Skip to content

C
cpython
Commit 1f7df359 authored by Tim Peters's avatar Tim Peters
Browse files
Options

Remove the CACHE_HASH and INTERN_STRINGS preprocessor symbols.

parent 83584056
Hide whitespace changes
Inline Side-by-side
Showing with 19 additions and 102 deletions
Include/stringobject.h
View file @ 1f7df359
...@@ -24,28 +24,18 @@ variant that assumes a zero-terminated string. Note that none of the ...@@ -24,28 +24,18 @@ variant that assumes a zero-terminated string. Note that none of the
functions should be applied to nil objects. functions should be applied to nil objects.
*/ */
/* Two speedup hacks. Caching the hash saves recalculation of a /* Caching the hash (ob_shash) saves recalculation of a string's hash value.
string's hash value. Interning strings (which requires hash Interning strings (ob_sinterned) tries to ensure that only one string
caching) tries to ensure that only one string object with a given object with a given value exists, so equality tests can be one pointer
value exists, so equality tests are one pointer comparison. comparison. This is generally restricted to strings that "look like"
Together, these can speed the interpreter up by as much as 20%. Python identifiers, although the intern() builtin can be used to force
Each costs the size of a long or pointer per string object. In interning of any string.
addition, interned strings live until the end of times. If you are Together, these sped the interpreter by up to 20%. */
concerned about memory footprint, simply comment the #define out
here (and rebuild everything!). */
#define CACHE_HASH
#ifdef CACHE_HASH
#define INTERN_STRINGS
#endif
typedef struct { typedef struct {
PyObject_VAR_HEAD PyObject_VAR_HEAD
#ifdef CACHE_HASH
long ob_shash; long ob_shash;
#endif
#ifdef INTERN_STRINGS
PyObject *ob_sinterned; PyObject *ob_sinterned;
#endif
char ob_sval[1]; char ob_sval[1];
} PyStringObject; } PyStringObject;
...@@ -70,15 +60,9 @@ extern DL_IMPORT(PyObject *) PyString_Format(PyObject *, PyObject *); ...@@ -70,15 +60,9 @@ extern DL_IMPORT(PyObject *) PyString_Format(PyObject *, PyObject *);
extern DL_IMPORT(PyObject *) _PyString_FormatLong(PyObject*, int, int, extern DL_IMPORT(PyObject *) _PyString_FormatLong(PyObject*, int, int,
int, char**, int*); int, char**, int*);
#ifdef INTERN_STRINGS
extern DL_IMPORT(void) PyString_InternInPlace(PyObject **); extern DL_IMPORT(void) PyString_InternInPlace(PyObject **);
extern DL_IMPORT(PyObject *) PyString_InternFromString(const char *); extern DL_IMPORT(PyObject *) PyString_InternFromString(const char *);
extern DL_IMPORT(void) _Py_ReleaseInternedStrings(void); extern DL_IMPORT(void) _Py_ReleaseInternedStrings(void);
#else
#define PyString_InternInPlace(p)
#define PyString_InternFromString(cp) PyString_FromString(cp)
#define _Py_ReleaseInternedStrings()
#endif
/* Macro, trading safety for speed */ /* Macro, trading safety for speed */
#define PyString_AS_STRING(op) (((PyStringObject *)(op))->ob_sval) #define PyString_AS_STRING(op) (((PyStringObject *)(op))->ob_sval)
......
Misc/NEWS
View file @ 1f7df359
...@@ -89,6 +89,10 @@ Tools/Demos ...@@ -89,6 +89,10 @@ Tools/Demos
Build Build
- References to the CACHE_HASH and INTERN_STRINGS preprocessor symbols
were eliminated. They were always defined, and the internal features
they enabled stopped being experimental long ago.
C API C API
- Objects allocated using the new PyMalloc_New and PyMalloc_NewVar - Objects allocated using the new PyMalloc_New and PyMalloc_NewVar
......
Objects/bufferobject.c
View file @ 1f7df359
...@@ -10,9 +10,7 @@ typedef struct { ...@@ -10,9 +10,7 @@ typedef struct {
void *b_ptr; void *b_ptr;
int b_size; int b_size;
int b_readonly; int b_readonly;
#ifdef CACHE_HASH
long b_hash; long b_hash;
#endif
} PyBufferObject; } PyBufferObject;
...@@ -36,9 +34,7 @@ _PyBuffer_FromMemory(PyObject *base, void *ptr, int size, int readonly) ...@@ -36,9 +34,7 @@ _PyBuffer_FromMemory(PyObject *base, void *ptr, int size, int readonly)
b->b_ptr = ptr; b->b_ptr = ptr;
b->b_size = size; b->b_size = size;
b->b_readonly = readonly; b->b_readonly = readonly;
#ifdef CACHE_HASH
b->b_hash = -1; b->b_hash = -1;
#endif
return (PyObject *) b; return (PyObject *) b;
} }
...@@ -152,9 +148,7 @@ PyBuffer_New(int size) ...@@ -152,9 +148,7 @@ PyBuffer_New(int size)
b->b_ptr = (void *)(b + 1); b->b_ptr = (void *)(b + 1);
b->b_size = size; b->b_size = size;
b->b_readonly = 0; b->b_readonly = 0;
#ifdef CACHE_HASH
b->b_hash = -1; b->b_hash = -1;
#endif
return o; return o;
} }
...@@ -211,10 +205,8 @@ buffer_hash(PyBufferObject *self) ...@@ -211,10 +205,8 @@ buffer_hash(PyBufferObject *self)
register unsigned char *p; register unsigned char *p;
register long x; register long x;
#ifdef CACHE_HASH
if ( self->b_hash != -1 ) if ( self->b_hash != -1 )
return self->b_hash; return self->b_hash;
#endif
if ( !self->b_readonly ) if ( !self->b_readonly )
{ {
...@@ -231,9 +223,7 @@ buffer_hash(PyBufferObject *self) ...@@ -231,9 +223,7 @@ buffer_hash(PyBufferObject *self)
x ^= self->b_size; x ^= self->b_size;
if (x == -1) if (x == -1)
x = -2; x = -2;
#ifdef CACHE_HASH
self->b_hash = x; self->b_hash = x;
#endif
return x; return x;
} }
......
Objects/dictobject.c
View file @ 1f7df359
...@@ -480,10 +480,8 @@ PyDict_GetItem(PyObject *op, PyObject *key) ...@@ -480,10 +480,8 @@ PyDict_GetItem(PyObject *op, PyObject *key)
if (!PyDict_Check(op)) { if (!PyDict_Check(op)) {
return NULL; return NULL;
} }
#ifdef CACHE_HASH
if (!PyString_CheckExact(key) || if (!PyString_CheckExact(key) ||
(hash = ((PyStringObject *) key)->ob_shash) == -1) (hash = ((PyStringObject *) key)->ob_shash) == -1)
#endif
{ {
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) { if (hash == -1) {
...@@ -512,24 +510,18 @@ PyDict_SetItem(register PyObject *op, PyObject *key, PyObject *value) ...@@ -512,24 +510,18 @@ PyDict_SetItem(register PyObject *op, PyObject *key, PyObject *value)
return -1; return -1;
} }
mp = (dictobject *)op; mp = (dictobject *)op;
#ifdef CACHE_HASH
if (PyString_CheckExact(key)) { if (PyString_CheckExact(key)) {
#ifdef INTERN_STRINGS
if (((PyStringObject *)key)->ob_sinterned != NULL) { if (((PyStringObject *)key)->ob_sinterned != NULL) {
key = ((PyStringObject *)key)->ob_sinterned; key = ((PyStringObject *)key)->ob_sinterned;
hash = ((PyStringObject *)key)->ob_shash; hash = ((PyStringObject *)key)->ob_shash;
} }
else else {
#endif
{
hash = ((PyStringObject *)key)->ob_shash; hash = ((PyStringObject *)key)->ob_shash;
if (hash == -1) if (hash == -1)
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
} }
} }
else else {
#endif
{
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) if (hash == -1)
return -1; return -1;
...@@ -564,11 +556,8 @@ PyDict_DelItem(PyObject *op, PyObject *key) ...@@ -564,11 +556,8 @@ PyDict_DelItem(PyObject *op, PyObject *key)
PyErr_BadInternalCall(); PyErr_BadInternalCall();
return -1; return -1;
} }
#ifdef CACHE_HASH
if (!PyString_CheckExact(key) || if (!PyString_CheckExact(key) ||
(hash = ((PyStringObject *) key)->ob_shash) == -1) (hash = ((PyStringObject *) key)->ob_shash) == -1) {
#endif
{
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) if (hash == -1)
return -1; return -1;
...@@ -844,11 +833,8 @@ dict_subscript(dictobject *mp, register PyObject *key) ...@@ -844,11 +833,8 @@ dict_subscript(dictobject *mp, register PyObject *key)
PyObject *v; PyObject *v;
long hash; long hash;
assert(mp->ma_table != NULL); assert(mp->ma_table != NULL);
#ifdef CACHE_HASH
if (!PyString_CheckExact(key) || if (!PyString_CheckExact(key) ||
(hash = ((PyStringObject *) key)->ob_shash) == -1) (hash = ((PyStringObject *) key)->ob_shash) == -1) {
#endif
{
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) if (hash == -1)
return NULL; return NULL;
...@@ -1436,11 +1422,8 @@ dict_has_key(register dictobject *mp, PyObject *key) ...@@ -1436,11 +1422,8 @@ dict_has_key(register dictobject *mp, PyObject *key)
{ {
long hash; long hash;
register long ok; register long ok;
#ifdef CACHE_HASH
if (!PyString_CheckExact(key) || if (!PyString_CheckExact(key) ||
(hash = ((PyStringObject *) key)->ob_shash) == -1) (hash = ((PyStringObject *) key)->ob_shash) == -1) {
#endif
{
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) if (hash == -1)
return NULL; return NULL;
...@@ -1460,11 +1443,8 @@ dict_get(register dictobject *mp, PyObject *args) ...@@ -1460,11 +1443,8 @@ dict_get(register dictobject *mp, PyObject *args)
if (!PyArg_ParseTuple(args, "O|O:get", &key, &failobj)) if (!PyArg_ParseTuple(args, "O|O:get", &key, &failobj))
return NULL; return NULL;
#ifdef CACHE_HASH
if (!PyString_CheckExact(key) || if (!PyString_CheckExact(key) ||
(hash = ((PyStringObject *) key)->ob_shash) == -1) (hash = ((PyStringObject *) key)->ob_shash) == -1) {
#endif
{
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) if (hash == -1)
return NULL; return NULL;
...@@ -1489,11 +1469,8 @@ dict_setdefault(register dictobject *mp, PyObject *args) ...@@ -1489,11 +1469,8 @@ dict_setdefault(register dictobject *mp, PyObject *args)
if (!PyArg_ParseTuple(args, "O|O:setdefault", &key, &failobj)) if (!PyArg_ParseTuple(args, "O|O:setdefault", &key, &failobj))
return NULL; return NULL;
#ifdef CACHE_HASH
if (!PyString_CheckExact(key) || if (!PyString_CheckExact(key) ||
(hash = ((PyStringObject *) key)->ob_shash) == -1) (hash = ((PyStringObject *) key)->ob_shash) == -1) {
#endif
{
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) if (hash == -1)
return NULL; return NULL;
...@@ -1725,11 +1702,8 @@ dict_contains(dictobject *mp, PyObject *key) ...@@ -1725,11 +1702,8 @@ dict_contains(dictobject *mp, PyObject *key)
{ {
long hash; long hash;
#ifdef CACHE_HASH
if (!PyString_CheckExact(key) || if (!PyString_CheckExact(key) ||
(hash = ((PyStringObject *) key)->ob_shash) == -1) (hash = ((PyStringObject *) key)->ob_shash) == -1) {
#endif
{
hash = PyObject_Hash(key); hash = PyObject_Hash(key);
if (hash == -1) if (hash == -1)
return -1; return -1;
......
Objects/stringobject.c
View file @ 1f7df359
...@@ -72,12 +72,8 @@ PyString_FromStringAndSize(const char *str, int size) ...@@ -72,12 +72,8 @@ PyString_FromStringAndSize(const char *str, int size)
if (op == NULL) if (op == NULL)
return PyErr_NoMemory(); return PyErr_NoMemory();
PyObject_INIT_VAR(op, &PyString_Type, size); PyObject_INIT_VAR(op, &PyString_Type, size);
#ifdef CACHE_HASH
op->ob_shash = -1; op->ob_shash = -1;
#endif
#ifdef INTERN_STRINGS
op->ob_sinterned = NULL; op->ob_sinterned = NULL;
#endif
if (str != NULL) if (str != NULL)
memcpy(op->ob_sval, str, size); memcpy(op->ob_sval, str, size);
op->ob_sval[size] = '\0'; op->ob_sval[size] = '\0';
...@@ -135,12 +131,8 @@ PyString_FromString(const char *str) ...@@ -135,12 +131,8 @@ PyString_FromString(const char *str)
if (op == NULL) if (op == NULL)
return PyErr_NoMemory(); return PyErr_NoMemory();
PyObject_INIT_VAR(op, &PyString_Type, size); PyObject_INIT_VAR(op, &PyString_Type, size);
#ifdef CACHE_HASH
op->ob_shash = -1; op->ob_shash = -1;
#endif
#ifdef INTERN_STRINGS
op->ob_sinterned = NULL; op->ob_sinterned = NULL;
#endif
memcpy(op->ob_sval, str, size+1); memcpy(op->ob_sval, str, size+1);
#ifndef DONT_SHARE_SHORT_STRINGS #ifndef DONT_SHARE_SHORT_STRINGS
if (size == 0) { if (size == 0) {
...@@ -737,12 +729,8 @@ string_concat(register PyStringObject *a, register PyObject *bb) ...@@ -737,12 +729,8 @@ string_concat(register PyStringObject *a, register PyObject *bb)
if (op == NULL) if (op == NULL)
return PyErr_NoMemory(); return PyErr_NoMemory();
PyObject_INIT_VAR(op, &PyString_Type, size); PyObject_INIT_VAR(op, &PyString_Type, size);
#ifdef CACHE_HASH
op->ob_shash = -1; op->ob_shash = -1;
#endif
#ifdef INTERN_STRINGS
op->ob_sinterned = NULL; op->ob_sinterned = NULL;
#endif
memcpy(op->ob_sval, a->ob_sval, (int) a->ob_size); memcpy(op->ob_sval, a->ob_sval, (int) a->ob_size);
memcpy(op->ob_sval + a->ob_size, b->ob_sval, (int) b->ob_size); memcpy(op->ob_sval + a->ob_size, b->ob_sval, (int) b->ob_size);
op->ob_sval[size] = '\0'; op->ob_sval[size] = '\0';
...@@ -784,12 +772,8 @@ string_repeat(register PyStringObject *a, register int n) ...@@ -784,12 +772,8 @@ string_repeat(register PyStringObject *a, register int n)
if (op == NULL) if (op == NULL)
return PyErr_NoMemory(); return PyErr_NoMemory();
PyObject_INIT_VAR(op, &PyString_Type, size); PyObject_INIT_VAR(op, &PyString_Type, size);
#ifdef CACHE_HASH
op->ob_shash = -1; op->ob_shash = -1;
#endif
#ifdef INTERN_STRINGS
op->ob_sinterned = NULL; op->ob_sinterned = NULL;
#endif
for (i = 0; i < size; i += a->ob_size) for (i = 0; i < size; i += a->ob_size)
memcpy(op->ob_sval+i, a->ob_sval, (int) a->ob_size); memcpy(op->ob_sval+i, a->ob_sval, (int) a->ob_size);
op->ob_sval[size] = '\0'; op->ob_sval[size] = '\0';
...@@ -945,15 +929,11 @@ string_hash(PyStringObject *a) ...@@ -945,15 +929,11 @@ string_hash(PyStringObject *a)
register unsigned char *p; register unsigned char *p;
register long x; register long x;
#ifdef CACHE_HASH
if (a->ob_shash != -1) if (a->ob_shash != -1)
return a->ob_shash; return a->ob_shash;
#ifdef INTERN_STRINGS
if (a->ob_sinterned != NULL) if (a->ob_sinterned != NULL)
return (a->ob_shash = return (a->ob_shash =
((PyStringObject *)(a->ob_sinterned))->ob_shash); ((PyStringObject *)(a->ob_sinterned))->ob_shash);
#endif
#endif
len = a->ob_size; len = a->ob_size;
p = (unsigned char *) a->ob_sval; p = (unsigned char *) a->ob_sval;
x = *p << 7; x = *p << 7;
...@@ -962,9 +942,7 @@ string_hash(PyStringObject *a) ...@@ -962,9 +942,7 @@ string_hash(PyStringObject *a)
x ^= a->ob_size; x ^= a->ob_size;
if (x == -1) if (x == -1)
x = -2; x = -2;
#ifdef CACHE_HASH
a->ob_shash = x; a->ob_shash = x;
#endif
return x; return x;
} }
...@@ -2730,14 +2708,10 @@ str_subtype_new(PyTypeObject *type, PyObject *args, PyObject *kwds) ...@@ -2730,14 +2708,10 @@ str_subtype_new(PyTypeObject *type, PyObject *args, PyObject *kwds)
pnew = type->tp_alloc(type, n); pnew = type->tp_alloc(type, n);
if (pnew != NULL) { if (pnew != NULL) {
memcpy(PyString_AS_STRING(pnew), PyString_AS_STRING(tmp), n+1); memcpy(PyString_AS_STRING(pnew), PyString_AS_STRING(tmp), n+1);
#ifdef CACHE_HASH
((PyStringObject *)pnew)->ob_shash = ((PyStringObject *)pnew)->ob_shash =
((PyStringObject *)tmp)->ob_shash; ((PyStringObject *)tmp)->ob_shash;
#endif
#ifdef INTERN_STRINGS
((PyStringObject *)pnew)->ob_sinterned = ((PyStringObject *)pnew)->ob_sinterned =
((PyStringObject *)tmp)->ob_sinterned; ((PyStringObject *)tmp)->ob_sinterned;
#endif
} }
Py_DECREF(tmp); Py_DECREF(tmp);
return pnew; return pnew;
...@@ -3579,7 +3553,6 @@ PyString_Format(PyObject *format, PyObject *args) ...@@ -3579,7 +3553,6 @@ PyString_Format(PyObject *format, PyObject *args)
} }
#ifdef INTERN_STRINGS
/* This dictionary will leak at PyString_Fini() time. That's acceptable /* This dictionary will leak at PyString_Fini() time. That's acceptable
* because PyString_Fini() specifically frees interned strings that are * because PyString_Fini() specifically frees interned strings that are
...@@ -3656,8 +3629,6 @@ PyString_InternFromString(const char *cp) ...@@ -3656,8 +3629,6 @@ PyString_InternFromString(const char *cp)
return s; return s;
} }
#endif
void void
PyString_Fini(void) PyString_Fini(void)
{ {
...@@ -3670,7 +3641,6 @@ PyString_Fini(void) ...@@ -3670,7 +3641,6 @@ PyString_Fini(void)
Py_XDECREF(nullstring); Py_XDECREF(nullstring);
nullstring = NULL; nullstring = NULL;
#endif #endif
#ifdef INTERN_STRINGS
if (interned) { if (interned) {
int pos, changed; int pos, changed;
PyObject *key, *value; PyObject *key, *value;
...@@ -3685,10 +3655,8 @@ PyString_Fini(void) ...@@ -3685,10 +3655,8 @@ PyString_Fini(void)
} }
} while (changed); } while (changed);
} }
#endif
} }
#ifdef INTERN_STRINGS
void _Py_ReleaseInternedStrings(void) void _Py_ReleaseInternedStrings(void)
{ {
if (interned) { if (interned) {
...@@ -3698,4 +3666,3 @@ void _Py_ReleaseInternedStrings(void) ...@@ -3698,4 +3666,3 @@ void _Py_ReleaseInternedStrings(void)
interned = NULL; interned = NULL;
} }
} }
#endif /* INTERN_STRINGS */
Python/import.c
View file @ 1f7df359
...@@ -978,7 +978,6 @@ find_module(char *realname, PyObject *path, char *buf, size_t buflen, ...@@ -978,7 +978,6 @@ find_module(char *realname, PyObject *path, char *buf, size_t buflen,
if (strlen(buf) != len) if (strlen(buf) != len)
continue; /* v contains '\0' */ continue; /* v contains '\0' */
#ifdef macintosh #ifdef macintosh
#ifdef INTERN_STRINGS
/* /*
** Speedup: each sys.path item is interned, and ** Speedup: each sys.path item is interned, and
** FindResourceModule remembers which items refer to ** FindResourceModule remembers which items refer to
...@@ -987,7 +986,6 @@ find_module(char *realname, PyObject *path, char *buf, size_t buflen, ...@@ -987,7 +986,6 @@ find_module(char *realname, PyObject *path, char *buf, size_t buflen,
*/ */
PyString_InternInPlace(&PyList_GET_ITEM(path, i)); PyString_InternInPlace(&PyList_GET_ITEM(path, i));
v = PyList_GET_ITEM(path, i); v = PyList_GET_ITEM(path, i);
#endif
if (PyMac_FindResourceModule((PyStringObject *)v, name, buf)) { if (PyMac_FindResourceModule((PyStringObject *)v, name, buf)) {
static struct filedescr resfiledescr = static struct filedescr resfiledescr =
{"", "", PY_RESOURCE}; {"", "", PY_RESOURCE};
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7