diff options
Diffstat (limited to 'Objects')
-rw-r--r-- | Objects/abstract.c | 14 | ||||
-rw-r--r-- | Objects/classobject.c | 2 | ||||
-rw-r--r-- | Objects/dictobject.c | 24 | ||||
-rw-r--r-- | Objects/exceptions.c | 52 | ||||
-rw-r--r-- | Objects/fileobject.c | 33 | ||||
-rw-r--r-- | Objects/intobject.c | 42 | ||||
-rw-r--r-- | Objects/listobject.c | 31 | ||||
-rw-r--r-- | Objects/longobject.c | 71 | ||||
-rw-r--r-- | Objects/obmalloc.c | 3 | ||||
-rw-r--r-- | Objects/setobject.c | 123 | ||||
-rw-r--r-- | Objects/stringlib/partition.h | 8 | ||||
-rw-r--r-- | Objects/stringobject.c | 45 | ||||
-rw-r--r-- | Objects/tupleobject.c | 5 | ||||
-rw-r--r-- | Objects/typeobject.c | 23 | ||||
-rw-r--r-- | Objects/unicodeobject.c | 52 |
15 files changed, 366 insertions, 162 deletions
diff --git a/Objects/abstract.c b/Objects/abstract.c index 8810762..03883be 100644 --- a/Objects/abstract.c +++ b/Objects/abstract.c @@ -1524,20 +1524,18 @@ _PySequence_IterSearch(PyObject *seq, PyObject *obj, int operation) if (cmp > 0) { switch (operation) { case PY_ITERSEARCH_COUNT: - ++n; - if (n <= 0) { - /* XXX(nnorwitz): int means ssize_t */ + if (n == PY_SSIZE_T_MAX) { PyErr_SetString(PyExc_OverflowError, - "count exceeds C int size"); + "count exceeds C integer size"); goto Fail; } + ++n; break; case PY_ITERSEARCH_INDEX: if (wrapped) { - /* XXX(nnorwitz): int means ssize_t */ PyErr_SetString(PyExc_OverflowError, - "index exceeds C int size"); + "index exceeds C integer size"); goto Fail; } goto Done; @@ -1552,9 +1550,9 @@ _PySequence_IterSearch(PyObject *seq, PyObject *obj, int operation) } if (operation == PY_ITERSEARCH_INDEX) { - ++n; - if (n <= 0) + if (n == PY_SSIZE_T_MAX) wrapped = 1; + ++n; } } diff --git a/Objects/classobject.c b/Objects/classobject.c index 1107977..85826a9 100644 --- a/Objects/classobject.c +++ b/Objects/classobject.c @@ -152,6 +152,8 @@ method_new(PyTypeObject* type, PyObject* args, PyObject *kw) PyObject *self; PyObject *classObj = NULL; + if (!_PyArg_NoKeywords("instancemethod", kw)) + return NULL; if (!PyArg_UnpackTuple(args, "method", 2, 3, &func, &self, &classObj)) return NULL; diff --git a/Objects/dictobject.c b/Objects/dictobject.c index 64585a4..09aba61 100644 --- a/Objects/dictobject.c +++ b/Objects/dictobject.c @@ -12,6 +12,20 @@ typedef PyDictEntry dictentry; typedef PyDictObject dictobject; +/* Set a key error with the specified argument, wrapping it in a + * tuple automatically so that tuple keys are not unpacked as the + * exception arguments. */ +static void +set_key_error(PyObject *arg) +{ + PyObject *tup; + tup = PyTuple_Pack(1, arg); + if (!tup) + return; /* caller will expect error to be set anyway */ + PyErr_SetObject(PyExc_KeyError, tup); + Py_DECREF(tup); +} + /* Define this out if you don't want conversion statistics on exit. */ #undef SHOW_CONVERSION_COUNTS @@ -307,6 +321,8 @@ lookdict(dictobject *mp, PyObject *key, register long hash) else if (ep->me_key == dummy && freeslot == NULL) freeslot = ep; } + assert(0); /* NOT REACHED */ + return 0; } /* @@ -366,6 +382,8 @@ lookdict_string(dictobject *mp, PyObject *key, register long hash) if (ep->me_key == dummy && freeslot == NULL) freeslot = ep; } + assert(0); /* NOT REACHED */ + return 0; } /* @@ -691,7 +709,7 @@ PyDict_DelItem(PyObject *op, PyObject *key) if (ep == NULL) return -1; if (ep->me_value == NULL) { - PyErr_SetObject(PyExc_KeyError, key); + set_key_error(key); return -1; } old_key = ep->me_key; @@ -1000,7 +1018,7 @@ dict_subscript(dictobject *mp, register PyObject *key) return PyObject_CallFunctionObjArgs(missing, (PyObject *)mp, key, NULL); } - PyErr_SetObject(PyExc_KeyError, key); + set_key_error(key); return NULL; } else @@ -1644,7 +1662,7 @@ dict_pop(dictobject *mp, PyObject *args) Py_INCREF(deflt); return deflt; } - PyErr_SetObject(PyExc_KeyError, key); + set_key_error(key); return NULL; } old_key = ep->me_key; diff --git a/Objects/exceptions.c b/Objects/exceptions.c index c3ead69..0cd819c 100644 --- a/Objects/exceptions.c +++ b/Objects/exceptions.c @@ -81,6 +81,7 @@ BaseException_clear(PyBaseExceptionObject *self) static void BaseException_dealloc(PyBaseExceptionObject *self) { + _PyObject_GC_UNTRACK(self); BaseException_clear(self); self->ob_type->tp_free((PyObject *)self); } @@ -174,27 +175,10 @@ BaseException_setstate(PyObject *self, PyObject *state) Py_RETURN_NONE; } -#ifdef Py_USING_UNICODE -/* while this method generates fairly uninspired output, it a least - * guarantees that we can display exceptions that have unicode attributes - */ -static PyObject * -BaseException_unicode(PyBaseExceptionObject *self) -{ - if (PyTuple_GET_SIZE(self->args) == 0) - return PyUnicode_FromUnicode(NULL, 0); - if (PyTuple_GET_SIZE(self->args) == 1) - return PyObject_Unicode(PyTuple_GET_ITEM(self->args, 0)); - return PyObject_Unicode(self->args); -} -#endif /* Py_USING_UNICODE */ static PyMethodDef BaseException_methods[] = { {"__reduce__", (PyCFunction)BaseException_reduce, METH_NOARGS }, {"__setstate__", (PyCFunction)BaseException_setstate, METH_O }, -#ifdef Py_USING_UNICODE - {"__unicode__", (PyCFunction)BaseException_unicode, METH_NOARGS }, -#endif {NULL, NULL, 0, NULL}, }; @@ -206,12 +190,19 @@ BaseException_getitem(PyBaseExceptionObject *self, Py_ssize_t index) return PySequence_GetItem(self->args, index); } +static PyObject * +BaseException_getslice(PyBaseExceptionObject *self, + Py_ssize_t start, Py_ssize_t stop) +{ + return PySequence_GetSlice(self->args, start, stop); +} + static PySequenceMethods BaseException_as_sequence = { 0, /* sq_length; */ 0, /* sq_concat; */ 0, /* sq_repeat; */ (ssizeargfunc)BaseException_getitem, /* sq_item; */ - 0, /* sq_slice; */ + (ssizessizeargfunc)BaseException_getslice, /* sq_slice; */ 0, /* sq_ass_item; */ 0, /* sq_ass_slice; */ 0, /* sq_contains; */ @@ -456,6 +447,7 @@ SystemExit_clear(PySystemExitObject *self) static void SystemExit_dealloc(PySystemExitObject *self) { + _PyObject_GC_UNTRACK(self); SystemExit_clear(self); self->ob_type->tp_free((PyObject *)self); } @@ -518,7 +510,7 @@ EnvironmentError_init(PyEnvironmentErrorObject *self, PyObject *args, if (BaseException_init((PyBaseExceptionObject *)self, args, kwds) == -1) return -1; - if (PyTuple_GET_SIZE(args) <= 1) { + if (PyTuple_GET_SIZE(args) <= 1 || PyTuple_GET_SIZE(args) > 3) { return 0; } @@ -562,6 +554,7 @@ EnvironmentError_clear(PyEnvironmentErrorObject *self) static void EnvironmentError_dealloc(PyEnvironmentErrorObject *self) { + _PyObject_GC_UNTRACK(self); EnvironmentError_clear(self); self->ob_type->tp_free((PyObject *)self); } @@ -760,6 +753,7 @@ WindowsError_clear(PyWindowsErrorObject *self) static void WindowsError_dealloc(PyWindowsErrorObject *self) { + _PyObject_GC_UNTRACK(self); WindowsError_clear(self); self->ob_type->tp_free((PyObject *)self); } @@ -834,9 +828,9 @@ WindowsError_str(PyWindowsErrorObject *self) return NULL; } - if (self->myerrno) { - Py_INCREF(self->myerrno); - PyTuple_SET_ITEM(tuple, 0, self->myerrno); + if (self->winerror) { + Py_INCREF(self->winerror); + PyTuple_SET_ITEM(tuple, 0, self->winerror); } else { Py_INCREF(Py_None); @@ -858,7 +852,7 @@ WindowsError_str(PyWindowsErrorObject *self) Py_DECREF(fmt); Py_DECREF(tuple); } - else if (self->myerrno && self->strerror) { + else if (self->winerror && self->strerror) { PyObject *fmt; PyObject *tuple; @@ -872,9 +866,9 @@ WindowsError_str(PyWindowsErrorObject *self) return NULL; } - if (self->myerrno) { - Py_INCREF(self->myerrno); - PyTuple_SET_ITEM(tuple, 0, self->myerrno); + if (self->winerror) { + Py_INCREF(self->winerror); + PyTuple_SET_ITEM(tuple, 0, self->winerror); } else { Py_INCREF(Py_None); @@ -1035,6 +1029,7 @@ SyntaxError_clear(PySyntaxErrorObject *self) static void SyntaxError_dealloc(PySyntaxErrorObject *self) { + _PyObject_GC_UNTRACK(self); SyntaxError_clear(self); self->ob_type->tp_free((PyObject *)self); } @@ -1551,6 +1546,7 @@ UnicodeError_clear(PyUnicodeErrorObject *self) static void UnicodeError_dealloc(PyUnicodeErrorObject *self) { + _PyObject_GC_UNTRACK(self); UnicodeError_clear(self); self->ob_type->tp_free((PyObject *)self); } @@ -1637,7 +1633,7 @@ UnicodeEncodeError_str(PyObject *self) static PyTypeObject _PyExc_UnicodeEncodeError = { PyObject_HEAD_INIT(NULL) 0, - "UnicodeEncodeError", + EXC_MODULE_NAME "UnicodeEncodeError", sizeof(PyUnicodeErrorObject), 0, (destructor)UnicodeError_dealloc, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, (reprfunc)UnicodeEncodeError_str, 0, 0, 0, @@ -1812,7 +1808,7 @@ static PyTypeObject _PyExc_UnicodeTranslateError = { (destructor)UnicodeError_dealloc, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, (reprfunc)UnicodeTranslateError_str, 0, 0, 0, Py_TPFLAGS_DEFAULT | Py_TPFLAGS_BASETYPE | Py_TPFLAGS_HAVE_GC, - PyDoc_STR("Unicode decoding error."), (traverseproc)UnicodeError_traverse, + PyDoc_STR("Unicode translation error."), (traverseproc)UnicodeError_traverse, (inquiry)UnicodeError_clear, 0, 0, 0, 0, 0, UnicodeError_members, 0, &_PyExc_UnicodeError, 0, 0, 0, offsetof(PyUnicodeErrorObject, dict), (initproc)UnicodeTranslateError_init, 0, BaseException_new, diff --git a/Objects/fileobject.c b/Objects/fileobject.c index fb8a542..f2aeb9d 100644 --- a/Objects/fileobject.c +++ b/Objects/fileobject.c @@ -540,7 +540,7 @@ file_seek(PyFileObject *f, PyObject *args) int whence; int ret; Py_off_t offset; - PyObject *offobj; + PyObject *offobj, *off_index; if (f->f_fp == NULL) return err_closed(); @@ -548,12 +548,25 @@ file_seek(PyFileObject *f, PyObject *args) whence = 0; if (!PyArg_ParseTuple(args, "O|i:seek", &offobj, &whence)) return NULL; + off_index = PyNumber_Index(offobj); + if (!off_index) { + if (!PyFloat_Check(offobj)) + return NULL; + /* Deprecated in 2.6 */ + PyErr_Clear(); + if (PyErr_Warn(PyExc_DeprecationWarning, + "integer argument expected, got float")) + return NULL; + off_index = offobj; + Py_INCREF(offobj); + } #if !defined(HAVE_LARGEFILE_SUPPORT) - offset = PyInt_AsLong(offobj); + offset = PyInt_AsLong(off_index); #else - offset = PyLong_Check(offobj) ? - PyLong_AsLongLong(offobj) : PyInt_AsLong(offobj); + offset = PyLong_Check(off_index) ? + PyLong_AsLongLong(off_index) : PyInt_AsLong(off_index); #endif + Py_DECREF(off_index); if (PyErr_Occurred()) return NULL; @@ -927,7 +940,7 @@ file_readinto(PyFileObject *f, PyObject *args) ndone += nnow; ntodo -= nnow; } - return PyInt_FromLong((long)ndone); + return PyInt_FromSsize_t(ndone); } /************************************************************************** @@ -1006,6 +1019,7 @@ getline_via_fgets(FILE *fp) size_t nfree; /* # of free buffer slots; pvend-pvfree */ size_t total_v_size; /* total # of slots in buffer */ size_t increment; /* amount to increment the buffer */ + size_t prev_v_size; /* Optimize for normal case: avoid _PyString_Resize if at all * possible via first reading into stack buffer "buf". @@ -1120,8 +1134,11 @@ getline_via_fgets(FILE *fp) /* expand buffer and try again */ assert(*(pvend-1) == '\0'); increment = total_v_size >> 2; /* mild exponential growth */ + prev_v_size = total_v_size; total_v_size += increment; - if (total_v_size > PY_SSIZE_T_MAX) { + /* check for overflow */ + if (total_v_size <= prev_v_size || + total_v_size > PY_SSIZE_T_MAX) { PyErr_SetString(PyExc_OverflowError, "line is longer than a Python string can hold"); Py_DECREF(v); @@ -1130,7 +1147,7 @@ getline_via_fgets(FILE *fp) if (_PyString_Resize(&v, (int)total_v_size) < 0) return NULL; /* overwrite the trailing null byte */ - pvfree = BUF(v) + (total_v_size - increment - 1); + pvfree = BUF(v) + (prev_v_size - 1); } if (BUF(v) + total_v_size != p) _PyString_Resize(&v, p - BUF(v)); @@ -2014,7 +2031,7 @@ file_init(PyObject *self, PyObject *args, PyObject *kwds) if (!PyArg_ParseTupleAndKeywords(args, kwds, "O|si:file", kwlist, &o_name, &mode, &bufsize)) - return -1; + goto Error; if (fill_file_fields(foself, NULL, o_name, mode, fclose) == NULL) diff --git a/Objects/intobject.c b/Objects/intobject.c index b420c12..0b746de 100644 --- a/Objects/intobject.c +++ b/Objects/intobject.c @@ -557,6 +557,17 @@ int_mul(PyObject *v, PyObject *w) } } +/* Integer overflow checking for unary negation: on a 2's-complement + * box, -x overflows iff x is the most negative long. In this case we + * get -x == x. However, -x is undefined (by C) if x /is/ the most + * negative long (it's a signed overflow case), and some compilers care. + * So we cast x to unsigned long first. However, then other compilers + * warn about applying unary minus to an unsigned operand. Hence the + * weird "0-". + */ +#define UNARY_NEG_WOULD_OVERFLOW(x) \ + ((x) < 0 && (unsigned long)(x) == 0-(unsigned long)(x)) + /* Return type of i_divmod */ enum divmod_result { DIVMOD_OK, /* Correct result */ @@ -576,7 +587,7 @@ i_divmod(register long x, register long y, return DIVMOD_ERROR; } /* (-sys.maxint-1)/-1 is the only overflow case. */ - if (y == -1 && x < 0 && x == -x) + if (y == -1 && UNARY_NEG_WOULD_OVERFLOW(x)) return DIVMOD_OVERFLOW; xdivy = x / y; xmody = x - xdivy * y; @@ -744,10 +755,10 @@ int_pow(PyIntObject *v, PyIntObject *w, PyIntObject *z) static PyObject * int_neg(PyIntObject *v) { - register long a, x; + register long a; a = v->ob_ival; - x = -a; - if (a < 0 && x < 0) { + /* check for overflow */ + if (UNARY_NEG_WOULD_OVERFLOW(a)) { PyObject *o = PyLong_FromLong(a); if (o != NULL) { PyObject *result = PyNumber_Negative(o); @@ -756,7 +767,7 @@ int_neg(PyIntObject *v) } return NULL; } - return PyInt_FromLong(x); + return PyInt_FromLong(-a); } static PyObject * @@ -955,8 +966,25 @@ int_new(PyTypeObject *type, PyObject *args, PyObject *kwds) return PyInt_FromLong(0L); if (base == -909) return PyNumber_Int(x); - if (PyString_Check(x)) - return PyInt_FromString(PyString_AS_STRING(x), NULL, base); + if (PyString_Check(x)) { + /* Since PyInt_FromString doesn't have a length parameter, + * check here for possible NULs in the string. */ + char *string = PyString_AS_STRING(x); + if (strlen(string) != PyString_Size(x)) { + /* create a repr() of the input string, + * just like PyInt_FromString does */ + PyObject *srepr; + srepr = PyObject_Repr(x); + if (srepr == NULL) + return NULL; + PyErr_Format(PyExc_ValueError, + "invalid literal for int() with base %d: %s", + base, PyString_AS_STRING(srepr)); + Py_DECREF(srepr); + return NULL; + } + return PyInt_FromString(string, NULL, base); + } #ifdef Py_USING_UNICODE if (PyUnicode_Check(x)) return PyInt_FromUnicode(PyUnicode_AS_UNICODE(x), diff --git a/Objects/listobject.c b/Objects/listobject.c index 9a11680..4c9c4c3 100644 --- a/Objects/listobject.c +++ b/Objects/listobject.c @@ -863,17 +863,12 @@ static PyObject * listpop(PyListObject *self, PyObject *args) { Py_ssize_t i = -1; - PyObject *v, *arg = NULL; + PyObject *v; int status; - if (!PyArg_UnpackTuple(args, "pop", 0, 1, &arg)) + if (!PyArg_ParseTuple(args, "|n:pop", &i)) return NULL; - if (arg != NULL) { - if (PyInt_Check(arg)) - i = PyInt_AS_LONG((PyIntObject*) arg); - else if (!PyArg_ParseTuple(args, "|n:pop", &i)) - return NULL; - } + if (self->ob_size == 0) { /* Special-case most common failure cause */ PyErr_SetString(PyExc_IndexError, "pop from empty list"); @@ -951,9 +946,10 @@ islt(PyObject *x, PyObject *y, PyObject *compare) if (res == NULL) return -1; if (!PyInt_Check(res)) { + PyErr_Format(PyExc_TypeError, + "comparison function must return int, not %.200s", + res->ob_type->tp_name); Py_DECREF(res); - PyErr_SetString(PyExc_TypeError, - "comparison function must return int"); return -1; } i = PyInt_AsLong(res); @@ -2507,8 +2503,9 @@ list_subscript(PyListObject* self, PyObject* item) } } else { - PyErr_SetString(PyExc_TypeError, - "list indices must be integers"); + PyErr_Format(PyExc_TypeError, + "list indices must be integers, not %.200s", + item->ob_type->tp_name); return NULL; } } @@ -2624,6 +2621,11 @@ list_ass_subscript(PyListObject* self, PyObject* item, PyObject* value) garbage = (PyObject**) PyMem_MALLOC(slicelength*sizeof(PyObject*)); + if (!garbage) { + Py_DECREF(seq); + PyErr_NoMemory(); + return -1; + } selfitems = self->ob_item; seqitems = PySequence_Fast_ITEMS(seq); @@ -2646,8 +2648,9 @@ list_ass_subscript(PyListObject* self, PyObject* item, PyObject* value) } } else { - PyErr_SetString(PyExc_TypeError, - "list indices must be integers"); + PyErr_Format(PyExc_TypeError, + "list indices must be integers, not %.200s", + item->ob_type->tp_name); return -1; } } diff --git a/Objects/longobject.c b/Objects/longobject.c index 566e6a7..f68656b 100644 --- a/Objects/longobject.c +++ b/Objects/longobject.c @@ -191,6 +191,18 @@ PyLong_FromDouble(double dval) return (PyObject *)v; } +/* Checking for overflow in PyLong_AsLong is a PITA since C doesn't define + * anything about what happens when a signed integer operation overflows, + * and some compilers think they're doing you a favor by being "clever" + * then. The bit pattern for the largest postive signed long is + * (unsigned long)LONG_MAX, and for the smallest negative signed long + * it is abs(LONG_MIN), which we could write -(unsigned long)LONG_MIN. + * However, some other compilers warn about applying unary minus to an + * unsigned operand. Hence the weird "0-". + */ +#define PY_ABS_LONG_MIN (0-(unsigned long)LONG_MIN) +#define PY_ABS_SSIZE_T_MIN (0-(size_t)PY_SSIZE_T_MIN) + /* Get a C long int from a long int object. Returns -1 and sets an error condition if overflow occurs. */ @@ -223,14 +235,16 @@ PyLong_AsLong(PyObject *vv) if ((x >> SHIFT) != prev) goto overflow; } - /* Haven't lost any bits, but if the sign bit is set we're in - * trouble *unless* this is the min negative number. So, - * trouble iff sign bit set && (positive || some bit set other - * than the sign bit). - */ - if ((long)x < 0 && (sign > 0 || (x << 1) != 0)) - goto overflow; - return (long)x * sign; + /* Haven't lost any bits, but casting to long requires extra care + * (see comment above). + */ + if (x <= (unsigned long)LONG_MAX) { + return (long)x * sign; + } + else if (sign < 0 && x == PY_ABS_LONG_MIN) { + return LONG_MIN; + } + /* else overflow */ overflow: PyErr_SetString(PyExc_OverflowError, @@ -266,14 +280,16 @@ _PyLong_AsSsize_t(PyObject *vv) { if ((x >> SHIFT) != prev) goto overflow; } - /* Haven't lost any bits, but if the sign bit is set we're in - * trouble *unless* this is the min negative number. So, - * trouble iff sign bit set && (positive || some bit set other - * than the sign bit). + /* Haven't lost any bits, but casting to a signed type requires + * extra care (see comment above). */ - if ((Py_ssize_t)x < 0 && (sign > 0 || (x << 1) != 0)) - goto overflow; - return (Py_ssize_t)x * sign; + if (x <= (size_t)PY_SSIZE_T_MAX) { + return (Py_ssize_t)x * sign; + } + else if (sign < 0 && x == PY_ABS_SSIZE_T_MIN) { + return PY_SSIZE_T_MIN; + } + /* else overflow */ overflow: PyErr_SetString(PyExc_OverflowError, @@ -1165,7 +1181,7 @@ long_format(PyObject *aa, int base) { register PyLongObject *a = (PyLongObject *)aa; PyStringObject *str; - Py_ssize_t i; + Py_ssize_t i, j, sz; Py_ssize_t size_a; char *p; int bits; @@ -1185,11 +1201,18 @@ long_format(PyObject *aa, int base) ++bits; i >>= 1; } - i = 5 + (size_a*SHIFT + bits-1) / bits; - str = (PyStringObject *) PyString_FromStringAndSize((char *)0, i); + i = 5; + j = size_a*SHIFT + bits-1; + sz = i + j / bits; + if (j / SHIFT < size_a || sz < i) { + PyErr_SetString(PyExc_OverflowError, + "long is too large to format"); + return NULL; + } + str = (PyStringObject *) PyString_FromStringAndSize((char *)0, sz); if (str == NULL) return NULL; - p = PyString_AS_STRING(str) + i; + p = PyString_AS_STRING(str) + sz; *p = '\0'; if (a->ob_size < 0) sign = '-'; @@ -1301,7 +1324,7 @@ long_format(PyObject *aa, int base) } while ((*q++ = *p++) != '\0'); q--; _PyString_Resize((PyObject **)&str, - (int) (q - PyString_AS_STRING(str))); + (Py_ssize_t) (q - PyString_AS_STRING(str))); } return (PyObject *)str; } @@ -1359,14 +1382,14 @@ long_from_binary_base(char **str, int base) while (_PyLong_DigitValue[Py_CHARMASK(*p)] < base) ++p; *str = p; - n = (p - start) * bits_per_char; - if (n / bits_per_char != p - start) { + /* n <- # of Python digits needed, = ceiling(n/SHIFT). */ + n = (p - start) * bits_per_char + SHIFT - 1; + if (n / bits_per_char < p - start) { PyErr_SetString(PyExc_ValueError, "long string too large to convert"); return NULL; } - /* n <- # of Python digits needed, = ceiling(n/SHIFT). */ - n = (n + SHIFT - 1) / SHIFT; + n = n / SHIFT; z = _PyLong_New(n); if (z == NULL) return NULL; diff --git a/Objects/obmalloc.c b/Objects/obmalloc.c index 0ca8f2b..840570e 100644 --- a/Objects/obmalloc.c +++ b/Objects/obmalloc.c @@ -691,7 +691,8 @@ extremely desirable that it be this fast. #undef Py_ADDRESS_IN_RANGE -#if defined(__GNUC__) && (__GNUC__ == 3) && (__GNUC_MINOR__ >= 1) +#if defined(__GNUC__) && ((__GNUC__ == 3) && (__GNUC_MINOR__ >= 1) || \ + (__GNUC__ >= 4)) #define Py_NO_INLINE __attribute__((__noinline__)) #else #define Py_NO_INLINE diff --git a/Objects/setobject.c b/Objects/setobject.c index d489711..0af1f15 100644 --- a/Objects/setobject.c +++ b/Objects/setobject.c @@ -10,6 +10,20 @@ #include "Python.h" #include "structmember.h" +/* Set a key error with the specified argument, wrapping it in a + * tuple automatically so that tuple keys are not unpacked as the + * exception arguments. */ +static void +set_key_error(PyObject *arg) +{ + PyObject *tup; + tup = PyTuple_Pack(1, arg); + if (!tup) + return; /* caller will expect error to be set anyway */ + PyErr_SetObject(PyExc_KeyError, tup); + Py_DECREF(tup); +} + /* This must be >= 1. */ #define PERTURB_SHIFT 5 @@ -179,11 +193,13 @@ set_lookkey_string(PySetObject *so, PyObject *key, register long hash) if (entry->key == dummy && freeslot == NULL) freeslot = entry; } + assert(0); /* NOT REACHED */ + return 0; } /* Internal routine to insert a new key into the table. -Used both by the internal resize routine and by the public insert routine. +Used by the public insert routine. Eats a reference to key. */ static int @@ -216,6 +232,35 @@ set_insert_key(register PySetObject *so, PyObject *key, long hash) } /* +Internal routine used by set_table_resize() to insert an item which is +known to be absent from the set. This routine also assumes that +the set contains no deleted entries. Besides the performance benefit, +using set_insert_clean() in set_table_resize() is dangerous (SF bug #1456209). +Note that no refcounts are changed by this routine; if needed, the caller +is responsible for incref'ing `key`. +*/ +static void +set_insert_clean(register PySetObject *so, PyObject *key, long hash) +{ + register size_t i; + register size_t perturb; + register size_t mask = (size_t)so->mask; + setentry *table = so->table; + register setentry *entry; + + i = hash & mask; + entry = &table[i]; + for (perturb = hash; entry->key != NULL; perturb >>= PERTURB_SHIFT) { + i = (i << 2) + i + perturb + 1; + entry = &table[i & mask]; + } + so->fill++; + entry->key = key; + entry->hash = hash; + so->used++; +} + +/* Restructure the table by allocating a new table and reinserting all keys again. When entries have been deleted, the new table may actually be smaller than the old one. @@ -296,11 +341,7 @@ set_table_resize(PySetObject *so, Py_ssize_t minused) } else { /* ACTIVE */ --i; - if(set_insert_key(so, entry->key, entry->hash) == -1) { - if (is_oldtable_malloced) - PyMem_DEL(oldtable); - return -1; - } + set_insert_clean(so, entry->key, entry->hash); } } @@ -319,8 +360,10 @@ set_add_entry(register PySetObject *so, setentry *entry) assert(so->fill <= so->mask); /* at least one empty slot */ n_used = so->used; Py_INCREF(entry->key); - if (set_insert_key(so, entry->key, entry->hash) == -1) + if (set_insert_key(so, entry->key, entry->hash) == -1) { + Py_DECREF(entry->key); return -1; + } if (!(so->used > n_used && so->fill*3 >= (so->mask+1)*2)) return 0; return set_table_resize(so, so->used>50000 ? so->used*2 : so->used*4); @@ -1155,7 +1198,12 @@ set_intersection(PySetObject *so, PyObject *other) } while (set_next((PySetObject *)other, &pos, &entry)) { - if (set_contains_entry(so, entry)) { + int rv = set_contains_entry(so, entry); + if (rv == -1) { + Py_DECREF(result); + return NULL; + } + if (rv) { if (set_add_entry(result, entry) == -1) { Py_DECREF(result); return NULL; @@ -1172,8 +1220,27 @@ set_intersection(PySetObject *so, PyObject *other) } while ((key = PyIter_Next(it)) != NULL) { - if (set_contains_key(so, key)) { - if (set_add_key(result, key) == -1) { + int rv; + setentry entry; + long hash = PyObject_Hash(key); + + if (hash == -1) { + Py_DECREF(it); + Py_DECREF(result); + Py_DECREF(key); + return NULL; + } + entry.hash = hash; + entry.key = key; + rv = set_contains_entry(so, &entry); + if (rv == -1) { + Py_DECREF(it); + Py_DECREF(result); + Py_DECREF(key); + return NULL; + } + if (rv) { + if (set_add_entry(result, &entry) == -1) { Py_DECREF(it); Py_DECREF(result); Py_DECREF(key); @@ -1249,7 +1316,8 @@ set_difference_update_internal(PySetObject *so, PyObject *other) Py_ssize_t pos = 0; while (set_next((PySetObject *)other, &pos, &entry)) - set_discard_entry(so, entry); + if (set_discard_entry(so, entry) == -1) + return -1; } else { PyObject *key, *it; it = PyObject_GetIter(other); @@ -1312,17 +1380,26 @@ set_difference(PySetObject *so, PyObject *other) entrycopy.hash = entry->hash; entrycopy.key = entry->key; if (!PyDict_Contains(other, entry->key)) { - if (set_add_entry((PySetObject *)result, &entrycopy) == -1) + if (set_add_entry((PySetObject *)result, &entrycopy) == -1) { + Py_DECREF(result); return NULL; + } } } return result; } while (set_next(so, &pos, &entry)) { - if (!set_contains_entry((PySetObject *)other, entry)) { - if (set_add_entry((PySetObject *)result, entry) == -1) + int rv = set_contains_entry((PySetObject *)other, entry); + if (rv == -1) { + Py_DECREF(result); + return NULL; + } + if (!rv) { + if (set_add_entry((PySetObject *)result, entry) == -1) { + Py_DECREF(result); return NULL; + } } } return result; @@ -1374,11 +1451,18 @@ set_symmetric_difference_update(PySetObject *so, PyObject *other) PyObject *value; int rv; while (PyDict_Next(other, &pos, &key, &value)) { - rv = set_discard_key(so, key); + setentry an_entry; + long hash = PyObject_Hash(key); + + if (hash == -1) + return NULL; + an_entry.hash = hash; + an_entry.key = key; + rv = set_discard_entry(so, &an_entry); if (rv == -1) return NULL; if (rv == DISCARD_NOTFOUND) { - if (set_add_key(so, key) == -1) + if (set_add_entry(so, &an_entry) == -1) return NULL; } } @@ -1481,7 +1565,10 @@ set_issubset(PySetObject *so, PyObject *other) Py_RETURN_FALSE; while (set_next(so, &pos, &entry)) { - if (!set_contains_entry((PySetObject *)other, entry)) + int rv = set_contains_entry((PySetObject *)other, entry); + if (rv == -1) + return NULL; + if (!rv) Py_RETURN_FALSE; } Py_RETURN_TRUE; @@ -1628,7 +1715,7 @@ set_remove(PySetObject *so, PyObject *key) Py_DECREF(tmpkey); return result; } else if (rv == DISCARD_NOTFOUND) { - PyErr_SetObject(PyExc_KeyError, key); + set_key_error(key); return NULL; } Py_RETURN_NONE; diff --git a/Objects/stringlib/partition.h b/Objects/stringlib/partition.h index 1486347..105ba31 100644 --- a/Objects/stringlib/partition.h +++ b/Objects/stringlib/partition.h @@ -78,12 +78,12 @@ stringlib_rpartition( } if (pos < 0) { - Py_INCREF(str_obj); - PyTuple_SET_ITEM(out, 0, (PyObject*) str_obj); Py_INCREF(STRINGLIB_EMPTY); - PyTuple_SET_ITEM(out, 1, (PyObject*) STRINGLIB_EMPTY); + PyTuple_SET_ITEM(out, 0, (PyObject*) STRINGLIB_EMPTY); Py_INCREF(STRINGLIB_EMPTY); - PyTuple_SET_ITEM(out, 2, (PyObject*) STRINGLIB_EMPTY); + PyTuple_SET_ITEM(out, 1, (PyObject*) STRINGLIB_EMPTY); + Py_INCREF(str_obj); + PyTuple_SET_ITEM(out, 2, (PyObject*) str_obj); return out; } diff --git a/Objects/stringobject.c b/Objects/stringobject.c index 028b375..61e6af5 100644 --- a/Objects/stringobject.c +++ b/Objects/stringobject.c @@ -804,10 +804,22 @@ string_print(PyStringObject *op, FILE *fp, int flags) return ret; } if (flags & Py_PRINT_RAW) { + char *data = op->ob_sval; + Py_ssize_t size = op->ob_size; + while (size > INT_MAX) { + /* Very long strings cannot be written atomically. + * But don't write exactly INT_MAX bytes at a time + * to avoid memory aligment issues. + */ + const int chunk_size = INT_MAX & ~0x3FFF; + fwrite(data, 1, chunk_size, fp); + data += chunk_size; + size -= chunk_size; + } #ifdef __VMS - if (op->ob_size) fwrite(op->ob_sval, (int) op->ob_size, 1, fp); + if (size) fwrite(data, (int)size, 1, fp); #else - fwrite(op->ob_sval, 1, (int) op->ob_size, fp); + fwrite(data, 1, (int)size, fp); #endif return 0; } @@ -844,7 +856,7 @@ PyString_Repr(PyObject *obj, int smartquotes) register PyStringObject* op = (PyStringObject*) obj; size_t newsize = 2 + 4 * op->ob_size; PyObject *v; - if (newsize > PY_SSIZE_T_MAX) { + if (newsize > PY_SSIZE_T_MAX || newsize / 4 != op->ob_size) { PyErr_SetString(PyExc_OverflowError, "string is too large to make repr"); } @@ -1059,8 +1071,9 @@ string_contains(PyObject *str_obj, PyObject *sub_obj) return PyUnicode_Contains(str_obj, sub_obj); #endif if (!PyString_Check(sub_obj)) { - PyErr_SetString(PyExc_TypeError, - "'in <string>' requires string as left operand"); + PyErr_Format(PyExc_TypeError, + "'in <string>' requires string as left operand, " + "not %.200s", sub_obj->ob_type->tp_name); return -1; } } @@ -1228,8 +1241,9 @@ string_subscript(PyStringObject* self, PyObject* item) } } else { - PyErr_SetString(PyExc_TypeError, - "string indices must be integers"); + PyErr_Format(PyExc_TypeError, + "string indices must be integers, not %.200s", + item->ob_type->tp_name); return NULL; } } @@ -1543,11 +1557,11 @@ string_partition(PyStringObject *self, PyObject *sep_obj) } PyDoc_STRVAR(rpartition__doc__, -"S.rpartition(sep) -> (head, sep, tail)\n\ +"S.rpartition(sep) -> (tail, sep, head)\n\ \n\ Searches for the separator sep in S, starting at the end of S, and returns\n\ the part before it, the separator itself, and the part after it. If the\n\ -separator is not found, returns S and two empty strings."); +separator is not found, returns two empty strings and S."); static PyObject * string_rpartition(PyStringObject *self, PyObject *sep_obj) @@ -4136,7 +4150,8 @@ formatfloat(char *buf, size_t buflen, int flags, double x; x = PyFloat_AsDouble(v); if (x == -1.0 && PyErr_Occurred()) { - PyErr_SetString(PyExc_TypeError, "float argument required"); + PyErr_Format(PyExc_TypeError, "float argument required, " + "not %.200s", v->ob_type->tp_name); return -1; } if (prec < 0) @@ -4237,7 +4252,7 @@ _PyString_FormatLong(PyObject *val, int flags, int prec, int type, return NULL; } llen = PyString_Size(result); - if (llen > PY_SSIZE_T_MAX) { + if (llen > INT_MAX) { PyErr_SetString(PyExc_ValueError, "string too large in _PyString_FormatLong"); return NULL; } @@ -4331,7 +4346,8 @@ formatint(char *buf, size_t buflen, int flags, x = PyInt_AsLong(v); if (x == -1 && PyErr_Occurred()) { - PyErr_SetString(PyExc_TypeError, "int argument required"); + PyErr_Format(PyExc_TypeError, "int argument required, not %.200s", + v->ob_type->tp_name); return -1; } if (x < 0 && type == 'u') { @@ -4726,9 +4742,10 @@ PyString_Format(PyObject *format, PyObject *args) default: PyErr_Format(PyExc_ValueError, "unsupported format character '%c' (0x%x) " - "at index %i", + "at index %zd", c, c, - (int)(fmt - 1 - PyString_AsString(format))); + (Py_ssize_t)(fmt - 1 - + PyString_AsString(format))); goto error; } if (sign) { diff --git a/Objects/tupleobject.c b/Objects/tupleobject.c index 6f3711f..c85b35a 100644 --- a/Objects/tupleobject.c +++ b/Objects/tupleobject.c @@ -620,8 +620,9 @@ tuplesubscript(PyTupleObject* self, PyObject* item) } } else { - PyErr_SetString(PyExc_TypeError, - "tuple indices must be integers"); + PyErr_Format(PyExc_TypeError, + "tuple indices must be integers, not %.200s", + item->ob_type->tp_name); return NULL; } } diff --git a/Objects/typeobject.c b/Objects/typeobject.c index ea1d1a2..1bef208 100644 --- a/Objects/typeobject.c +++ b/Objects/typeobject.c @@ -98,7 +98,7 @@ type_module(PyTypeObject *type, void *context) s = strrchr(type->tp_name, '.'); if (s != NULL) return PyString_FromStringAndSize( - type->tp_name, (int)(s - type->tp_name)); + type->tp_name, (Py_ssize_t)(s - type->tp_name)); return PyString_FromString("__builtin__"); } } @@ -3644,7 +3644,7 @@ hackcheck(PyObject *self, setattrofunc func, char *what) while (type && type->tp_flags & Py_TPFLAGS_HEAPTYPE) type = type->tp_base; /* If type is NULL now, this is a really weird type. - In the same of backwards compatibility (?), just shut up. */ + In the spirit of backwards compatibility (?), just shut up. */ if (type && type->tp_setattro != func) { PyErr_Format(PyExc_TypeError, "can't apply this %s to %s object", @@ -3861,7 +3861,7 @@ tp_new_wrapper(PyObject *self, PyObject *args, PyObject *kwds) while (staticbase && (staticbase->tp_flags & Py_TPFLAGS_HEAPTYPE)) staticbase = staticbase->tp_base; /* If staticbase is NULL now, it is a really weird type. - In the same of backwards compatibility (?), just shut up. */ + In the spirit of backwards compatibility (?), just shut up. */ if (staticbase && staticbase->tp_new != type->tp_new) { PyErr_Format(PyExc_TypeError, "%s.__new__(%s) is not safe, use %s.__new__()", @@ -4017,19 +4017,10 @@ slot_sq_length(PyObject *self) return -1; len = PyInt_AsSsize_t(res); Py_DECREF(res); - if (len == -1 && PyErr_Occurred()) - return -1; -#if SIZEOF_SIZE_T < SIZEOF_INT - /* Overflow check -- range of PyInt is more than C ssize_t */ - if (len != (int)len) { - PyErr_SetString(PyExc_OverflowError, - "__len__() should return 0 <= outcome < 2**31"); - return -1; - } -#endif if (len < 0) { - PyErr_SetString(PyExc_ValueError, - "__len__() should return >= 0"); + if (!PyErr_Occurred()) + PyErr_SetString(PyExc_ValueError, + "__len__() should return >= 0"); return -1; } return len; @@ -5583,6 +5574,8 @@ super_init(PyObject *self, PyObject *args, PyObject *kwds) PyObject *obj = NULL; PyTypeObject *obj_type = NULL; + if (!_PyArg_NoKeywords("super", kwds)) + return -1; if (!PyArg_ParseTuple(args, "O!|O:super", &PyType_Type, &type, &obj)) return -1; if (obj == Py_None) diff --git a/Objects/unicodeobject.c b/Objects/unicodeobject.c index 92301c0..bde3454 100644 --- a/Objects/unicodeobject.c +++ b/Objects/unicodeobject.c @@ -2040,7 +2040,32 @@ PyObject *unicodeescape_string(const Py_UNICODE *s, static const char *hexdigit = "0123456789abcdef"; - repr = PyString_FromStringAndSize(NULL, 2 + 6*size + 1); + /* XXX(nnorwitz): rather than over-allocating, it would be + better to choose a different scheme. Perhaps scan the + first N-chars of the string and allocate based on that size. + */ + /* Initial allocation is based on the longest-possible unichr + escape. + + In wide (UTF-32) builds '\U00xxxxxx' is 10 chars per source + unichr, so in this case it's the longest unichr escape. In + narrow (UTF-16) builds this is five chars per source unichr + since there are two unichrs in the surrogate pair, so in narrow + (UTF-16) builds it's not the longest unichr escape. + + In wide or narrow builds '\uxxxx' is 6 chars per source unichr, + so in the narrow (UTF-16) build case it's the longest unichr + escape. + */ + + repr = PyString_FromStringAndSize(NULL, + 2 +#ifdef Py_UNICODE_WIDE + + 10*size +#else + + 6*size +#endif + + 1); if (repr == NULL) return NULL; @@ -2065,15 +2090,6 @@ PyObject *unicodeescape_string(const Py_UNICODE *s, #ifdef Py_UNICODE_WIDE /* Map 21-bit characters to '\U00xxxxxx' */ else if (ch >= 0x10000) { - Py_ssize_t offset = p - PyString_AS_STRING(repr); - - /* Resize the string if necessary */ - if (offset + 12 > PyString_GET_SIZE(repr)) { - if (_PyString_Resize(&repr, PyString_GET_SIZE(repr) + 100)) - return NULL; - p = PyString_AS_STRING(repr) + offset; - } - *p++ = '\\'; *p++ = 'U'; *p++ = hexdigit[(ch >> 28) & 0x0000000F]; @@ -2086,8 +2102,8 @@ PyObject *unicodeescape_string(const Py_UNICODE *s, *p++ = hexdigit[ch & 0x0000000F]; continue; } -#endif - /* Map UTF-16 surrogate pairs to Unicode \UXXXXXXXX escapes */ +#else + /* Map UTF-16 surrogate pairs to '\U00xxxxxx' */ else if (ch >= 0xD800 && ch < 0xDC00) { Py_UNICODE ch2; Py_UCS4 ucs; @@ -2112,6 +2128,7 @@ PyObject *unicodeescape_string(const Py_UNICODE *s, s--; size++; } +#endif /* Map 16-bit characters to '\uxxxx' */ if (ch >= 256) { @@ -2367,6 +2384,7 @@ PyObject *_PyUnicode_DecodeUnicodeInternal(const char *s, Py_UNICODE unimax = PyUnicode_GetMax(); #endif + /* XXX overflow detection missing */ v = _PyUnicode_New((size+Py_UNICODE_SIZE-1)/ Py_UNICODE_SIZE); if (v == NULL) goto onError; @@ -3153,6 +3171,7 @@ PyObject *PyUnicode_DecodeCharmap(const char *s, Py_ssize_t needed = (targetsize - extrachars) + \ (targetsize << 2); extrachars += needed; + /* XXX overflow detection missing */ if (_PyUnicode_Resize(&v, PyUnicode_GET_SIZE(v) + needed) < 0) { Py_DECREF(x); @@ -6695,11 +6714,11 @@ unicode_partition(PyUnicodeObject *self, PyObject *separator) } PyDoc_STRVAR(rpartition__doc__, -"S.rpartition(sep) -> (head, sep, tail)\n\ +"S.rpartition(sep) -> (tail, sep, head)\n\ \n\ Searches for the separator sep in S, starting at the end of S, and returns\n\ the part before it, the separator itself, and the part after it. If the\n\ -separator is not found, returns S and two empty strings."); +separator is not found, returns two empty strings and S."); static PyObject* unicode_rpartition(PyUnicodeObject *self, PyObject *separator) @@ -7744,10 +7763,11 @@ PyObject *PyUnicode_Format(PyObject *format, default: PyErr_Format(PyExc_ValueError, "unsupported format character '%c' (0x%x) " - "at index %i", + "at index %zd", (31<=c && c<=126) ? (char)c : '?', (int)c, - (int)(fmt -1 - PyUnicode_AS_UNICODE(uformat))); + (Py_ssize_t)(fmt - 1 - + PyUnicode_AS_UNICODE(uformat))); goto onError; } if (sign) { |