Issue #12170: The count(), find(), rfind(), index() and rindex() methods

of bytes and bytearray objects now accept an integer between 0 and 255
as their first argument.  Patch by Petri Lehtinen.

3 files changed, 133 insertions, 31 deletions

diff --git a/Objects/bytearrayobject.c b/Objects/bytearrayobject.c
index 0c9add0..6f0b28d 100644
--- a/Objects/bytearrayobject.c
+++ b/Objects/bytearrayobject.c
@@ -1071,24 +1071,41 @@ Py_LOCAL_INLINE(Py_ssize_t)
 bytearray_find_internal(PyByteArrayObject *self, PyObject *args, int dir)
 {
     PyObject *subobj;
+    char byte;
     Py_buffer subbuf;
+    const char *sub;
+    Py_ssize_t sub_len;
     Py_ssize_t start=0, end=PY_SSIZE_T_MAX;
     Py_ssize_t res;
 
-    if (!stringlib_parse_args_finds("find/rfind/index/rindex",
-                                    args, &subobj, &start, &end))
-        return -2;
-    if (_getbuffer(subobj, &subbuf) < 0)
+    if (!stringlib_parse_args_finds_byte("find/rfind/index/rindex",
+                                         args, &subobj, &byte, &start, &end))
         return -2;
+
+    if (subobj) {
+        if (_getbuffer(subobj, &subbuf) < 0)
+            return -2;
+
+        sub = subbuf.buf;
+        sub_len = subbuf.len;
+    }
+    else {
+        sub = &byte;
+        sub_len = 1;
+    }
+
     if (dir > 0)
         res = stringlib_find_slice(
             PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
-            subbuf.buf, subbuf.len, start, end);
+            sub, sub_len, start, end);
     else
         res = stringlib_rfind_slice(
             PyByteArray_AS_STRING(self), PyByteArray_GET_SIZE(self),
-            subbuf.buf, subbuf.len, start, end);
-    PyBuffer_Release(&subbuf);
+            sub, sub_len, start, end);
+
+    if (subobj)
+        PyBuffer_Release(&subbuf);
+
     return res;
 }
 
@@ -1121,23 +1138,39 @@ static PyObject *
 bytearray_count(PyByteArrayObject *self, PyObject *args)
 {
     PyObject *sub_obj;
-    const char *str = PyByteArray_AS_STRING(self);
+    const char *str = PyByteArray_AS_STRING(self), *sub;
+    Py_ssize_t sub_len;
+    char byte;
     Py_ssize_t start = 0, end = PY_SSIZE_T_MAX;
+
     Py_buffer vsub;
     PyObject *count_obj;
 
-    if (!stringlib_parse_args_finds("count", args, &sub_obj, &start, &end))
+    if (!stringlib_parse_args_finds_byte("count", args, &sub_obj, &byte,
+                                         &start, &end))
         return NULL;
 
-    if (_getbuffer(sub_obj, &vsub) < 0)
-        return NULL;
+    if (sub_obj) {
+        if (_getbuffer(sub_obj, &vsub) < 0)
+            return NULL;
+
+        sub = vsub.buf;
+        sub_len = vsub.len;
+    }
+    else {
+        sub = &byte;
+        sub_len = 1;
+    }
 
     ADJUST_INDICES(start, end, PyByteArray_GET_SIZE(self));
 
     count_obj = PyLong_FromSsize_t(
-        stringlib_count(str + start, end - start, vsub.buf, vsub.len, PY_SSIZE_T_MAX)
+        stringlib_count(str + start, end - start, sub, sub_len, PY_SSIZE_T_MAX)
         );
-    PyBuffer_Release(&vsub);
+
+    if (sub_obj)
+        PyBuffer_Release(&vsub);
+
     return count_obj;
 }
 
diff --git a/Objects/bytesobject.c b/Objects/bytesobject.c
index 17e31b9..7438a70 100644
--- a/Objects/bytesobject.c
+++ b/Objects/bytesobject.c
@@ -1230,31 +1230,42 @@ Py_LOCAL_INLINE(Py_ssize_t)
 bytes_find_internal(PyBytesObject *self, PyObject *args, int dir)
 {
     PyObject *subobj;
+    char byte;
+    Py_buffer subbuf;
     const char *sub;
     Py_ssize_t sub_len;
     Py_ssize_t start=0, end=PY_SSIZE_T_MAX;
+    Py_ssize_t res;
 
-    if (!stringlib_parse_args_finds("find/rfind/index/rindex",
-                                    args, &subobj, &start, &end))
+    if (!stringlib_parse_args_finds_byte("find/rfind/index/rindex",
+                                         args, &subobj, &byte, &start, &end))
         return -2;
 
-    if (PyBytes_Check(subobj)) {
-        sub = PyBytes_AS_STRING(subobj);
-        sub_len = PyBytes_GET_SIZE(subobj);
+    if (subobj) {
+        if (_getbuffer(subobj, &subbuf) < 0)
+            return -2;
+
+        sub = subbuf.buf;
+        sub_len = subbuf.len;
+    }
+    else {
+        sub = &byte;
+        sub_len = 1;
     }
-    else if (PyObject_AsCharBuffer(subobj, &sub, &sub_len))
-        /* XXX - the "expected a character buffer object" is pretty
-           confusing for a non-expert.  remap to something else ? */
-        return -2;
 
     if (dir > 0)
-        return stringlib_find_slice(
+        res = stringlib_find_slice(
             PyBytes_AS_STRING(self), PyBytes_GET_SIZE(self),
             sub, sub_len, start, end);
     else
-        return stringlib_rfind_slice(
+        res = stringlib_rfind_slice(
             PyBytes_AS_STRING(self), PyBytes_GET_SIZE(self),
             sub, sub_len, start, end);
+
+    if (subobj)
+        PyBuffer_Release(&subbuf);
+
+    return res;
 }
 
 
@@ -1480,23 +1491,38 @@ bytes_count(PyBytesObject *self, PyObject *args)
     PyObject *sub_obj;
     const char *str = PyBytes_AS_STRING(self), *sub;
     Py_ssize_t sub_len;
+    char byte;
     Py_ssize_t start = 0, end = PY_SSIZE_T_MAX;
 
-    if (!stringlib_parse_args_finds("count", args, &sub_obj, &start, &end))
+    Py_buffer vsub;
+    PyObject *count_obj;
+
+    if (!stringlib_parse_args_finds_byte("count", args, &sub_obj, &byte,
+                                         &start, &end))
         return NULL;
 
-    if (PyBytes_Check(sub_obj)) {
-        sub = PyBytes_AS_STRING(sub_obj);
-        sub_len = PyBytes_GET_SIZE(sub_obj);
+    if (sub_obj) {
+        if (_getbuffer(sub_obj, &vsub) < 0)
+            return NULL;
+
+        sub = vsub.buf;
+        sub_len = vsub.len;
+    }
+    else {
+        sub = &byte;
+        sub_len = 1;
     }
-    else if (PyObject_AsCharBuffer(sub_obj, &sub, &sub_len))
-        return NULL;
 
     ADJUST_INDICES(start, end, PyBytes_GET_SIZE(self));
 
-    return PyLong_FromSsize_t(
+    count_obj = PyLong_FromSsize_t(
         stringlib_count(str + start, end - start, sub, sub_len, PY_SSIZE_T_MAX)
         );
+
+    if (sub_obj)
+        PyBuffer_Release(&vsub);
+
+    return count_obj;
 }
 
 
diff --git a/Objects/stringlib/find.h b/Objects/stringlib/find.h
index 7cce156..00eaf1b 100644
--- a/Objects/stringlib/find.h
+++ b/Objects/stringlib/find.h
@@ -167,4 +167,47 @@ STRINGLIB(parse_args_finds_unicode)(const char * function_name, PyObject *args,
     return 0;
 }
 
+#else /* !STRINGLIB_IS_UNICODE */
+
+/*
+Wraps stringlib_parse_args_finds() and additionally checks whether the
+first argument is an integer in range(0, 256).
+
+If this is the case, writes the integer value to the byte parameter
+and sets subobj to NULL. Otherwise, sets the first argument to subobj
+and doesn't touch byte. The other parameters are similar to those of
+stringlib_parse_args_finds().
+*/
+
+Py_LOCAL_INLINE(int)
+STRINGLIB(parse_args_finds_byte)(const char *function_name, PyObject *args,
+                                 PyObject **subobj, char *byte,
+                                 Py_ssize_t *start, Py_ssize_t *end)
+{
+    PyObject *tmp_subobj;
+    Py_ssize_t ival;
+
+    if(!STRINGLIB(parse_args_finds)(function_name, args, &tmp_subobj,
+                                    start, end))
+        return 0;
+
+    ival = PyNumber_AsSsize_t(tmp_subobj, PyExc_ValueError);
+    if (ival == -1 && PyErr_Occurred()) {
+        PyErr_Clear();
+        *subobj = tmp_subobj;
+    }
+    else {
+        /* The first argument was an integer */
+        if(ival < 0 || ival > 255) {
+            PyErr_SetString(PyExc_ValueError, "byte must be in range(0, 256)");
+            return 0;
+        }
+
+        *subobj = NULL;
+        *byte = (char)ival;
+    }
+
+    return 1;
+}
+
 #endif /* STRINGLIB_IS_UNICODE */