diff options
Diffstat (limited to 'Objects')
| -rw-r--r-- | Objects/stringobject.c | 39 |
1 files changed, 26 insertions, 13 deletions
diff --git a/Objects/stringobject.c b/Objects/stringobject.c index b90221a..0cbf439 100644 --- a/Objects/stringobject.c +++ b/Objects/stringobject.c @@ -1002,8 +1002,12 @@ string_slice(register PyStringObject *a, register int i, register int j) static int string_contains(PyObject *a, PyObject *el) { - const char *lhs, *rhs, *end; - int size; + char *s = PyString_AS_STRING(a); + const char *sub = PyString_AS_STRING(el); + char *last; + int len_sub = PyString_GET_SIZE(el); + int shortsub; + char firstchar, lastchar; if (!PyString_CheckExact(el)) { #ifdef Py_USING_UNICODE @@ -1016,20 +1020,29 @@ string_contains(PyObject *a, PyObject *el) return -1; } } - size = PyString_GET_SIZE(el); - rhs = PyString_AS_STRING(el); - lhs = PyString_AS_STRING(a); - /* optimize for a single character */ - if (size == 1) - return memchr(lhs, *rhs, PyString_GET_SIZE(a)) != NULL; - - end = lhs + (PyString_GET_SIZE(a) - size); - while (lhs <= end) { - if (memcmp(lhs++, rhs, size) == 0) + if (len_sub == 0) + return 1; + /* last points to one char beyond the start of the rightmost + substring. When s<last, there is still room for a possible match + and s[0] through s[len_sub-1] will be in bounds. + shortsub is len_sub minus the last character which is checked + separately just before the memcmp(). That check helps prevent + false starts and saves the setup time for memcmp(). + */ + firstchar = sub[0]; + shortsub = len_sub - 1; + lastchar = sub[shortsub]; + last = s + PyString_GET_SIZE(a) - len_sub + 1; + while (s < last) { + s = memchr(s, firstchar, last-s); + if (s == NULL) + return 0; + assert(s < last); + if (s[shortsub] == lastchar && memcmp(s, sub, shortsub) == 0) return 1; + s++; } - return 0; } |