CLN: PY3 Remove cPickle (#25956)

* CLN: PY3 cPickle

* add back pickle_compat import

* Move pickle imports to top

* Clarify pickle read operation

Author: mroeschke

pandas/compat/__init__.py

@@ -44,13 +44,11 @@
     # always writeable
     from StringIO import StringIO
     BytesIO = StringIO
-    import cPickle
     import httplib
 except ImportError:
     import builtins
     from io import StringIO, BytesIO
     cStringIO = StringIO
-    import pickle as cPickle
     import http.client as httplib
 
 from pandas.compat.chainmap import DeepChainMap

pandas/compat/pickle_compat.py

@@ -7,7 +7,7 @@
 import sys
 
 import pandas  # noqa
-from pandas import Index, compat
+from pandas import Index
 
 
 def load_reduce(self):
@@ -138,27 +138,14 @@ def load_reduce(self):
 # our Unpickler sub-class to override methods and some dispatcher
 # functions for compat
 
-if compat.PY3:
-    class Unpickler(pkl._Unpickler):
+class Unpickler(pkl._Unpickler):
 
-        def find_class(self, module, name):
-            # override superclass
-            key = (module, name)
-            module, name = _class_locations_map.get(key, key)
-            return super(Unpickler, self).find_class(module, name)
+    def find_class(self, module, name):
+        # override superclass
+        key = (module, name)
+        module, name = _class_locations_map.get(key, key)
+        return super(Unpickler, self).find_class(module, name)
 
-else:
-
-    class Unpickler(pkl.Unpickler):
-
-        def find_class(self, module, name):
-            # override superclass
-            key = (module, name)
-            module, name = _class_locations_map.get(key, key)
-            __import__(module)
-            mod = sys.modules[module]
-            klass = getattr(mod, name)
-            return klass
 
 Unpickler.dispatch = copy.copy(Unpickler.dispatch)
 Unpickler.dispatch[pkl.REDUCE[0]] = load_reduce

pandas/core/generic.py

@@ -5,6 +5,7 @@
 import gc
 import json
 import operator
+import pickle
 from textwrap import dedent
 import warnings
 import weakref
@@ -15,8 +16,7 @@
 
 from pandas._libs import Timestamp, iNaT, properties
 import pandas.compat as compat
-from pandas.compat import (
-    cPickle as pkl, isidentifier, lrange, lzip, set_function_name, to_str)
+from pandas.compat import isidentifier, lrange, lzip, set_function_name, to_str
 from pandas.compat.numpy import function as nv
 from pandas.errors import AbstractMethodError
 from pandas.util._decorators import (
@@ -2564,7 +2564,7 @@ def to_sql(self, name, con, schema=None, if_exists='fail', index=True,
                    dtype=dtype, method=method)
 
     def to_pickle(self, path, compression='infer',
-                  protocol=pkl.HIGHEST_PROTOCOL):
+                  protocol=pickle.HIGHEST_PROTOCOL):
         """
         Pickle (serialize) object to file.
 

pandas/io/pickle.py

@@ -1,14 +1,16 @@
 """ pickle compat """
+import pickle
 import warnings
 
 from numpy.lib.format import read_array
 
-from pandas.compat import BytesIO, cPickle as pkl, pickle_compat as pc
+from pandas.compat import BytesIO, pickle_compat as pc
 
 from pandas.io.common import _get_handle, _stringify_path
 
 
-def to_pickle(obj, path, compression='infer', protocol=pkl.HIGHEST_PROTOCOL):
+def to_pickle(obj, path, compression='infer',
+              protocol=pickle.HIGHEST_PROTOCOL):
     """
     Pickle (serialize) object to file.
 
@@ -71,9 +73,9 @@ def to_pickle(obj, path, compression='infer', protocol=pkl.HIGHEST_PROTOCOL):
                         compression=compression,
                         is_text=False)
     if protocol < 0:
-        protocol = pkl.HIGHEST_PROTOCOL
+        protocol = pickle.HIGHEST_PROTOCOL
     try:
-        f.write(pkl.dumps(obj, protocol=protocol))
+        f.write(pickle.dumps(obj, protocol=protocol))
     finally:
         f.close()
         for _f in fh:
@@ -140,15 +142,15 @@ def read_pickle(path, compression='infer'):
     path = _stringify_path(path)
     f, fh = _get_handle(path, 'rb', compression=compression, is_text=False)
 
-    # 1) try with cPickle
-    # 2) try with the compat pickle to handle subclass changes
-    # 3) pass encoding only if its not None as py2 doesn't handle the param
+    # 1) try standard libary Pickle
+    # 2) try pickle_compat (older pandas version) to handle subclass changes
+    # 3) try pickle_compat with latin1 encoding
 
     try:
         with warnings.catch_warnings(record=True):
             # We want to silence any warnings about, e.g. moved modules.
             warnings.simplefilter("ignore", Warning)
-            return pkl.load(f)
+            return pickle.load(f)
     except Exception:  # noqa: E722
         try:
             return pc.load(f, encoding=None)

pandas/tests/io/generate_legacy_storage_files.py

@@ -37,6 +37,7 @@
 from datetime import timedelta
 from distutils.version import LooseVersion
 import os
+import pickle
 import platform as pl
 import sys
 
@@ -285,11 +286,6 @@ def platform_name():
 def write_legacy_pickles(output_dir):
 
     # make sure we are < 0.13 compat (in py3)
-    try:
-        from pandas.compat import cPickle as pickle  # noqa
-    except ImportError:
-        import pickle
-
     version = pandas.__version__
 
     print("This script generates a storage file for the current arch, system, "

pandas/tests/io/test_pickle.py

@@ -18,6 +18,7 @@
 import gzip
 import lzma
 import os
+import pickle
 import shutil
 from warnings import catch_warnings, simplefilter
 import zipfile
@@ -217,37 +218,20 @@ def test_pickles(current_pickle_data, legacy_pickle):
 
 def test_round_trip_current(current_pickle_data):
 
-    try:
-        import cPickle as c_pickle
-
-        def c_pickler(obj, path):
-            with open(path, 'wb') as fh:
-                c_pickle.dump(obj, fh, protocol=-1)
-
-        def c_unpickler(path):
-            with open(path, 'rb') as fh:
-                fh.seek(0)
-                return c_pickle.load(fh)
-    except ImportError:
-        c_pickler = None
-        c_unpickler = None
-
-    import pickle as python_pickle
-
     def python_pickler(obj, path):
         with open(path, 'wb') as fh:
-            python_pickle.dump(obj, fh, protocol=-1)
+            pickle.dump(obj, fh, protocol=-1)
 
     def python_unpickler(path):
         with open(path, 'rb') as fh:
             fh.seek(0)
-            return python_pickle.load(fh)
+            return pickle.load(fh)
 
     data = current_pickle_data
     for typ, dv in data.items():
         for dt, expected in dv.items():
 
-            for writer in [pd.to_pickle, c_pickler, python_pickler]:
+            for writer in [pd.to_pickle, python_pickler]:
                 if writer is None:
                     continue
 
@@ -260,10 +244,6 @@ def python_unpickler(path):
                     result = pd.read_pickle(path)
                     compare_element(result, expected, typ)
 
-                    if c_unpickler is not None:
-                        result = c_unpickler(path)
-                        compare_element(result, expected, typ)
-
                     result = python_unpickler(path)
                     compare_element(result, expected, typ)