gh-117431: Adapt str.find and friends to Argument Clinic (#117468)

This change gives a significant speedup, as the METH_FASTCALL calling
convention is now used. The following methods are adapted:

- str.count
- str.find
- str.index
- str.rfind
- str.rindex

1 files changed, 280 insertions, 1 deletions

diff --git a/Objects/clinic/unicodeobject.c.h b/Objects/clinic/unicodeobject.c.h
index 3f6dd8b..01c40b9 100644
--- a/Objects/clinic/unicodeobject.c.h
+++ b/Objects/clinic/unicodeobject.c.h
@@ -136,6 +136,61 @@ exit:
     return return_value;
 }
 
+PyDoc_STRVAR(unicode_count__doc__,
+"count($self, sub[, start[, end]], /)\n"
+"--\n"
+"\n"
+"Return the number of non-overlapping occurrences of substring sub in string S[start:end].\n"
+"\n"
+"Optional arguments start and end are interpreted as in slice notation.");
+
+#define UNICODE_COUNT_METHODDEF    \
+    {"count", _PyCFunction_CAST(unicode_count), METH_FASTCALL, unicode_count__doc__},
+
+static Py_ssize_t
+unicode_count_impl(PyObject *str, PyObject *substr, Py_ssize_t start,
+                   Py_ssize_t end);
+
+static PyObject *
+unicode_count(PyObject *str, PyObject *const *args, Py_ssize_t nargs)
+{
+    PyObject *return_value = NULL;
+    PyObject *substr;
+    Py_ssize_t start = 0;
+    Py_ssize_t end = PY_SSIZE_T_MAX;
+    Py_ssize_t _return_value;
+
+    if (!_PyArg_CheckPositional("count", nargs, 1, 3)) {
+        goto exit;
+    }
+    if (!PyUnicode_Check(args[0])) {
+        _PyArg_BadArgument("count", "argument 1", "str", args[0]);
+        goto exit;
+    }
+    substr = args[0];
+    if (nargs < 2) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[1], &start)) {
+        goto exit;
+    }
+    if (nargs < 3) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[2], &end)) {
+        goto exit;
+    }
+skip_optional:
+    _return_value = unicode_count_impl(str, substr, start, end);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyLong_FromSsize_t(_return_value);
+
+exit:
+    return return_value;
+}
+
 PyDoc_STRVAR(unicode_encode__doc__,
 "encode($self, /, encoding=\'utf-8\', errors=\'strict\')\n"
 "--\n"
@@ -301,6 +356,118 @@ exit:
     return return_value;
 }
 
+PyDoc_STRVAR(unicode_find__doc__,
+"find($self, sub, start=None, end=None, /)\n"
+"--\n"
+"\n"
+"Return the lowest index in S where substring sub is found, such that sub is contained within S[start:end].\n"
+"\n"
+"Optional arguments start and end are interpreted as in slice notation.\n"
+"Return -1 on failure.");
+
+#define UNICODE_FIND_METHODDEF    \
+    {"find", _PyCFunction_CAST(unicode_find), METH_FASTCALL, unicode_find__doc__},
+
+static Py_ssize_t
+unicode_find_impl(PyObject *str, PyObject *substr, Py_ssize_t start,
+                  Py_ssize_t end);
+
+static PyObject *
+unicode_find(PyObject *str, PyObject *const *args, Py_ssize_t nargs)
+{
+    PyObject *return_value = NULL;
+    PyObject *substr;
+    Py_ssize_t start = 0;
+    Py_ssize_t end = PY_SSIZE_T_MAX;
+    Py_ssize_t _return_value;
+
+    if (!_PyArg_CheckPositional("find", nargs, 1, 3)) {
+        goto exit;
+    }
+    if (!PyUnicode_Check(args[0])) {
+        _PyArg_BadArgument("find", "argument 1", "str", args[0]);
+        goto exit;
+    }
+    substr = args[0];
+    if (nargs < 2) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[1], &start)) {
+        goto exit;
+    }
+    if (nargs < 3) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[2], &end)) {
+        goto exit;
+    }
+skip_optional:
+    _return_value = unicode_find_impl(str, substr, start, end);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyLong_FromSsize_t(_return_value);
+
+exit:
+    return return_value;
+}
+
+PyDoc_STRVAR(unicode_index__doc__,
+"index($self, sub, start=None, end=None, /)\n"
+"--\n"
+"\n"
+"Return the lowest index in S where substring sub is found, such that sub is contained within S[start:end].\n"
+"\n"
+"Optional arguments start and end are interpreted as in slice notation.\n"
+"Raises ValueError when the substring is not found.");
+
+#define UNICODE_INDEX_METHODDEF    \
+    {"index", _PyCFunction_CAST(unicode_index), METH_FASTCALL, unicode_index__doc__},
+
+static Py_ssize_t
+unicode_index_impl(PyObject *str, PyObject *substr, Py_ssize_t start,
+                   Py_ssize_t end);
+
+static PyObject *
+unicode_index(PyObject *str, PyObject *const *args, Py_ssize_t nargs)
+{
+    PyObject *return_value = NULL;
+    PyObject *substr;
+    Py_ssize_t start = 0;
+    Py_ssize_t end = PY_SSIZE_T_MAX;
+    Py_ssize_t _return_value;
+
+    if (!_PyArg_CheckPositional("index", nargs, 1, 3)) {
+        goto exit;
+    }
+    if (!PyUnicode_Check(args[0])) {
+        _PyArg_BadArgument("index", "argument 1", "str", args[0]);
+        goto exit;
+    }
+    substr = args[0];
+    if (nargs < 2) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[1], &start)) {
+        goto exit;
+    }
+    if (nargs < 3) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[2], &end)) {
+        goto exit;
+    }
+skip_optional:
+    _return_value = unicode_index_impl(str, substr, start, end);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyLong_FromSsize_t(_return_value);
+
+exit:
+    return return_value;
+}
+
 PyDoc_STRVAR(unicode_isascii__doc__,
 "isascii($self, /)\n"
 "--\n"
@@ -892,6 +1059,118 @@ exit:
     return return_value;
 }
 
+PyDoc_STRVAR(unicode_rfind__doc__,
+"rfind($self, sub, start=None, end=None, /)\n"
+"--\n"
+"\n"
+"Return the highest index in S where substring sub is found, such that sub is contained within S[start:end].\n"
+"\n"
+"Optional arguments start and end are interpreted as in slice notation.\n"
+"Return -1 on failure.");
+
+#define UNICODE_RFIND_METHODDEF    \
+    {"rfind", _PyCFunction_CAST(unicode_rfind), METH_FASTCALL, unicode_rfind__doc__},
+
+static Py_ssize_t
+unicode_rfind_impl(PyObject *str, PyObject *substr, Py_ssize_t start,
+                   Py_ssize_t end);
+
+static PyObject *
+unicode_rfind(PyObject *str, PyObject *const *args, Py_ssize_t nargs)
+{
+    PyObject *return_value = NULL;
+    PyObject *substr;
+    Py_ssize_t start = 0;
+    Py_ssize_t end = PY_SSIZE_T_MAX;
+    Py_ssize_t _return_value;
+
+    if (!_PyArg_CheckPositional("rfind", nargs, 1, 3)) {
+        goto exit;
+    }
+    if (!PyUnicode_Check(args[0])) {
+        _PyArg_BadArgument("rfind", "argument 1", "str", args[0]);
+        goto exit;
+    }
+    substr = args[0];
+    if (nargs < 2) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[1], &start)) {
+        goto exit;
+    }
+    if (nargs < 3) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[2], &end)) {
+        goto exit;
+    }
+skip_optional:
+    _return_value = unicode_rfind_impl(str, substr, start, end);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyLong_FromSsize_t(_return_value);
+
+exit:
+    return return_value;
+}
+
+PyDoc_STRVAR(unicode_rindex__doc__,
+"rindex($self, sub, start=None, end=None, /)\n"
+"--\n"
+"\n"
+"Return the highest index in S where substring sub is found, such that sub is contained within S[start:end].\n"
+"\n"
+"Optional arguments start and end are interpreted as in slice notation.\n"
+"Raises ValueError when the substring is not found.");
+
+#define UNICODE_RINDEX_METHODDEF    \
+    {"rindex", _PyCFunction_CAST(unicode_rindex), METH_FASTCALL, unicode_rindex__doc__},
+
+static Py_ssize_t
+unicode_rindex_impl(PyObject *str, PyObject *substr, Py_ssize_t start,
+                    Py_ssize_t end);
+
+static PyObject *
+unicode_rindex(PyObject *str, PyObject *const *args, Py_ssize_t nargs)
+{
+    PyObject *return_value = NULL;
+    PyObject *substr;
+    Py_ssize_t start = 0;
+    Py_ssize_t end = PY_SSIZE_T_MAX;
+    Py_ssize_t _return_value;
+
+    if (!_PyArg_CheckPositional("rindex", nargs, 1, 3)) {
+        goto exit;
+    }
+    if (!PyUnicode_Check(args[0])) {
+        _PyArg_BadArgument("rindex", "argument 1", "str", args[0]);
+        goto exit;
+    }
+    substr = args[0];
+    if (nargs < 2) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[1], &start)) {
+        goto exit;
+    }
+    if (nargs < 3) {
+        goto skip_optional;
+    }
+    if (!_PyEval_SliceIndex(args[2], &end)) {
+        goto exit;
+    }
+skip_optional:
+    _return_value = unicode_rindex_impl(str, substr, start, end);
+    if ((_return_value == -1) && PyErr_Occurred()) {
+        goto exit;
+    }
+    return_value = PyLong_FromSsize_t(_return_value);
+
+exit:
+    return return_value;
+}
+
 PyDoc_STRVAR(unicode_rjust__doc__,
 "rjust($self, width, fillchar=\' \', /)\n"
 "--\n"
@@ -1609,4 +1888,4 @@ skip_optional_pos:
 exit:
     return return_value;
 }
-/*[clinic end generated code: output=1734aa1fcc9b076a input=a9049054013a1b77]*/
+/*[clinic end generated code: output=3aa49013ffa3fa93 input=a9049054013a1b77]*/