Mercurial > cpython-lru_cache-weakref

diff Objects/stringobject.c @ 17710:6199e5bd370b legacy-trunk

Find changesets by keywords (author, files, the commit message), revision number or hash, or revset expression.
Intern 1-character strings as soon as they're created. As-is, they aren't interned when created, so the cached versions generally aren't ever interned. With the patch, the Py_INCREF(t); *p = t; Py_DECREF(s); return; indirection block in PyString_InternInPlace() is never executed during a full run of the test suite, but was executed very many times before. So I'm trading more work when creating one-character strings for doing less work later. Note that the "more work" here can happen at most 256 times per program run, so it's trivial. The same reasoning accounts for the patch's simplification of string_item (the new version can call PyString_FromStringAndSize() no more than 256 times per run, so there's no point to inlining that stuff -- if we were serious about saving time here, we'd pre-initialize the characters vector so that no runtime testing at all was needed!).
author Tim Peters <tim.peters@gmail.com>
date Tue, 08 May 2001 22:33:50 +0000
parents 9e093bf8f6a9
children 350be19ca6a5
line wrap: on
line diff
--- a/Objects/stringobject.c	Tue May 08 15:43:37 2001 +0000
+++ b/Objects/stringobject.c	Tue May 08 22:33:50 2001 +0000
@@ -36,7 +36,7 @@
 PyObject *
 PyString_FromStringAndSize(const char *str, int size)
 {
-	register PyStringObject *op;
+	PyStringObject *op;
 #ifndef DONT_SHARE_SHORT_STRINGS
 	if (size == 0 && (op = nullstring) != NULL) {
 #ifdef COUNT_ALLOCS
@@ -73,9 +73,11 @@
 	op->ob_sval[size] = '\0';
 #ifndef DONT_SHARE_SHORT_STRINGS
 	if (size == 0) {
+		PyString_InternInPlace(&(PyObject *)op);
 		nullstring = op;
 		Py_INCREF(op);
 	} else if (size == 1 && str != NULL) {
+		PyString_InternInPlace(&(PyObject *)op);
 		characters[*str & UCHAR_MAX] = op;
 		Py_INCREF(op);
 	}
@@ -87,7 +89,7 @@
 PyString_FromString(const char *str)
 {
 	register size_t size = strlen(str);
-	register PyStringObject *op;
+	PyStringObject *op;
 	if (size > INT_MAX) {
 		PyErr_SetString(PyExc_OverflowError,
 			"string is too long for a Python string");
@@ -125,9 +127,11 @@
 	strcpy(op->ob_sval, str);
 #ifndef DONT_SHARE_SHORT_STRINGS
 	if (size == 0) {
+		PyString_InternInPlace(&(PyObject *)op);
 		nullstring = op;
 		Py_INCREF(op);
 	} else if (size == 1) {
+		PyString_InternInPlace(&(PyObject *)op);
 		characters[*str & UCHAR_MAX] = op;
 		Py_INCREF(op);
 	}
@@ -551,24 +555,17 @@
 {
 	int c;
 	PyObject *v;
+	char *pchar;
 	if (i < 0 || i >= a->ob_size) {
 		PyErr_SetString(PyExc_IndexError, "string index out of range");
 		return NULL;
 	}
-	c = a->ob_sval[i] & UCHAR_MAX;
+	pchar = a->ob_sval + i;
+	c = *pchar & UCHAR_MAX;
 	v = (PyObject *) characters[c];
-#ifdef COUNT_ALLOCS
-	if (v != NULL)
-		one_strings++;
-#endif
-	if (v == NULL) {
-		v = PyString_FromStringAndSize((char *)NULL, 1);
-		if (v == NULL)
-			return NULL;
-		characters[c] = (PyStringObject *) v;
-		((PyStringObject *)v)->ob_sval[0] = c;
-	}
-	Py_INCREF(v);
+	if (v == NULL)
+		v = PyString_FromStringAndSize(pchar, 1);
+	Py_XINCREF(v);
 	return v;
 }