merged in latest cython-devel

Cython/Compiler/ExprNodes.py

@@ -1996,6 +1996,15 @@ class IndexNode(ExprNode):
             elif not skip_child_analysis:
                 self.index.analyse_types(env)
             self.original_index_type = self.index.type
+            if base_type is PyrexTypes.c_py_unicode_type:
+                # we infer Py_UNICODE for unicode strings in some
+                # cases, but indexing must still work for them
+                if self.index.constant_result in (0, -1):
+                    # FIXME: we know that this node is redundant -
+                    # currently, this needs to get handled in Optimize.py
+                    pass
+                self.base = self.base.coerce_to_pyobject(env)
+                base_type = self.base.type
             if base_type.is_pyobject:
                 if self.index.type.is_int:
                     if (not setting

Cython/Compiler/Optimize.py

@@ -1616,7 +1616,6 @@ class OptimizeBuiltinCalls(Visitor.EnvTransform):
 
     _map_to_capi_len_function = {
         Builtin.unicode_type   : "PyUnicode_GET_SIZE",
-        Builtin.str_type       : "Py_SIZE", # works in Py2 and Py3
         Builtin.bytes_type     : "PyBytes_GET_SIZE",
         Builtin.list_type      : "PyList_GET_SIZE",
         Builtin.tuple_type     : "PyTuple_GET_SIZE",
@@ -1645,13 +1644,15 @@ class OptimizeBuiltinCalls(Visitor.EnvTransform):
             cfunc_name = self._map_to_capi_len_function(arg.type)
             if cfunc_name is None:
                 return node
-            if not arg.is_literal:
-                arg = arg.as_none_safe_node(
-                    "object of type 'NoneType' has no len()")
+            arg = arg.as_none_safe_node(
+                "object of type 'NoneType' has no len()")
             new_node = ExprNodes.PythonCapiCallNode(
                 node.pos, cfunc_name, self.PyObject_Size_func_type,
                 args = [arg],
                 is_temp = node.is_temp)
+        elif arg.type is PyrexTypes.c_py_unicode_type:
+            return ExprNodes.IntNode(node.pos, value='1', constant_result=1,
+                                     type=node.type)
         else:
             return node
         if node.type not in (PyrexTypes.c_size_t_type, PyrexTypes.c_py_ssize_t_type):

tests/run/builtin_len.pyx

@@ -38,9 +38,9 @@ def len_bytes(bytes s):
     """
     return len(s)
 
-@cython.test_assert_path_exists(
-    "//CoerceToPyTypeNode",
-    "//PythonCapiCallNode")
+#@cython.test_assert_path_exists(
+#    "//CoerceToPyTypeNode",
+#    "//PythonCapiCallNode")
 def len_str(str s):
     """
     >>> len('abcdefg')

tests/run/py_unicode_type.pyx

@@ -119,3 +119,46 @@ def unicode_methods(Py_UNICODE uchar):
         uchar.upper(),
         uchar.title(),
         ]
+
+@cython.test_assert_path_exists('//IntNode')
+@cython.test_fail_if_path_exists('//SimpleCallNode',
+                                 '//PythonCapiCallNode')
+def len_uchar(Py_UNICODE uchar):
+    """
+    >>> len_uchar(ord('A'))
+    1
+    """
+    return len(uchar)
+
+def index_uchar(Py_UNICODE uchar, Py_ssize_t i):
+    """
+    >>> index_uchar(ord('A'), 0) == ('A', 'A', 'A')
+    True
+    >>> index_uchar(ord('A'), -1) == ('A', 'A', 'A')
+    True
+    >>> index_uchar(ord('A'), 1)
+    Traceback (most recent call last):
+    IndexError: string index out of range
+    """
+    return uchar[0], uchar[-1], uchar[i]
+
+mixed_ustring = u'AbcDefGhIjKlmnoP'
+lower_ustring = mixed_ustring.lower()
+upper_ustring = mixed_ustring.lower()
+
+@cython.test_assert_path_exists('//PythonCapiCallNode',
+                                '//ForFromStatNode')
+@cython.test_fail_if_path_exists('//SimpleCallNode',
+                                 '//ForInStatNode')
+def count_lower_case_characters(unicode ustring):
+    """
+    >>> count_lower_case_characters(mixed_ustring)
+    10
+    >>> count_lower_case_characters(lower_ustring)
+    16
+    """
+    cdef Py_ssize_t count = 0
+    for uchar in ustring:
+         if uchar.islower():
+             count += 1
+    return count