Patch #976880: add mmap .rfind() method, and 'end' paramter to .find().

Contributed by John Lenton.

Doc/library/mmap.rst

@@ -139,11 +139,12 @@ Memory-mapped file objects support the following methods:
    an exception being raised.
 
 
-.. method:: mmap.find(string[, start])
+.. method:: mmap.find(string[, start[, end]])
 
-   Returns the lowest index in the object where the substring *string* is found.
+   Returns the lowest index in the object where the substring *string* is found,
-   Returns ``-1`` on failure.  *start* is the index at which the search begins, and
+   such that *string* is contained in the range [*start*, *end*]. Optional
-   defaults to zero.
+   arguments *start* and *end* are interpreted as in slice notation.
+   Returns ``-1`` on failure.
 
 
 .. method:: mmap.flush([offset, size])
@@ -188,6 +189,14 @@ Memory-mapped file objects support the following methods:
    :exc:`TypeError` exception.
 
 
+.. method:: mmap.rfind(string[, start[, end]])
+
+   Returns the highest index in the object where the substring *string* is
+   found, such that *string* is contained in the range [*start*,
+   *end*]. Optional arguments *start* and *end* are interpreted as in slice
+   notation.  Returns ``-1`` on failure.
+
+
 .. method:: mmap.seek(pos[, whence])
 
    Set the file's current position.  *whence* argument is optional and defaults to

Lib/test/test_mmap.py

@@ -255,6 +255,42 @@ class MmapTests(unittest.TestCase):
                 self.assertEqual(m.find(slice + 'x'), -1)
         m.close()
 
+    def test_find_end(self):
+        # test the new 'end' parameter works as expected
+        f = open(TESTFN, 'w+')
+        data = 'one two ones'
+        n = len(data)
+        f.write(data)
+        f.flush()
+        m = mmap.mmap(f.fileno(), n)
+        f.close()
+
+        self.assertEqual(m.find('one'), 0)
+        self.assertEqual(m.find('ones'), 8)
+        self.assertEqual(m.find('one', 0, -1), 0)
+        self.assertEqual(m.find('one', 1), 8)
+        self.assertEqual(m.find('one', 1, -1), 8)
+        self.assertEqual(m.find('one', 1, -2), -1)
+
+
+    def test_rfind(self):
+        # test the new 'end' parameter works as expected
+        f = open(TESTFN, 'w+')
+        data = 'one two ones'
+        n = len(data)
+        f.write(data)
+        f.flush()
+        m = mmap.mmap(f.fileno(), n)
+        f.close()
+
+        self.assertEqual(m.rfind('one'), 8)
+        self.assertEqual(m.rfind('one '), 0)
+        self.assertEqual(m.rfind('one', 0, -1), 8)
+        self.assertEqual(m.rfind('one', 0, -2), 0)
+        self.assertEqual(m.rfind('one', 1, -1), 8)
+        self.assertEqual(m.rfind('one', 1, -2), -1)
+
+
     def test_double_close(self):
         # make sure a double close doesn't crash on Solaris (Bug# 665913)
         f = open(TESTFN, 'w+')

Misc/ACKS

@@ -395,6 +395,7 @@ Kip Lehman
 Joerg Lehmann
 Luke Kenneth Casson Leighton
 Marc-Andre Lemburg
+John Lenton
 Mark Levinson
 William Lewis
 Robert van Liere

Misc/NEWS

@@ -1011,6 +1011,10 @@ Library
 Extension Modules
 -----------------
 
+- Patch 976880: ``mmap`` objects now have an ``rfind`` method that
+  works as expected.  ``mmap.find`` also takes an optional ``end``
+  parameter.
+
 - _winreg's HKEY object has gained __enter__ and __exit__ methods to support
   the context manager protocol.  The _winreg module also gained a new function
   ``ExpandEnvironmentStrings`` to expand REG_EXPAND_SZ keys.

Modules/mmapmodule.c

@@ -248,19 +248,22 @@ mmap_read_method(mmap_object *self,
 }
 
 static PyObject *
-mmap_find_method(mmap_object *self,
+mmap_gfind(mmap_object *self,
-		 PyObject *args)
+	   PyObject *args,
+	   int reverse)
 {
 	Py_ssize_t start = self->pos;
+        Py_ssize_t end = self->size;
 	char *needle;
 	Py_ssize_t len;
 
 	CHECK_VALID(NULL);
-	if (!PyArg_ParseTuple(args, "s#|n:find", &needle, &len, &start)) {
+	if (!PyArg_ParseTuple(args, reverse ? "s#|nn:rfind" : "s#|nn:find",
+			      &needle, &len, &start, &end)) {
 		return NULL;
 	} else {
 		char *p;
-		char *e = self->data + self->size;
+		char sign = reverse ? -1 : 1;
 
                 if (start < 0)
 			start += self->size;
@@ -269,7 +272,18 @@ mmap_find_method(mmap_object *self,
                 else if ((size_t)start > self->size)
 			start = self->size;
 
-		for (p = self->data + start; p + len <= e; ++p) {
+                if (end < 0)
+			end += self->size;
+		if (end < 0)
+			end = 0;
+		else if ((size_t)end > self->size)
+			end = self->size;
+
+		start += (Py_ssize_t)self->data;
+		end += (Py_ssize_t)self->data;
+
+		for (p = (char *)(reverse ? end - len : start);
+		     p >= (char *)start && p + len <= (char *)end; p+=sign) {
 			Py_ssize_t i;
 			for (i = 0; i < len && needle[i] == p[i]; ++i)
 				/* nothing */;
@@ -281,6 +295,20 @@ mmap_find_method(mmap_object *self,
 	}
 }
 
+static PyObject *
+mmap_find_method(mmap_object *self,
+		 PyObject *args)
+{
+	return mmap_gfind(self, args, 0);
+}
+
+static PyObject *
+mmap_rfind_method(mmap_object *self,
+		 PyObject *args)
+{
+	return mmap_gfind(self, args, 1);
+}
+
 static int
 is_writeable(mmap_object *self)
 {
@@ -593,6 +621,7 @@ mmap_move_method(mmap_object *self, PyObject *args)
 static struct PyMethodDef mmap_object_methods[] = {
 	{"close",	(PyCFunction) mmap_close_method,	METH_NOARGS},
 	{"find",	(PyCFunction) mmap_find_method,		METH_VARARGS},
+	{"rfind",	(PyCFunction) mmap_rfind_method,	METH_VARARGS},
 	{"flush",	(PyCFunction) mmap_flush_method,	METH_VARARGS},
 	{"move",	(PyCFunction) mmap_move_method,		METH_VARARGS},
 	{"read",	(PyCFunction) mmap_read_method,		METH_VARARGS},