Fix #71: make cloudpickle Python 3.6 compatible

Also add a cache to the function which extracts global-referencing opcodes,
speeding up repetitive pickling of non-tiny functions.

cloudpickle/cloudpickle.py

@@ -42,17 +42,19 @@
 """
 from __future__ import print_function
 
-import operator
-import io
+import dis
+from functools import partial
 import imp
+import io
+import itertools
+import opcode
+import operator
 import pickle
 import struct
 import sys
-import types
-from functools import partial
-import itertools
-import dis
 import traceback
+import types
+import weakref
 
 if sys.version < '3':
     from pickle import Pickler
@@ -68,10 +70,10 @@
     PY3 = True
 
 #relevant opcodes
-STORE_GLOBAL = dis.opname.index('STORE_GLOBAL')
-DELETE_GLOBAL = dis.opname.index('DELETE_GLOBAL')
-LOAD_GLOBAL = dis.opname.index('LOAD_GLOBAL')
-GLOBAL_OPS = [STORE_GLOBAL, DELETE_GLOBAL, LOAD_GLOBAL]
+STORE_GLOBAL = opcode.opmap['STORE_GLOBAL']
+DELETE_GLOBAL = opcode.opmap['DELETE_GLOBAL']
+LOAD_GLOBAL = opcode.opmap['LOAD_GLOBAL']
+GLOBAL_OPS = (STORE_GLOBAL, DELETE_GLOBAL, LOAD_GLOBAL)
 HAVE_ARGUMENT = dis.HAVE_ARGUMENT
 EXTENDED_ARG = dis.EXTENDED_ARG
 
@@ -90,6 +92,46 @@ def _builtin_type(name):
     return getattr(types, name)
 
 
+if sys.version_info < (3, 4):
+    def _walk_global_ops_real(code):
+        code = getattr(code, 'co_code', b'')
+        if not PY3:
+            code = map(ord, code)
+
+        n = len(code)
+        i = 0
+        extended_arg = 0
+        while i < n:
+            op = code[i]
+            i += 1
+            if op >= HAVE_ARGUMENT:
+                oparg = code[i] + code[i + 1] * 256 + extended_arg
+                extended_arg = 0
+                i += 2
+                if op == EXTENDED_ARG:
+                    extended_arg = oparg * 65536
+                if op in GLOBAL_OPS:
+                    yield op, oparg
+
+else:
+    def _walk_global_ops_real(code):
+        for instr in dis.get_instructions(code):
+            op = instr.opcode
+            if op in GLOBAL_OPS:
+                yield op, instr.arg
+
+
+def _walk_global_ops(code, _cache=weakref.WeakKeyDictionary()):
+    """
+    Return a list of (opcode, argument number) tuples for all
+    global-referencing instructions in *code*.
+    """
+    res = _cache.get(code)
+    if res is None:
+        res = _cache[code] = list(_walk_global_ops_real(code))
+    return res
+
+
 class CloudPickler(Pickler):
 
     dispatch = Pickler.dispatch.copy()
@@ -281,41 +323,22 @@ def save_function_tuple(self, func):
         write(pickle.TUPLE)
         write(pickle.REDUCE)  # applies _fill_function on the tuple
 
-    @staticmethod
-    def extract_code_globals(co):
+    @classmethod
+    def extract_code_globals(cls, co):
         """
         Find all globals names read or written to by codeblock co
         """
-
-        code = getattr(co, 'co_code', None)
-        if code is None:
-            return set()
-        if not PY3:
-            code = [ord(c) for c in code]
         names = co.co_names
         out_names = set()
 
-        n = len(code)
-        i = 0
-        extended_arg = 0
-        while i < n:
-            op = code[i]
-
-            i += 1
-            if op >= HAVE_ARGUMENT:
-                oparg = code[i] + code[i+1] * 256 + extended_arg
-                extended_arg = 0
-                i += 2
-                if op == EXTENDED_ARG:
-                    extended_arg = oparg*65536
-                if op in GLOBAL_OPS:
-                    out_names.add(names[oparg])
+        for op, oparg in _walk_global_ops(co):
+            out_names.add(names[oparg])
 
         # see if nested function have any global refs
         if co.co_consts:
             for const in co.co_consts:
                 if type(const) is types.CodeType:
-                    out_names |= CloudPickler.extract_code_globals(const)
+                    out_names |= cls.extract_code_globals(const)
 
         return out_names