fix: make sure PyUnicode_READY() is called before unicode iteration

jparyani · Apr 15, 2012 · ee471d5 · ee471d5
1 parent 04ff4ec
commit ee471d5
Show file tree

Hide file tree

Showing 2 changed files with 26 additions and 6 deletions.
diff --git a/Cython/Compiler/Optimize.py b/Cython/Compiler/Optimize.py
@@ -283,14 +283,20 @@ def _transform_bytes_iteration(self, node, slice_node, reversed=False):
                     ),
                 reversed = reversed))
 
+    PyUnicode_READY_func_type = PyrexTypes.CFuncType(
+        PyrexTypes.c_int_type, [
+            PyrexTypes.CFuncTypeArg("s", PyrexTypes.py_object_type, None)
+        ],
+        exception_value='-1')
+
     PyUnicode_GET_LENGTH_func_type = PyrexTypes.CFuncType(
         PyrexTypes.c_py_ssize_t_type, [
-            PyrexTypes.CFuncTypeArg("s", Builtin.unicode_type, None)
+            PyrexTypes.CFuncTypeArg("s", PyrexTypes.py_object_type, None)
         ])
 
     PyUnicode_KIND_func_type = PyrexTypes.CFuncType(
         PyrexTypes.c_int_type, [
-            PyrexTypes.CFuncTypeArg("s", Builtin.unicode_type, None)
+            PyrexTypes.CFuncTypeArg("s", PyrexTypes.py_object_type, None)
         ])
 
     PyUnicode_READ_func_type = PyrexTypes.CFuncType(
@@ -302,7 +308,7 @@ def _transform_bytes_iteration(self, node, slice_node, reversed=False):
 
     PyUnicode_DATA_func_type = PyrexTypes.CFuncType(
         PyrexTypes.c_void_ptr_type, [
-            PyrexTypes.CFuncTypeArg("s", Builtin.unicode_type, None)
+            PyrexTypes.CFuncTypeArg("s", PyrexTypes.py_object_type, None)
         ])
 
     def _transform_unicode_iteration(self, node, slice_node, reversed=False):
@@ -349,7 +355,7 @@ def _transform_unicode_iteration(self, node, slice_node, reversed=False):
                 slice_node.pos, "__Pyx_PyUnicode_READ",
                 self.PyUnicode_READ_func_type,
                 args = [kind_temp, data_temp, counter_temp],
-                is_temp = 0,
+                is_temp = False,
                 ))
         body = Nodes.StatListNode(
             node.pos,
@@ -366,9 +372,21 @@ def _transform_unicode_iteration(self, node, slice_node, reversed=False):
 
         loop_node = UtilNodes.TempsBlockNode(
             node.pos, temps=[counter], body=loop_node)
-        for temp in (kind_temp, data_temp, unpack_temp_node): # last is outermost temp
+        for temp in (kind_temp, data_temp):
             loop_node = UtilNodes.LetNode(temp, loop_node)
-        return loop_node
+
+        setup_node = Nodes.ExprStatNode(
+            node.pos,
+            expr = ExprNodes.PythonCapiCallNode(
+                slice_node.pos, "__Pyx_PyUnicode_READY",
+                self.PyUnicode_READY_func_type,
+                args = [unpack_temp_node],
+                is_temp = True,
+                result_is_used = False,
+                ))
+        return UtilNodes.LetNode(
+            unpack_temp_node,
+            Nodes.StatListNode(node.pos, stats=[setup_node, loop_node]))
 
     def _transform_carray_iteration(self, node, slice_node, reversed=False):
         neg_step = False

diff --git a/Cython/Utility/ModuleSetupCode.c b/Cython/Utility/ModuleSetupCode.c
@@ -119,13 +119,15 @@
 /* new Py3.3 unicode type (PEP 393) */
 #if PY_VERSION_HEX > 0x03030000 && defined(PyUnicode_KIND)
   #define CYTHON_PEP393_ENABLED 1
+  #define __Pyx_PyUnicode_READY(u)        PyUnicode_READY(u)
   #define __Pyx_PyUnicode_KIND(u)         PyUnicode_KIND(u)
   #define __Pyx_PyUnicode_DATA(u)         PyUnicode_DATA(u)
   #define __Pyx_PyUnicode_GET_LENGTH(u)   PyUnicode_GET_LENGTH(u)
   #define __Pyx_PyUnicode_READ_CHAR(u, i) PyUnicode_READ_CHAR(u, i)
   #define __Pyx_PyUnicode_READ(k, d, i)   PyUnicode_READ(k, d, i)
 #else
   #define CYTHON_PEP393_ENABLED 0
+  #define __Pyx_PyUnicode_READY(u)        (0)
   #define __Pyx_PyUnicode_KIND(u)         (0)    /* PyUnicode_WCHAR_KIND */
   #define __Pyx_PyUnicode_DATA(u)         PyUnicode_AS_UNICODE(u)
   #define __Pyx_PyUnicode_GET_LENGTH(u)   PyUnicode_GET_SIZE(u)