Squashed commit of the following:

commit b29dfc6
Author: Tom Augspurger <tom.w.augspurger@gmail.com>
Date:   Thu Aug 16 10:45:38 2018 -0500

    Support NDFrame.shift with EAs

    Uses take internally.

    Closes pandas-dev#22386

commit b5d81cf
Author: William Ayd <william.ayd@icloud.com>
Date:   Thu Aug 16 03:54:18 2018 -0700

    Bump pytest (pandas-dev#22320)

commit f07a790
Author: jbrockmendel <jbrockmendel@gmail.com>
Date:   Thu Aug 16 03:46:58 2018 -0700

    Make more of numpy_helper unnecessary (pandas-dev#22344)

commit 7b80d4d
Author: Graham Inggs <graham.inggs+github@gmail.com>
Date:   Thu Aug 16 12:43:02 2018 +0200

    Drop redundant TestLocale (pandas-dev#22349)

commit 6bcfc46
Author: Matthew Roeschke <emailformattr@gmail.com>
Date:   Thu Aug 16 03:32:31 2018 -0700

    Fix failing dateutil test (pandas-dev#22354)

commit 86e8f23
Author: jbrockmendel <jbrockmendel@gmail.com>
Date:   Thu Aug 16 03:08:09 2018 -0700

    remove last cython: nprofile comments (pandas-dev#22371)

commit 70e6f7c
Author: Joris Van den Bossche <jorisvandenbossche@gmail.com>
Date:   Wed Aug 15 18:09:50 2018 +0200

    DOC: edit docstring example to prevent segfault (pandas-dev#21824) (pandas-dev#22368)

ci/environment-dev.yaml

@@ -8,7 +8,7 @@ dependencies:
   - flake8
   - flake8-comprehensions
   - moto
-  - pytest>=3.1
+  - pytest>=3.6
   - python-dateutil>=2.5.0
   - python=3
   - pytz

ci/requirements_dev.txt

@@ -5,7 +5,7 @@ NumPy
 flake8
 flake8-comprehensions
 moto
-pytest>=3.1
+pytest>=3.6
 python-dateutil>=2.5.0
 pytz
 setuptools>=24.2.0

doc/source/contributing.rst

@@ -652,13 +652,13 @@ Adding tests is one of the most common requests after code is pushed to *pandas*
 it is worth getting in the habit of writing tests ahead of time so this is never an issue.
 
 Like many packages, *pandas* uses `pytest
-<http://doc.pytest.org/en/latest/>`_ and the convenient
+<http://docs.pytest.org/en/latest/>`_ and the convenient
 extensions in `numpy.testing
 <http://docs.scipy.org/doc/numpy/reference/routines.testing.html>`_.
 
 .. note::
 
-   The earliest supported pytest version is 3.1.0.
+   The earliest supported pytest version is 3.6.0.
 
 Writing tests
 ~~~~~~~~~~~~~
@@ -702,7 +702,7 @@ Transitioning to ``pytest``
     class TestReallyCoolFeature(object):
         ....
 
-Going forward, we are moving to a more *functional* style using the `pytest <http://doc.pytest.org/en/latest/>`__ framework, which offers a richer testing
+Going forward, we are moving to a more *functional* style using the `pytest <http://docs.pytest.org/en/latest/>`__ framework, which offers a richer testing
 framework that will facilitate testing and developing. Thus, instead of writing test classes, we will write test functions like this:
 
 .. code-block:: python
@@ -766,7 +766,7 @@ A test run of this yields
 
    ((pandas) bash-3.2$ pytest  test_cool_feature.py  -v
    =========================== test session starts ===========================
-   platform darwin -- Python 3.6.2, pytest-3.2.1, py-1.4.31, pluggy-0.4.0
+   platform darwin -- Python 3.6.2, pytest-3.6.0, py-1.4.31, pluggy-0.4.0
    collected 11 items
 
    tester.py::test_dtypes[int8] PASSED
@@ -788,7 +788,7 @@ Tests that we have ``parametrized`` are now accessible via the test name, for ex
 
    ((pandas) bash-3.2$ pytest  test_cool_feature.py  -v -k int8
    =========================== test session starts ===========================
-   platform darwin -- Python 3.6.2, pytest-3.2.1, py-1.4.31, pluggy-0.4.0
+   platform darwin -- Python 3.6.2, pytest-3.6.0, py-1.4.31, pluggy-0.4.0
    collected 11 items
 
    test_cool_feature.py::test_dtypes[int8] PASSED
@@ -837,7 +837,7 @@ On Windows, one can type::
 This can significantly reduce the time it takes to locally run tests before
 submitting a pull request.
 
-For more, see the `pytest <http://doc.pytest.org/en/latest/>`_ documentation.
+For more, see the `pytest <http://docs.pytest.org/en/latest/>`_ documentation.
 
     .. versionadded:: 0.20.0
 

doc/source/install.rst

@@ -202,15 +202,15 @@ pandas is equipped with an exhaustive set of unit tests, covering about 97% of
 the code base as of this writing. To run it on your machine to verify that
 everything is working (and that you have all of the dependencies, soft and hard,
 installed), make sure you have `pytest
-<http://doc.pytest.org/en/latest/>`__ and run:
+<http://docs.pytest.org/en/latest/>`__ >= 3.6 and run:
 
 ::
 
     >>> import pandas as pd
     >>> pd.test()
     running: pytest --skip-slow --skip-network C:\Users\TP\Anaconda3\envs\py36\lib\site-packages\pandas
     ============================= test session starts =============================
-    platform win32 -- Python 3.6.2, pytest-3.2.1, py-1.4.34, pluggy-0.4.0
+    platform win32 -- Python 3.6.2, pytest-3.6.0, py-1.4.34, pluggy-0.4.0
     rootdir: C:\Users\TP\Documents\Python\pandasdev\pandas, inifile: setup.cfg
     collected 12145 items / 3 skipped
 

doc/source/whatsnew/v0.23.5.txt

@@ -26,6 +26,11 @@ Fixed Regressions
 -
 -
 
+
+Development
+~~~~~~~~~~~
+- The minimum required pytest version has been increased to 3.6 (:issue:`22319`)
+
 .. _whatsnew_0235.bug_fixes:
 
 Bug Fixes

doc/source/whatsnew/v0.24.0.txt

@@ -462,6 +462,7 @@ ExtensionType Changes
 - Added ``ExtensionDtype._is_numeric`` for controlling whether an extension dtype is considered numeric (:issue:`22290`).
 - The ``ExtensionArray`` constructor, ``_from_sequence`` now take the keyword arg ``copy=False`` (:issue:`21185`)
 - Bug in :meth:`Series.get` for ``Series`` using ``ExtensionArray`` and integer index (:issue:`21257`)
+- :meth:`~Series.shift` now works with extension arrays, rather than raising an AttributeError (:isseu:`22386`)
 - :meth:`Series.combine()` works correctly with :class:`~pandas.api.extensions.ExtensionArray` inside of :class:`Series` (:issue:`20825`)
 - :meth:`Series.combine()` with scalar argument now works for any function type (:issue:`21248`)
 - Added ``ExtensionDtype._is_numeric`` for controlling whether an extension dtype is considered numeric.

pandas/_libs/algos.pyx

@@ -1,4 +1,4 @@
-# cython: profile=False
+# -*- coding: utf-8 -*-
 
 cimport cython
 from cython cimport Py_ssize_t

pandas/_libs/groupby.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 cimport cython
 from cython cimport Py_ssize_t

pandas/_libs/hashing.pyx

@@ -1,4 +1,4 @@
-# cython: profile=False
+# -*- coding: utf-8 -*-
 # Translated from the reference implementation
 # at https://github.com/veorq/SipHash
 

pandas/_libs/hashtable.pyx

@@ -1,4 +1,4 @@
-# cython: profile=False
+# -*- coding: utf-8 -*-
 
 cimport cython
 

pandas/_libs/index.pyx

@@ -1,4 +1,4 @@
-# cython: profile=False
+# -*- coding: utf-8 -*-
 from datetime import datetime, timedelta, date
 
 cimport cython
@@ -319,15 +319,14 @@ cdef class IndexEngine:
         # form the set of the results (like ismember)
         members = np.empty(n, dtype=np.uint8)
         for i in range(n):
-            val = util.get_value_1d(values, i)
+            val = values[i]
             if val in stargets:
                 if val not in d:
                     d[val] = []
                 d[val].append(i)
 
         for i in range(n_t):
-
-            val = util.get_value_1d(targets, i)
+            val = targets[i]
 
             # found
             if val in d:

pandas/_libs/indexing.pyx

@@ -1,4 +1,5 @@
-# cython: profile=False
+# -*- coding: utf-8 -*-
+
 
 cdef class _NDFrameIndexerBase:
     """

pandas/_libs/join.pyx

@@ -1,4 +1,4 @@
-# cython: profile=False
+# -*- coding: utf-8 -*-
 
 cimport cython
 from cython cimport Py_ssize_t

pandas/_libs/khash.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 from cpython cimport PyObject
 from numpy cimport int64_t, uint64_t, int32_t, uint32_t, float64_t
 

pandas/_libs/lib.pyx

@@ -1153,7 +1153,7 @@ def infer_dtype(object value, bint skipna=False):
 
     # try to use a valid value
     for i in range(n):
-        val = util.get_value_1d(values, i)
+        val = values[i]
 
         # do not use is_nul_datetimelike to keep
         # np.datetime64('nat') and np.timedelta64('nat')
@@ -1240,7 +1240,7 @@ def infer_dtype(object value, bint skipna=False):
             return 'interval'
 
     for i in range(n):
-        val = util.get_value_1d(values, i)
+        val = values[i]
         if (util.is_integer_object(val) and
                 not util.is_timedelta64_object(val) and
                 not util.is_datetime64_object(val)):
@@ -2255,7 +2255,7 @@ def fast_multiget(dict mapping, ndarray keys, default=np.nan):
     keys = getattr(keys, 'values', keys)
 
     for i in range(n):
-        val = util.get_value_1d(keys, i)
+        val = keys[i]
         if val in mapping:
             output[i] = mapping[val]
         else:

pandas/_libs/missing.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from tslibs.nattype cimport is_null_datetimelike
 

pandas/_libs/missing.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cpython cimport PyFloat_Check, PyComplex_Check
 

pandas/_libs/reduction.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 from distutils.version import LooseVersion
 
 from cython cimport Py_ssize_t

pandas/_libs/reshape.pyx

@@ -1,4 +1,4 @@
-# cython: profile=False
+# -*- coding: utf-8 -*-
 
 cimport cython
 from cython cimport Py_ssize_t

pandas/_libs/skiplist.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cython cimport Py_ssize_t
 

pandas/_libs/src/numpy_helper.h

@@ -28,20 +28,4 @@ PANDAS_INLINE PyObject* get_value_1d(PyArrayObject* ap, Py_ssize_t i) {
     return PyArray_Scalar(item, PyArray_DESCR(ap), (PyObject*)ap);
 }
 
-// returns ASCII or UTF8 (py3) view on python str
-// python object owns memory, should not be freed
-PANDAS_INLINE const char* get_c_string(PyObject* obj) {
-#if PY_VERSION_HEX >= 0x03000000
-    return PyUnicode_AsUTF8(obj);
-#else
-    return PyString_AsString(obj);
-#endif
-}
-
-void set_array_not_contiguous(PyArrayObject* ao) {
-    // Numpy>=1.8-compliant equivalent to:
-    //  ao->flags &= ~(NPY_ARRAY_C_CONTIGUOUS | NPY_ARRAY_F_CONTIGUOUS);
-    PyArray_CLEARFLAGS(ao, (NPY_ARRAY_C_CONTIGUOUS | NPY_ARRAY_F_CONTIGUOUS));
-}
-
 #endif  // PANDAS__LIBS_SRC_NUMPY_HELPER_H_

pandas/_libs/tslib.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 from cython cimport Py_ssize_t
 
 from cpython cimport PyFloat_Check, PyUnicode_Check

pandas/_libs/tslibs/ccalendar.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cython cimport Py_ssize_t
 

pandas/_libs/tslibs/ccalendar.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 # cython: boundscheck=False
 """
 Cython implementations of functions resembling the stdlib calendar module

pandas/_libs/tslibs/conversion.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cpython.datetime cimport datetime, tzinfo
 

pandas/_libs/tslibs/conversion.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 cimport cython
 from cython cimport Py_ssize_t

pandas/_libs/tslibs/fields.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 """
 Functions for accessing attributes of Timestamp/datetime64/datetime-like
 objects and arrays

pandas/_libs/tslibs/frequencies.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 cpdef object get_rule_month(object source, object default=*)
 

pandas/_libs/tslibs/nattype.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from numpy cimport int64_t
 cdef int64_t NPY_NAT

pandas/_libs/tslibs/nattype.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cpython cimport (
     PyFloat_Check, PyComplex_Check,

pandas/_libs/tslibs/np_datetime.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cpython.datetime cimport date, datetime
 

pandas/_libs/tslibs/np_datetime.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cpython cimport (Py_EQ, Py_NE, Py_GE, Py_GT, Py_LT, Py_LE,
                       PyUnicode_Check, PyUnicode_AsASCIIString)

pandas/_libs/tslibs/timedeltas.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from numpy cimport int64_t
 

pandas/_libs/tslibs/timedeltas.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 import collections
 import textwrap
 import warnings

pandas/_libs/tslibs/timestamps.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from numpy cimport int64_t
 from np_datetime cimport npy_datetimestruct

pandas/_libs/tslibs/timestamps.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 import warnings
 
 from cpython cimport (PyObject_RichCompareBool, PyObject_RichCompare,

pandas/_libs/tslibs/timezones.pxd

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 cdef bint is_utc(object tz)
 cdef bint is_tzlocal(object tz)

pandas/_libs/tslibs/timezones.pyx

@@ -1,5 +1,4 @@
 # -*- coding: utf-8 -*-
-# cython: profile=False
 
 from cython cimport Py_ssize_t