr_get_denominator(self, space): return wrapint(space, 1) def descr_get_imag(self, space): return wrapint(space, 0) def int(self, space): """x.__int__() <==> int(x)""" raise NotImplementedError def asbigint(self): raise NotImplementedError def descr_format(self, space, w_format_spec): raise NotImplementedError def descr_pow(self, space, w_exponent, w_modulus=None): """x.__pow__(y[, z]) <==> pow(x, y[, z])""" raise NotImplementedError descr_rpow = func_with_new_name(descr_pow, 'descr_rpow') descr_rpow.__doc__ = "y.__rpow__(x[, z]) <==> pow(x, y[, z])" def _abstract_unaryop(opname, doc=SENTINEL): if doc is SENTINEL: doc = 'x.__%s__() <==> %s(x)' % (opname, opname) @func_renamer('descr_' + opname) def descr_unaryop(self, space): raise NotImplementedError descr_unaryop.__doc__ = doc return descr_unaryop descr_repr = _abstract_unaryop('repr') descr_str = _abstract_unaryop('str') descr_bit_length = _abstract_unaryop('bit_length', """\ int.bit_length() -> int Number of bits necessary to represent self in binary. >>> bin(37) '0b100101' >>> (37).bit_length() 6""") descr_bit_count = _abstract_unaryop('bit_count', """\ int.bit_count() -> int Number of bits set in the binary representation. >>> bin(37) '0b100101' >>> (37).bit_count() 3""") descr_hash = _abstract_unaryop('hash') descr_getnewargs = _abstract_unaryop('getnewargs', None) descr_float = _abstract_unaryop('float') descr_neg = _abstract_unaryop('neg', "x.__neg__() <==> -x") descr_abs = _abstract_unaryop('abs') descr_bool = _abstract_unaryop('bool', "x.__bool__() <==> x != 0") descr_invert = _abstract_unaryop('invert', "x.__invert__() <==> ~x") def _abstract_cmpop(opname): @func_renamer('descr_' + opname) def descr_cmp(self, space, w_other): raise NotImplementedError descr_cmp.__doc__ = 'x.__%s__(y) <==> x%sy' % (opname, CMP_OPS[opname]) return descr_cmp descr_lt = _abstract_cmpop('lt') descr_le = _abstract_cmpop('le') descr_eq = _abstract_cmpop('eq') descr_ne = _abstract_cmpop('ne') descr_gt = _abstract_cmpop('gt') descr_ge = _abstract_cmpop('ge') def _abstract_binop(opname): oper = BINARY_OPS.get(opname) if oper == '%': oper = '%%' oper = '%s(%%s, %%s)' % opname if not oper else '%%s%s%%s' % oper @func_renamer('descr_' + opname) def descr_binop(self, space, w_other): raise NotImplementedError descr_binop.__doc__ = "x.__%s__(y) <==> %s" % (opname, oper % ('x', 'y')) descr_rbinop = func_with_new_name(descr_binop, 'descr_r' + opname) descr_rbinop.__doc__ = "x.__r%s__(y) <==> %s" % (opname, oper % ('y', 'x')) return descr_binop, descr_rbinop descr_add, descr_radd = _abstract_binop('add') descr_sub, descr_rsub = _abstract_binop('sub') descr_mul, descr_rmul = _abstract_binop('mul') descr_matmul, descr_rmatmul = _abstract_binop('matmul') descr_and, descr_rand = _abstract_binop('and') descr_or, descr_ror = _abstract_binop('or') descr_xor, descr_rxor = _abstract_binop('xor') descr_lshift, descr_rlshift = _abstract_binop('lshift') descr_rshift, descr_rrshift = _abstract_binop('rshift') descr_floordiv, descr_rfloordiv = _abstract_binop('floordiv') descr_truediv, descr_rtruediv = _abstract_binop('truediv') descr_mod, descr_rmod = _abstract_binop('mod') descr_divmod, descr_rdivmod = _abstract_binop('divmod') def descr_as_integer_ratio(self, space): """ Return integer ratio. Return a pair of integers, whose ratio is exactly equal to the original int and with a positive denominator. >>> (10).as_integer_ratio() (10, 1) >>> (-10).as_integer_ratio() (-10, 1) >>> (0).as_integer_ratio() (0, 1) """ return space.newtuple([self.int(space), space.newint(1)]) def _floordiv(space, x, y): try: z = ovfcheck(x // y) except ZeroDivisionError: raise oefmt(space.w_ZeroDivisionError, "integer division or modulo by zero") return wrapint(space, z) def _truediv(space, x, y): if not y: raise oefmt(space.w_ZeroDivisionError, "division by zero") if (DBL_MANT_DIG < LONG_BIT and (r_uint(abs(x)) >> DBL_MANT_DIG or r_uint(abs(y)) >> DBL_MANT_DIG)): # large x or y, use long arithmetic raise OverflowError # both ints can be exactly represented as doubles, do a # floating-point division a = float(x) b = float(y) return space.newfloat(a / b) def _mod(space, x, y): try: z = ovfcheck(x % y) except ZeroDivisionError: raise oefmt(space.w_ZeroDivisionError, "integer modulo by zero") return wrapint(space, z) def _divmod(space, x, y): try: z = ovfcheck(x // y) except ZeroDivisionError: raise oefmt(space.w_ZeroDivisionError, "integer divmod by zero") # no overflow possible m = x % y return space.newtuple2(space.newint(z), space.newint(m)) def _divmod_ovf2small(space, x, y): from pypy.objspace.std.smalllongobject import W_SmallLongObject a = r_longlong(x) b = r_longlong(y) return space.newtuple2(W_SmallLongObject(a // b), W_SmallLongObject(a % b)) def _lshift(space, a, b): if r_uint(b) < LONG_BIT: # 0 <= b < LONG_BIT c = ovfcheck(a << b) return wrapint(space, c) if b < 0: raise oefmt(space.w_ValueError, "negative shift count") # b >= LONG_BIT if a == 0: return wrapint(space, a) raise OverflowError def _lshift_ovf2small(space, a, b): from pypy.objspace.std.smalllongobject import W_SmallLongObject w_a = W_SmallLongObject.fromint(a) w_b = W_SmallLongObject.fromint(b) return w_a.descr_lshift(space, w_b) def _rshift(space, a, b): if r_uint(b) >= LONG_BIT: # not (0 <= b < LONG_BIT) if b < 0: raise oefmt(space.w_ValueError, "negative shift count") # b >= LONG_BIT if a == 0: return wrapint(space, a) a = -1 if a < 0 else 0 else: a = a >> b return wrapint(space, a) def _pow(space, iv, iw, iz): """Helper for pow""" if iz == 0: return _pow_nomod(iv, iw) else: return _pow_mod(space, iv, iw, iz) @jit.look_inside_iff(lambda iv, iw: jit.isconstant(iw)) def _pow_nomod(iv, iw): if iw <= 0: if iw == 0: return 1 # bounce it, since it always returns float raise ValueError temp = iv ix = 1 while True: if iw & 1: try: ix = ovfcheck(ix * temp) except OverflowError: raise iw >>= 1 # Shift exponent down by 1 bit if iw == 0: break try: temp = ovfcheck(temp * temp) # Square the value of temp except OverflowError: raise return ix @jit.look_inside_iff(lambda space, iv, iw, iz: jit.isconstant(iw) and jit.isconstant(iz)) def _pow_mod(space, iv, iw, iz): from rpython.rlib.rarithmetic import mulmod if iw == 0: return 1 % iz # != 1, for iz == 1 or iz < 0 if iz < 0: try: iz = ovfcheck(-iz) except OverflowError: raise iz_negative = True else: iz_negative = False if iw <= 0: w_iv = invmod(space, space.newint(iv), space.newint(iz)) assert isinstance(w_iv, W_IntObject) iv = w_iv.intval try: iw = ovfcheck(-iw) except OverflowError: raise temp = iv ix = 1 while True: if iw & 1: ix = mulmod(ix, temp, iz) iw >>= 1 # Shift exponent down by 1 bit if iw == 0: break temp = mulmod(temp, temp, iz) if iz_negative and ix > 0: ix -= iz return ix def _pow_ovf2long(space, iv, w_iv, iw, w_iw, w_modulus): if space.is_none(w_modulus) and _recover_with_smalllong(space): from pypy.objspace.std.smalllongobject import _pow as _pow_small try: # XXX: shouldn't have to pass r_longlong(0) here (see # 4fa4c6b93a84) return _pow_small(space, r_longlong(iv), iw, r_longlong(0)) except (OverflowError, ValueError): pass from pypy.objspace.std.longobject import W_LongObject, W_AbstractLongObject if w_iv is None or not isinstance(w_iv, W_AbstractLongObject): w_iv = W_LongObject.fromint(space, iv) if w_iw is None or not isinstance(w_iw, W_AbstractLongObject): w_iw = W_LongObject.fromint(space, iw) return w_iv.descr_pow(space, w_iw, w_modulus) def _make_ovf2long(opname, ovf2small=None): op = getattr(operator, opname, None) assert op or ovf2small def ovf2long(space, x, w_x, y, w_y): """Handle overflowing to smalllong or long""" if _recover_with_smalllong(space): if ovf2small: return ovf2small(space, x, y) # Assume a generic operation without an explicit ovf2small # handler from pypy.objspace.std.smalllongobject import W_SmallLongObject a = r_longlong(x) b = r_longlong(y) return W_SmallLongObject(op(a, b)) if opname == 'mul': return space.newlong_from_rbigint(rbigint.mul_int_int_bigint_result(x, y)) from pypy.objspace.std.longobject import W_LongObject, W_AbstractLongObject w_x = W_LongObject.fromint(space, x) assert w_y is not None # call the W_LongObject implementation with the unconverted w_y. # W_LongObject can deal with W_IntObject arguments just fine, and it # has a slightly better code path for long/int combinations return getattr(w_x, 'descr_' + opname)(space, w_y) return ovf2long @jit.elidable def _bit_length(val): bits = 0 if val < 0: # warning, "-val" overflows here val = -((val + 1) >> 1) bits = 1 while val: bits += 1 val >>= 1 return bits @jit.elidable def _bit_count(val): if val == -sys.maxint - 1: return 1 elif val < 0: val = -val count = 0 while val: count += val & 1 val >>= 1 return count class W_IntObject(W_AbstractIntObject): __slots__ = 'intval' _immutable_fields_ = ['intval'] def __init__(self, intval): assert is_valid_int(intval) self.intval = int(intval) def __repr__(self): """representation for debugging purposes""" return "%s(%d)" % (self.__class__.__name__, self.intval) def is_w(self, space, w_other): from pypy.objspace.std.boolobject import W_BoolObject if (not isinstance(w_other, W_AbstractIntObject) or isinstance(w_other, W_BoolObject)): return False if self.user_overridden_class or w_other.user_overridden_class: return self is w_other x = self.intval try: y = space.int_w(w_other) except OperationError as e: if e.match(space, space.w_OverflowError): return False raise return x == y def int_w(self, space, allow_conversion=True): return self.intval def _int_w(self, space): return self.intval unwrap = _int_w def uint_w(self, space): intval = self.intval if intval < 0: raise oefmt(space.w_ValueError, "cannot convert negative integer to unsigned") return r_uint(intval) def bigint_w(self, space, allow_conversion=True): return self.asbigint() def _bigint_w(self, space): return self.asbigint() def float_w(self, space, allow_conversion=True): return float(self.intval) # note that we do NOT implement _float_w, because __float__ cannot return # an int def int(self, space): if type(self) is W_IntObject: return self if not space.is_overloaded(self, space.w_int, '__int__'): return space.newint(self.intval) return W_Root.int(self, space) def asbigint(self): return rbigint.fromint(self.intval) @staticmethod @unwrap_spec(w_x=WrappedDefault(0)) def descr_new(space, w_inttype, w_x, __posonly__, w_base=None): "Create and return a new object. See help(type) for accurate signature." return _new_int(space, w_inttype, w_x, w_base) def descr_hash(self, space): return space.newint(_hash_int(self.intval)) def as_w_long(self, space): from pypy.objspace.std.longobject import W_LongObject from pypy.objspace.std.smalllongobject import W_SmallLongObject if space.config.objspace.std.withsmalllong: return W_SmallLongObject.fromint(self.intval) return W_LongObject.fromint(space, self.intval) def descr_bool(self, space): return space.newbool(self.intval != 0) def descr_invert(self, space): return wrapint(space, ~self.intval) def descr_neg(self, space): a = self.intval if a == MININT: if _recover_with_smalllong(space): from pypy.objspace.std.smalllongobject import W_SmallLongObject x = r_longlong(a) return W_SmallLongObject(-x) return self.as_w_long(space).descr_neg(space) return wrapint(space, -a) def descr_abs(self, space): if type(self) is W_IntObject: if self.intval == MININT: return self.descr_neg(space) # branchless version x = self.intval mask = x >> (LONG_BIT - 1) return space.newint((x ^ mask) - mask) pos = self.intval >= 0 return self.int(space) if pos else self.descr_neg(space) def descr_float(self, space): a = self.intval x = float(a) return space.newfloat(x) def descr_getnewargs(self, space): return space.newtuple([wrapint(space, self.intval)]) def descr_bit_length(self, space): return space.newint(_bit_length(self.intval)) def descr_bit_count(self, space): return space.newint(_bit_count(self.intval)) def descr_repr(self, space): res = str(self.intval) return space.newutf8(res, len(res)) # res is always ASCII descr_str = func_with_new_name(descr_repr, 'descr_str') def descr_format(self, space, w_format_spec): return newformat.run_formatter(space, w_format_spec, "format_int_or_long", self, newformat.INT_KIND) @unwrap_spec(w_modulus=WrappedDefault(None)) def descr_pow(self, space, w_exponent, w_modulus=None): if isinstance(w_exponent, W_IntObject): y = w_exponent.intval elif isinstance(w_exponent, W_AbstractIntObject): self = self.as_w_long(space) return self.descr_pow(space, w_exponent, w_modulus) else: return space.w_NotImplemented x = self.intval y = w_exponent.intval if space.is_none(w_modulus): z = 0 elif isinstance(w_modulus, W_IntObject): z = w_modulus.intval if z == 0: raise oefmt(space.w_ValueError, "pow() 3rd argument cannot be 0") else: # can't return NotImplemented (space.pow doesn't do full # ternary, i.e. w_modulus.__zpow__(self, w_exponent)), so # handle it ourselves return _pow_ovf2long(space, x, self, y, w_exponent, w_modulus) try: result = _pow(space, x, y, z) except OverflowError: return _pow_ovf2long(space, x, self, y, w_exponent, w_modulus) except ValueError: # float result, so let avoid a roundtrip in rbigint. self = self.descr_float(space) w_exponent = w_exponent.descr_float(space) return space.pow(self, w_exponent, space.w_None) return space.newint(result) @unwrap_spec(w_modulus=WrappedDefault(None)) def descr_rpow(self, space, w_base, w_modulus=None): if isinstance(w_base, W_IntObject): return w_base.descr_pow(space, self, w_modulus) elif isinstance(w_base, W_AbstractIntObject): self = self.as_w_long(space) return self.descr_rpow(space, w_base, w_modulus) return space.w_NotImplemented def _make_descr_cmp(opname): op = getattr(operator, opname) descr_name = 'descr_' + opname if opname == 'lt': revopname = 'gt' elif opname == 'le': revopname = 'ge' elif opname == 'eq': revopname = 'eq' elif opname == 'ne': revopname = 'ne' elif opname == 'gt': revopname = 'lt' elif opname == 'ge': revopname = 'le' bigintintrevop = getattr(rbigint, "int_" + revopname) @func_renamer(descr_name) def descr_cmp(self, space, w_other): from pypy.objspace.std.longobject import W_LongObject, W_AbstractLongObject if isinstance(w_other, W_IntObject): i = self.intval j = w_other.intval return space.newbool(op(i, j)) elif isinstance(w_other, W_AbstractIntObject): if not space.config.objspace.std.withsmalllong: assert isinstance(w_other, W_LongObject) return space.newbool(bigintintrevop(w_other.asbigint(), self.intval)) return space.w_NotImplemented return descr_cmp descr_lt = _make_descr_cmp('lt') descr_le = _make_descr_cmp('le') descr_eq = _make_descr_cmp('eq') descr_ne = _make_descr_cmp('ne') descr_gt = _make_descr_cmp('gt') descr_ge = _make_descr_cmp('ge') def _make_generic_descr_binop(opname, ovf=True): op = getattr(operator, opname + '_' if opname in ('and', 'or') else opname) descr_name, descr_rname = 'descr_' + opname, 'descr_r' + opname if ovf: ovf2long = _make_ovf2long(opname) @func_renamer(descr_name) def descr_binop(self, space, w_other): if isinstance(w_other, W_IntObject): x = self.intval y = w_other.intval if ovf: try: z = ovfcheck(op(x, y)) except OverflowError: return ovf2long(space, x, self, y, w_other) else: z = op(x, y) return wrapint(space, z) elif isinstance(w_other, W_AbstractIntObject): self = self.as_w_long(space) return getattr(self, descr_name)(space, w_other) return space.w_NotImplemented if opname in COMMUTATIVE_OPS: @func_renamer(descr_rname) def descr_rbinop(self, space, w_other): return descr_binop(self, space, w_other) return descr_binop, descr_rbinop @func_renamer(descr_rname) def descr_rbinop(self, space, w_other): if isinstance(w_other, W_IntObject): x = self.intval y = w_other.intval if ovf: try: z = ovfcheck(op(y, x)) except OverflowError: return ovf2long(space, y, w_other, x, self) # XXX write a test else: z = op(y, x) return wrapint(space, z) elif isinstance(w_other, W_AbstractIntObject): self = self.as_w_long(space) return getattr(self, descr_rname)(space, w_other) return space.w_NotImplemented return descr_binop, descr_rbinop descr_add, descr_radd = _make_generic_descr_binop('add') descr_sub, descr_rsub = _make_generic_descr_binop('sub') descr_mul, descr_rmul = _make_generic_descr_binop('mul') descr_and, descr_rand = _make_generic_descr_binop('and', ovf=False) descr_or, descr_ror = _make_generic_descr_binop('or', ovf=False) descr_xor, descr_rxor = _make_generic_descr_binop('xor', ovf=False) def _make_descr_binop(func, ovf=True, ovf2small=None, ovf_func=None): opname = func.__name__[1:] descr_name, descr_rname = 'descr_' + opname, 'descr_r' + opname if ovf: if ovf_func: ovf2long = ovf_func assert not ovf2small # must be part of ovf_func else: ovf2long = _make_ovf2long(opname, ovf2small) @func_renamer(descr_name) def descr_binop(self, space, w_other): if isinstance(w_other, W_IntObject): x = self.intval y = w_other.intval if ovf: try: return func(space, x, y) except OverflowError: return ovf2long(space, x, self, y, w_other) else: return func(space, x, y) elif isinstance(w_other, W_AbstractIntObject): self = self.as_w_long(space) return getattr(self, descr_name)(space, w_other) return space.w_NotImplemented @func_renamer(descr_rname) def descr_rbinop(self, space, w_other): if isinstance(w_other, W_IntObject): x = self.intval y = w_other.intval if ovf: try: return func(space, y, x) except OverflowError: return ovf2long(space, y, w_other, x, self) else: return func(space, y, x) elif isinstance(w_other, W_AbstractIntObject): self = self.as_w_long(space) return getattr(self, descr_rname)(space, w_other) return space.w_NotImplemented return descr_binop, descr_rbinop def _ovf2long_lshift(space, x, w_x, y, w_y): if _recover_with_smalllong(space): return _lshift_ovf2small(space, x, y) from pypy.objspace.std.longobject import W_LongObject, W_AbstractLongObject if w_x is None or not isinstance(w_x, W_AbstractLongObject): w_x = W_LongObject.fromint(space, x) # crucially, *don't* convert w_y to W_LongObject, it will just be # converted back (huge lshifts always overflow) return w_x._int_lshift(space, y) descr_lshift, descr_rlshift = _make_descr_binop( _lshift, ovf_func=_ovf2long_lshift) descr_rshift, descr_rrshift = _make_descr_binop(_rshift, ovf=False) descr_floordiv, descr_rfloordiv = _make_descr_binop(_floordiv) descr_truediv, descr_rtruediv = _make_descr_binop(_truediv) descr_mod, descr_rmod = _make_descr_binop(_mod) descr_divmod, descr_rdivmod = _make_descr_binop( _divmod, ovf2small=_divmod_ovf2small) def setup_prebuilt(space): if space.config.objspace.std.withprebuiltint: W_IntObject.PREBUILT = [] for i in range(space.config.objspace.std.prebuiltintfrom, space.config.objspace.std.prebuiltintto): W_IntObject.PREBUILT.append(W_IntObject(i)) else: W_IntObject.PREBUILT = None def wrapint(space, x): if not space.config.objspace.std.withprebuiltint: return W_IntObject(x) lower = space.config.objspace.std.prebuiltintfrom upper = space.config.objspace.std.prebuiltintto # use r_uint to perform a single comparison (this whole function is # getting inlined into every caller so keeping the branching to a # minimum is a good idea) index = r_uint(x) - r_uint(lower) if index >= r_uint(upper - lower): w_res = instantiate(W_IntObject) else: w_res = W_IntObject.PREBUILT[index] # obscure hack to help the CPU cache: we store 'x' even into a # prebuilt integer's intval. This makes sure that the intval field # is present in the cache in the common case where it is quickly # reused. (we could use a prefetch hint if we had that) w_res.intval = x return w_res divmod_near = applevel(''' def divmod_near(a, b): """Return a pair (q, r) such that a = b * q + r, and abs(r) <= abs(b)/2, with equality possible only if q is even. In other words, q == a / b, rounded to the nearest integer using round-half-to-even.""" q, r = divmod(a, b) # round up if either r / b > 0.5, or r / b == 0.5 and q is # odd. The expression r / b > 0.5 is equivalent to 2 * r > b # if b is positive, 2 * r < b if b negative. greater_than_half = 2*r > b if b > 0 else 2*r < b exactly_half = 2*r == b if greater_than_half or exactly_half and q % 2 == 1: q += 1 r -= b return q, r ''', filename=__file__).interphook('divmod_near') invmod = applevel(''' def invmod(a, n): if n == 1: return 0 if a == 1: return 1 assert n >= 0 b, c = 1, 0 while n: q, r = divmod(a, n) a, b, c, n = n, c, b - q*c, r # at this point a is the gcd of the original inputs if a == 1: return b raise ValueError("base is not invertible for the given modulus") ''', filename=__file__).interphook("invmod") def _recover_with_smalllong(space): """True if there is a chance that a SmallLong would fit when an Int does not """ return (space.config.objspace.std.withsmalllong and sys.maxint == 2147483647) def _string_to_int_or_long(space, w_source, string, base=10): from pypy.module.sys.state import get_int_max_str_digits if (base & (base - 1) != 0): # Limit the size to avoid excessive computation attacks on non-binary bases max_str_digits = space.int_w(get_int_max_str_digits(space)) else: max_str_digits = 0 try: value = string_to_int(string, base, allow_underscores=True, no_implicit_octal=True, max_str_digits=max_str_digits) return wrapint(space, value) except MaxDigitsError as e: raise oefmt(space.w_ValueError, "Exceeds the limit (%d) for integer string conversion: value has %d digits", max_str_digits, e.digits) except ParseStringError as e: raise wrap_parsestringerror(space, e, w_source) except ParseStringOverflowError as e: return _retry_to_w_long(space, e.parser, w_source) def _retry_to_w_long(space, parser, w_source): from pypy.objspace.std.longobject import newbigint parser.rewind() try: bigint = rbigint._from_numberstring_parser(parser) except ParseStringError as e: raise wrap_parsestringerror(space, e, w_source) return newbigint(space, space.w_int, bigint) def _new_int(space, w_inttype, w_x, w_base=None): w_value = w_x # 'x' is the keyword argument name in CPython if w_inttype is space.w_int: return _new_baseint(space, w_x, w_base) else: w_tmp = _new_baseint(space, w_x, w_base) return _as_subint(space, w_inttype, w_tmp) def _new_baseint(space, w_value, w_base=None): if w_base is None: if space.is_w(space.type(w_value), space.w_int): assert isinstance(w_value, W_AbstractIntObject) return w_value elif space.lookup(w_value, '__int__') is not None: w_intvalue = space.int(w_value) return _ensure_baseint(space, w_intvalue) elif space.lookup(w_value, '__trunc__') is not None: space.warn(space.newtext( "The delegation of int() to __trunc__ is deprecated."), space.w_DeprecationWarning) w_obj = space.trunc(w_value) if not space.isinstance_w(w_obj, space.w_int): try: w_obj = space.index(w_obj) except OperationError as e: if not e.match(space, space.w_TypeError): raise raise oefmt( space.w_TypeError, "__trunc__ returned non-Integral (type '%T')", w_obj ) assert isinstance(w_obj, W_AbstractIntObject) return _ensure_baseint(space, w_obj) elif space.lookup(w_value, '__index__') is not None: w_obj = space.index(w_value) if not space.is_w(space.type(w_obj), space.w_int): if space.isinstance_w(w_obj, space.w_int): w_obj = space.int(w_obj) else: raise oefmt(space.w_TypeError, "int() argument must be a string, a bytes-like " "object or a number, not '%T'", w_value) assert isinstance(w_obj, W_AbstractIntObject) return _ensure_baseint(space, w_obj) elif space.isinstance_w(w_value, space.w_unicode): from pypy.objspace.std.unicodeobject import unicode_to_decimal_w try: b = unicode_to_decimal_w(space, w_value) except Exception: raise oefmt(space.w_ValueError, 'invalid literal for int() with base 10: %R', w_value) return _string_to_int_or_long(space, w_value, b) elif (space.isinstance_w(w_value, space.w_bytearray) or space.isinstance_w(w_value, space.w_bytes)): return _string_to_int_or_long(space, w_value, space.charbuf_w(w_value)) else: # If object supports the buffer interface try: buf = space.charbuf_w(w_value) except OperationError as e: if not e.match(space, space.w_TypeError): raise raise oefmt(space.w_TypeError, "int() argument must be a string, a bytes-like " "object or a number, not '%T'", w_value) else: return _string_to_int_or_long(space, w_value, buf) else: try: base = space.getindex_w(w_base, None) except OperationError as e: if not e.match(space, space.w_OverflowError): raise base = 37 # this raises the right error in string_to_bigint() if space.isinstance_w(w_value, space.w_unicode): from pypy.objspace.std.unicodeobject import unicode_to_decimal_w try: s = unicode_to_decimal_w(space, w_value) except Exception: raise oefmt(space.w_ValueError, 'invalid literal for int() with base %d: %R', base, w_value) elif (space.isinstance_w(w_value, space.w_bytes) or space.isinstance_w(w_value, space.w_bytearray)): s = space.charbuf_w(w_value) else: raise oefmt(space.w_TypeError, "int() can't convert non-string with explicit base") return _string_to_int_or_long(space, w_value, s, base) @enforceargs(None, None, W_AbstractIntObject, typecheck=False) def _as_subint(space, w_inttype, w_value): from pypy.objspace.std.longobject import W_LongObject, newbigint if space.config.objspace.std.withsmalllong: from pypy.objspace.std.smalllongobject import W_SmallLongObject else: W_SmallLongObject = None if type(w_value) is W_IntObject: w_obj = space.allocate_instance(W_IntObject, w_inttype) W_IntObject.__init__(w_obj, w_value.intval) return w_obj elif type(w_value) is W_LongObject: return newbigint(space, w_inttype, w_value.num) elif W_SmallLongObject and type(w_value) is W_SmallLongObject: return newbigint(space, w_inttype, space.bigint_w(w_value)) @enforceargs(None, W_AbstractIntObject, typecheck=False) def _ensure_baseint(space, w_intvalue): from pypy.objspace.std.longobject import ( W_LongObject, W_AbstractLongObject, newlong) if isinstance(w_intvalue, W_IntObject): if type(w_intvalue) is not W_IntObject: w_intvalue = wrapint(space, w_intvalue.intval) return w_intvalue elif isinstance(w_intvalue, W_AbstractLongObject): if type(w_intvalue) is not W_LongObject: w_intvalue = newlong(space, w_intvalue.asbigint()) return w_intvalue else: # shouldn't happen raise oefmt(space.w_RuntimeError, "internal error in int.__new__()") W_AbstractIntObject.typedef = TypeDef("int", __doc__ = """int([x]) -> integer int(x, base=10) -> integer Convert a number or string to an integer, or return 0 if no arguments are given. If x is a number, return x.__int__(). For floating point numbers, this truncates towards zero. If x is not a number or if base is given, then x must be a string, bytes, or bytearray instance representing an integer literal in the given base. The literal can be preceded by '+' or '-' and be surrounded by whitespace. The base defaults to 10. Valid bases are 0 and 2-36. Base 0 means to interpret the base from the string as an integer literal. >>> int('0b100', base=0) 4""", __new__ = interp2app(W_IntObject.descr_new), numerator = typedef.GetSetProperty( W_AbstractIntObject.descr_get_numerator, doc="the numerator of a rational number in lowest terms"), denominator = typedef.GetSetProperty( W_AbstractIntObject.descr_get_denominator, doc="the denominator of a rational number in lowest terms"), real = typedef.GetSetProperty( W_AbstractIntObject.descr_get_real, doc="the real part of a complex number"), imag = typedef.GetSetProperty( W_AbstractIntObject.descr_get_imag, doc="the imaginary part of a complex number"), as_integer_ratio = interp2app(W_AbstractIntObject.descr_as_integer_ratio), from_bytes = interp2app(W_AbstractIntObject.descr_from_bytes, as_classmethod=True), to_bytes = interpindirect2app(W_AbstractIntObject.descr_to_bytes), __repr__ = interpindirect2app(W_AbstractIntObject.descr_repr), __str__ = interpindirect2app(W_AbstractIntObject.descr_str), conjugate = interpindirect2app(W_AbstractIntObject.descr_conjugate), bit_length = interpindirect2app(W_AbstractIntObject.descr_bit_length), bit_count = interpindirect2app(W_AbstractIntObject.descr_bit_count), __format__ = interpindirect2app(W_AbstractIntObject.descr_format), __hash__ = interpindirect2app(W_AbstractIntObject.descr_hash), __getnewargs__ = interpindirect2app(W_AbstractIntObject.descr_getnewargs), __int__ = interpindirect2app(W_AbstractIntObject.int), __index__ = interpindirect2app(W_AbstractIntObject.descr_index), __trunc__ = interpindirect2app(W_AbstractIntObject.descr_trunc), __float__ = interpindirect2app(W_AbstractIntObject.descr_float), __round__ = interpindirect2app(W_AbstractIntObject.descr_round), __pos__ = interpindirect2app(W_AbstractIntObject.descr_pos), __neg__ = interpindirect2app(W_AbstractIntObject.descr_neg), __abs__ = interpindirect2app(W_AbstractIntObject.descr_abs), __bool__ = interpindirect2app(W_AbstractIntObject.descr_bool), __invert__ = interpindirect2app(W_AbstractIntObject.descr_invert), __floor__ = interpindirect2app(W_AbstractIntObject.descr_floor), __ceil__ = interpindirect2app(W_AbstractIntObject.descr_ceil), __lt__ = interpindirect2app(W_AbstractIntObject.descr_lt), __le__ = interpindirect2app(W_AbstractIntObject.descr_le), __eq__ = interpindirect2app(W_AbstractIntObject.descr_eq), __ne__ = interpindirect2app(W_AbstractIntObject.descr_ne), __gt__ = interpindirect2app(W_AbstractIntObject.descr_gt), __ge__ = interpindirect2app(W_AbstractIntObject.descr_ge), __add__ = interpindirect2app(W_AbstractIntObject.descr_add), __radd__ = interpindirect2app(W_AbstractIntObject.descr_radd), __sub__ = interpindirect2app(W_AbstractIntObject.descr_sub), __rsub__ = interpindirect2app(W_AbstractIntObject.descr_rsub), __mul__ = interpindirect2app(W_AbstractIntObject.descr_mul), __rmul__ = interpindirect2app(W_AbstractIntObject.descr_rmul), __and__ = interpindirect2app(W_AbstractIntObject.descr_and), __rand__ = interpindirect2app(W_AbstractIntObject.descr_rand), __or__ = interpindirect2app(W_AbstractIntObject.descr_or), __ror__ = interpindirect2app(W_AbstractIntObject.descr_ror), __xor__ = interpindirect2app(W_AbstractIntObject.descr_xor), __rxor__ = interpindirect2app(W_AbstractIntObject.descr_rxor), __lshift__ = interpindirect2app(W_AbstractIntObject.descr_lshift), __rlshift__ = interpindirect2app(W_AbstractIntObject.descr_rlshift), __rshift__ = interpindirect2app(W_AbstractIntObject.descr_rshift), __rrshift__ = interpindirect2app(W_AbstractIntObject.descr_rrshift), __floordiv__ = interpindirect2app(W_AbstractIntObject.descr_floordiv), __rfloordiv__ = interpindirect2app(W_AbstractIntObject.descr_rfloordiv), __truediv__ = interpindirect2app(W_AbstractIntObject.descr_truediv), __rtruediv__ = interpindirect2app(W_AbstractIntObject.descr_rtruediv), __mod__ = interpindirect2app(W_AbstractIntObject.descr_mod), __rmod__ = interpindirect2app(W_AbstractIntObject.descr_rmod), __divmod__ = interpindirect2app(W_AbstractIntObject.descr_divmod), __rdivmod__ = interpindirect2app(W_AbstractIntObject.descr_rdivmod), __pow__ = interpindirect2app(W_AbstractIntObject.descr_pow), __rpow__ = interpindirect2app(W_AbstractIntObject.descr_rpow), ) def _hash_int(a): # write it without branches for the benefit of the jit # it needs to give numbers that are equivalent to the hash function of # rbigint # compute the sign sign = 1 - ((a < 0) << 1) x = r_uint(a) # take absolute value x *= r_uint(sign) # efficient x % HASH_MODULUS: as HASH_MODULUS is a Mersenne # prime x = (x & HASH_MODULUS) + (x >> HASH_BITS) # if x >= HASH_MODULUS: # x -= HASH_MODULUS x -= HASH_MODULUS * (x >= HASH_MODULUS) h = intmask(intmask(x) * sign) return h - (h == -1)