Deprecate formatting_values

Author: TomAugspurger

doc/source/whatsnew/v0.24.0.txt

@@ -857,7 +857,7 @@ update the ``ExtensionDtype._metadata`` tuple to match the signature of your
 - :meth:`DataFrame.stack` no longer converts to object dtype for DataFrames where each column has the same extension dtype. The output Series will have the same dtype as the columns (:issue:`23077`).
 - :meth:`Series.unstack` and :meth:`DataFrame.unstack` no longer convert extension arrays to object-dtype ndarrays. Each column in the output ``DataFrame`` will now have the same dtype as the input (:issue:`23077`).
 - Bug when grouping :meth:`Dataframe.groupby()` and aggregating on ``ExtensionArray`` it was not returning the actual ``ExtensionArray`` dtype (:issue:`23227`).
-- A default repr is now provided.
+- A default repr for ExtensionArrays is now provided (:issue:`23601`).
 
 .. _whatsnew_0240.api.incompatibilities:
 
@@ -967,6 +967,7 @@ Deprecations
 - The class ``FrozenNDArray`` has been deprecated. When unpickling, ``FrozenNDArray`` will be unpickled to ``np.ndarray`` once this class is removed (:issue:`9031`)
 - Deprecated the `nthreads` keyword of :func:`pandas.read_feather` in favor of
   `use_threads` to reflect the changes in pyarrow 0.11.0. (:issue:`23053`)
+- :meth:`ExtensionArray._formatting_values` is deprecated. Use `ExtensionArray._formatter` instead. (:issue:`23601`)
 
 .. _whatsnew_0240.deprecations.datetimelike_int_ops:
 

pandas/core/arrays/base.py

@@ -673,20 +673,25 @@ def __repr__(self):
         # the short repr has no trailing newline, while the truncated
         # repr does. So we include a newline in our template, and strip
         # any trailing newlines from format_object_summary
-        data = format_object_summary(self, self._formatter, name=False,
+        data = format_object_summary(self, self._formatter(), name=False,
                                      trailing_comma=False).rstrip('\n')
         name = self.__class__.__name__
         return template.format(class_name=name, data=data,
                                length=len(self),
                                dtype=self.dtype)
 
-    @property
-    def _formatter(self):
-        # type: () -> Callable[Any]
+    def _formatter(self, boxed=False):
+        # type: (bool) -> Callable[Any]
         """Formatting function for scalar values.
 
         This is used in the default '__repr__'. The formatting function
         receives instances of your scalar type.
+
+        Parameters
+        ----------
+        boxed: bool, default False
+            Whether the formatter is to be used by pandas inside a Series
+            or DataFrame repr.
         """
         return str
 

pandas/core/arrays/categorical.py

@@ -2344,9 +2344,6 @@ def _concat_same_type(self, to_concat):
 
         return _concat_categorical(to_concat)
 
-    def _formatting_values(self):
-        return self
-
     def isin(self, values):
         """
         Check whether `values` are contained in Categorical.

pandas/core/arrays/integer.py

@@ -297,10 +297,6 @@ def __iter__(self):
             else:
                 yield self._data[i]
 
-    def _formatting_values(self):
-        # type: () -> np.ndarray
-        return self._coerce_to_ndarray()
-
     def take(self, indexer, allow_fill=False, fill_value=None):
         from pandas.api.extensions import take
 

pandas/core/arrays/interval.py

@@ -689,9 +689,6 @@ def copy(self, deep=False):
         # TODO: Could skip verify_integrity here.
         return type(self).from_arrays(left, right, closed=closed)
 
-    def _formatting_values(self):
-        return np.asarray(self)
-
     def isna(self):
         return isna(self.left)
 

pandas/core/arrays/period.py

@@ -330,6 +330,11 @@ def start_time(self):
     def end_time(self):
         return self.to_timestamp(how='end')
 
+    def _formatter(self, boxed=False):
+        if boxed:
+            return str
+        return "'{}'".format
+
     def __setitem__(
             self,
             key,   # type: Union[int, Sequence[int], Sequence[bool]]

pandas/core/internals/blocks.py

@@ -66,7 +66,7 @@
 import pandas.core.missing as missing
 from pandas.core.base import PandasObject
 
-from pandas.core.arrays import Categorical
+from pandas.core.arrays import Categorical, ExtensionArray
 
 from pandas.core.indexes.datetimes import DatetimeIndex
 from pandas.core.indexes.timedeltas import TimedeltaIndex
@@ -1951,7 +1951,21 @@ def _slice(self, slicer):
         return self.values[slicer]
 
     def formatting_values(self):
-        return self.values._formatting_values()
+        # Deprecating the ability to override _formatting_values.
+        # Do the warning here, it's only user in pandas, since we
+        # have to check if the subclass overrode it.
+        fv = getattr(type(self.values), '_formatting_values', None)
+        if fv is not ExtensionArray._formatting_values:
+            msg = (
+                "'ExtensionArray._formatting_values' is deprecated. "
+                "Specify 'ExtensionArray._formatter' instead."
+            )
+            warnings.warn(msg, FutureWarning)
+            return self.values._formatting_values()
+
+        # the future implementation (and current, if not overrode)
+        formatter = self.values._formatter(boxed=True)
+        return np.array([formatter(x) for x in self.values], dtype=object)
 
     def concat_same_type(self, to_concat, placement=None):
         """

pandas/io/formats/printing.py

@@ -341,19 +341,19 @@ def best_len(values):
             return 0
 
     if trailing_comma:
-        close = ', '
+        close = u', '
     else:
-        close = ''
+        close = u''
 
     if n == 0:
-        summary = '[]{}'.format(close)
+        summary = u'[]{}'.format(close)
     elif n == 1:
         first = formatter(obj[0])
-        summary = '[{}]{}'.format(first, close)
+        summary = u'[{}]{}'.format(first, close)
     elif n == 2:
         first = formatter(obj[0])
         last = formatter(obj[-1])
-        summary = '[{}, {}]{}'.format(first, last, close)
+        summary = u'[{}, {}]{}'.format(first, last, close)
     else:
 
         if n > max_seq_items:

pandas/tests/arrays/test_period.py

@@ -204,9 +204,9 @@ def test_repr_small():
     arr = period_array(['2000', '2001'], freq='D')
     result = str(arr)
     expected = (
-        '<PeriodArray>\n'
-        '[2000-01-01, 2001-01-01]\n'
-        'Length: 2, dtype: period[D]'
+        "<PeriodArray>\n"
+        "['2000-01-01', '2001-01-01']\n"
+        "Length: 2, dtype: period[D]"
     )
     assert result == expected
 

pandas/tests/extension/decimal/test_decimal.py

@@ -188,7 +188,8 @@ def test_value_counts(self, all_data, dropna):
 
 
 class TestCasting(BaseDecimal, base.BaseCastingTests):
-    pass
+    pytestmark = pytest.mark.skipif(compat.PY2,
+                                    reason="Unhashble dtype in Py2.")
 
 
 class TestGroupby(BaseDecimal, base.BaseGroupbyTests):