Rewrite the type lookup using flags where possible, eliminating need for "is_py3k" hack

1 files changed, 59 insertions, 43 deletions

diff --git a/libpython.py b/libpython.py
index 1ede863..2f1e0ec 100644
--- a/libpython.py
+++ b/libpython.py
@@ -49,6 +49,16 @@ TODO: better handling of "instance"
 
 import gdb
 
+Py_TPFLAGS_INT_SUBCLASS      = (1L<<23)
+Py_TPFLAGS_LONG_SUBCLASS     = (1L<<24)
+Py_TPFLAGS_LIST_SUBCLASS     = (1L<<25)
+Py_TPFLAGS_TUPLE_SUBCLASS    = (1L<<26)
+Py_TPFLAGS_STRING_SUBCLASS   = (1L<<27)
+Py_TPFLAGS_UNICODE_SUBCLASS  = (1L<<28)
+Py_TPFLAGS_DICT_SUBCLASS     = (1L<<29)
+Py_TPFLAGS_BASE_EXC_SUBCLASS = (1L<<30)
+Py_TPFLAGS_TYPE_SUBCLASS     = (1L<<31)
+
 class NullPyObjectPtr(RuntimeError):
     pass
 
@@ -63,17 +73,6 @@ def safe_range(val):
     # threshold in case the data was corrupted
     return xrange(safety_limit(val))
 
-def is_py3k():
-    # This code assumes that a libpython's DWARF data has actually been
-    # loaded by the point that this function is called
-    sym = gdb.lookup_symbol('PyBytes_Type')[0]
-    if sym:
-        #...then PyBytes_Type exists, assume this is libpython3.*
-        return True
-    else:
-        #...then PyBytes_Type doesn't exist, assume this is libpython2.*
-        return False
-
 class PyObjectPtr(object):
     """
     Class wrapping a gdb.Value that's a either a (PyObject*) within the
@@ -168,31 +167,62 @@ class PyObjectPtr(object):
                         long(self._gdbval))
 
     @classmethod
-    def subclass_for_tp_name(cls, tp_name):
-        if tp_name == 'str':
-            if is_py3k():
-                return PyUnicodeObjectPtr
-            else:
-                return PyStringObjectPtr
-        if tp_name == 'int':
-            if is_py3k():
-                return PyLongObjectPtr
-            else:
-                return PyIntObjectPtr
+    def subclass_from_type(cls, t):
+        '''
+        Given a PyTypeObjectPtr instance wrapping a gdb.Value that's a
+        (PyTypeObject*), determine the corresponding subclass of PyObjectPtr
+        to use
+
+        Ideally, we would look up the symbols for the global types, but that
+        isn't working yet:
+          (gdb) python print gdb.lookup_symbol('PyList_Type')[0].value
+          Traceback (most recent call last):
+            File "<string>", line 1, in <module>
+          NotImplementedError: Symbol type not yet supported in Python scripts.
+          Error while executing Python code.
+        
+        For now, we use tp_flags, after doing some string comparisons on the
+        tp_name for some special-cases that don't seem to be visible through flags
+        '''
+        try:
+            tp_name = t.field('tp_name').string()
+            tp_flags = int(t.field('tp_flags'))
+        except RuntimeError:
+            # Handle any kind of error e.g. NULL ptrs by simply using the base
+            # class
+            return cls
+
+        #print 'tp_flags = 0x%08x' % tp_flags
+        #print 'tp_name = %r' % tp_name
 
         name_map = {'bool' : PyBoolObjectPtr,
                     'classobj': PyClassObjectPtr,
-                    'dict': PyDictObjectPtr,
                     'instance': PyInstanceObjectPtr,
-                    'list': PyListObjectPtr,
-                    'long': PyLongObjectPtr,
                     'NoneType': PyNoneStructPtr,
-                    'tuple': PyTupleObjectPtr,
                     'frame': PyFrameObjectPtr,
-                    'unicode': PyUnicodeObjectPtr,
                     }
         if tp_name in name_map:
             return name_map[tp_name]
+
+        if tp_flags & Py_TPFLAGS_INT_SUBCLASS:
+            return PyIntObjectPtr
+        if tp_flags & Py_TPFLAGS_LONG_SUBCLASS:
+            return PyLongObjectPtr
+        if tp_flags & Py_TPFLAGS_LIST_SUBCLASS:
+            return PyListObjectPtr
+        if tp_flags & Py_TPFLAGS_TUPLE_SUBCLASS:
+            return PyTupleObjectPtr
+        if tp_flags & Py_TPFLAGS_STRING_SUBCLASS:
+            return PyStringObjectPtr
+        if tp_flags & Py_TPFLAGS_UNICODE_SUBCLASS:
+            return PyUnicodeObjectPtr
+        if tp_flags & Py_TPFLAGS_DICT_SUBCLASS:
+            return PyDictObjectPtr
+        #if tp_flags & Py_TPFLAGS_BASE_EXC_SUBCLASS:
+        #    return something
+        #if tp_flags & Py_TPFLAGS_TYPE_SUBCLASS:
+        #    return PyTypeObjectPtr
+
         # Use the base class:
         return cls
 
@@ -200,26 +230,12 @@ class PyObjectPtr(object):
     def from_pyobject_ptr(cls, gdbval):
         '''
         Try to locate the appropriate derived class dynamically, and cast
-        the pointer accordingly:
-        For now, we just do string comparison on the tp_name
-        Other approaches:
-        (i) look up the symbols for the global types, but that isn't working yet:
-            (gdb) python print gdb.lookup_symbol('PyList_Type')[0].value
-            Traceback (most recent call last):
-              File "<string>", line 1, in <module>
-            NotImplementedError: Symbol type not yet supported in Python scripts.
-            Error while executing Python code.
-        (ii) look at tp_flags, looking e.g. for Py_TPFLAGS_LIST_SUBCLASS however
-        this would rely on the values of those flags.
-        
-        So we go with the simple approach of looking at tp_name
+        the pointer accordingly.
         '''
         # 
         try:
             p = PyObjectPtr(gdbval)
-            t = p.type()
-            tp_name = t.field('tp_name').string()
-            cls = cls.subclass_for_tp_name(tp_name)
+            cls = cls.subclass_from_type(p.type())
             return cls(gdbval, cast_to=cls.get_gdb_type())
         except RuntimeError:
             # Handle any kind of error e.g. NULL ptrs by simply using the base