[WIP] style NA in reprs

LICENSES/COLORAMA_LICENSE

@@ -0,0 +1 @@
+

pandas/core/_formats.py

@@ -0,0 +1,67 @@
+"""
+Constants for formatting. Usable by both pandas.core and elsewhere.
+
+The names are chosen to match colorama/ansi.py, whose license is included
+in LICENSES/COLORAMA_LICENSE
+"""
+import re
+from typing import Any, Callable, List, Optional
+
+from pandas._libs import missing as libmissing
+
+CSI = "\033["
+ANSI_PAT = re.compile(r"\x1B[@-_][0-?]*[ -/]*[@-~]")
+
+
+def strip_ansi(x):
+    return ANSI_PAT.sub("", x)
+
+
+def format_with(value: str, formatters: List[str]):
+    return "".join(formatters + [value, AnsiStyle.RESET_ALL])
+
+
+class AnsiFore:
+    BLACK = f"{CSI}30m"
+    RED = f"{CSI}31m"
+    GREEN = f"{CSI}32m"
+    YELLOW = f"{CSI}33m"
+    BLUE = f"{CSI}34m"
+    MAGENTA = f"{CSI}35m"
+    CYAN = f"{CSI}36m"
+    WHITE = f"{CSI}37m"
+    RESET = f"{CSI}39m"
+
+
+class AnsiBack:
+    BLACK = f"{CSI}40m"
+    RED = f"{CSI}41m"
+    GREEN = f"{CSI}42m"
+    YELLOW = f"{CSI}43m"
+    BLUE = f"{CSI}44m"
+    MAGENTA = f"{CSI}45m"
+    CYAN = f"{CSI}46m"
+    WHITE = f"{CSI}47m"
+    RESET = f"{CSI}49m"
+
+
+class AnsiStyle:
+    BRIGHT = f"{CSI}1m"
+    DIM = f"{CSI}2m"
+    NORMAL = f"{CSI}22m"
+    RESET_ALL = f"{CSI}0m"
+
+
+class NAFormatterMixin:
+    def _formatter(
+        self, boxed: bool = False, terminal=False
+    ) -> Callable[[Any], Optional[str]]:
+        def formatter(x):
+            if x is libmissing.NA and terminal:
+                return format_with("NA", [AnsiFore.RED])
+            elif boxed:
+                return str(x)
+            else:
+                return repr(x)
+
+        return formatter

pandas/core/arrays/base.py

@@ -5,8 +5,10 @@
    This is an experimental API and subject to breaking changes
    without warning.
 """
+import inspect
 import operator
 from typing import Any, Callable, Dict, Optional, Sequence, Tuple, Union
+import warnings
 
 import numpy as np
 
@@ -910,20 +912,35 @@ def view(self, dtype=None) -> Union[ABCExtensionArray, np.ndarray]:
     # ------------------------------------------------------------------------
     # Printing
     # ------------------------------------------------------------------------
-
-    def __repr__(self) -> str:
+    def _repr_base(self, is_repr=True):
         from pandas.io.formats.printing import format_object_summary
 
         # 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
+
+        if is_repr:
+            terminal = True
+        else:
+            terminal = False
+        # compatibility for older EAs
+        kwargs = _check_formatter_signature(self._formatter, terminal=terminal)
+
         data = format_object_summary(
-            self, self._formatter(), indent_for_name=False
+            self, self._formatter(**kwargs), indent_for_name=False
         ).rstrip(", \n")
         class_name = f"<{type(self).__name__}>\n"
         return f"{class_name}{data}\nLength: {len(self)}, dtype: {self.dtype}"
 
-    def _formatter(self, boxed: bool = False) -> Callable[[Any], Optional[str]]:
+    def __repr__(self) -> str:
+        return self._repr_base(is_repr=True)
+
+    def __str__(self):
+        return self._repr_base(is_repr=False)
+
+    def _formatter(
+        self, boxed: bool = False, terminal: bool = False
+    ) -> Callable[[Any], Optional[str]]:
         """Formatting function for scalar values.
 
         This is used in the default '__repr__'. The returned formatting
@@ -937,6 +954,12 @@ def _formatter(self, boxed: bool = False) -> Callable[[Any], Optional[str]]:
             itself (False). This may be useful if you want scalar values
             to appear differently within a Series versus on its own (e.g.
             quoted or not).
+        terminal : bool, default False
+            Indicator whether the result is being printed to a terminal
+            screen. This may be used to detect whether ANSI codes should
+            be used to style terminal output.
+
+            .. versionadded:: 1.0.0
 
         Returns
         -------
@@ -1203,3 +1226,13 @@ def _create_arithmetic_method(cls, op):
     @classmethod
     def _create_comparison_method(cls, op):
         return cls._create_method(op, coerce_to_dtype=False)
+
+
+def _check_formatter_signature(formatter, **kwargs):
+    if len(inspect.signature(formatter).parameters) == 1:
+        warnings.warn(
+            "'_formatter' signature is incorrect. Ensure it matches "
+            "the base class' signature."
+        )
+        kwargs.pop("terminal", None)
+    return kwargs

pandas/core/arrays/boolean.py

@@ -27,6 +27,7 @@
 from pandas.core.dtypes.missing import isna, notna
 
 from pandas.core import nanops, ops
+from pandas.core._formats import NAFormatterMixin
 from pandas.core.algorithms import take
 from pandas.core.arrays import ExtensionArray, ExtensionOpsMixin
 import pandas.core.common as com
@@ -197,7 +198,7 @@ def coerce_to_array(values, mask=None, copy: bool = False):
     return values, mask
 
 
-class BooleanArray(ExtensionArray, ExtensionOpsMixin):
+class BooleanArray(NAFormatterMixin, ExtensionArray, ExtensionOpsMixin):
     """
     Array of boolean (True/False) data with missing values.
 
@@ -295,9 +296,6 @@ def _values_for_factorize(self) -> Tuple[np.ndarray, Any]:
     def _from_factorized(cls, values, original: "BooleanArray"):
         return cls._from_sequence(values, dtype=original.dtype)
 
-    def _formatter(self, boxed=False):
-        return str
-
     @property
     def _hasna(self) -> bool:
         # Note: this is expensive right now! The hope is that we can

pandas/core/arrays/categorical.py

@@ -459,7 +459,7 @@ def _constructor(self) -> Type["Categorical"]:
     def _from_sequence(cls, scalars, dtype=None, copy=False):
         return Categorical(scalars, dtype=dtype)
 
-    def _formatter(self, boxed=False):
+    def _formatter(self, boxed=False, terminal=False):
         # Defer to CategoricalFormatter's formatter.
         return None
 

pandas/core/arrays/datetimelike.py

@@ -389,7 +389,7 @@ def _format_native_types(self, na_rep="NaT", date_format=None):
         """
         raise AbstractMethodError(self)
 
-    def _formatter(self, boxed=False):
+    def _formatter(self, boxed=False, terminal=False):
         # TODO: Remove Datetime & DatetimeTZ formatters.
         return "'{}'".format
 

pandas/core/arrays/integer.py

@@ -24,6 +24,7 @@
 from pandas.core.dtypes.missing import isna, notna
 
 from pandas.core import nanops, ops
+from pandas.core._formats import NAFormatterMixin
 from pandas.core.algorithms import take
 from pandas.core.arrays import ExtensionArray, ExtensionOpsMixin
 import pandas.core.common as com
@@ -259,7 +260,7 @@ def coerce_to_array(values, dtype, mask=None, copy=False):
     return values, mask
 
 
-class IntegerArray(ExtensionArray, ExtensionOpsMixin):
+class IntegerArray(NAFormatterMixin, ExtensionArray, ExtensionOpsMixin):
     """
     Array of integer (optional missing) values.
 

pandas/core/arrays/period.py

@@ -615,7 +615,7 @@ def asfreq(self, freq=None, how="E"):
     # ------------------------------------------------------------------
     # Rendering Methods
 
-    def _formatter(self, boxed=False):
+    def _formatter(self, boxed=False, terminal=False):
         if boxed:
             return str
         return "'{}'".format

pandas/core/arrays/sparse/array.py

@@ -1471,7 +1471,7 @@ def __repr__(self) -> str:
         pp_index = printing.pprint_thing(self.sp_index)
         return f"{pp_str}\nFill: {pp_fill}\n{pp_index}"
 
-    def _formatter(self, boxed=False):
+    def _formatter(self, boxed=False, terminal=False):
         # Defer to the formatter from the GenericArrayFormatter calling us.
         # This will infer the correct formatter from the dtype of the values.
         return None

pandas/core/arrays/string_.py

@@ -13,6 +13,7 @@
 
 from pandas import compat
 from pandas.core import ops
+from pandas.core._formats import NAFormatterMixin
 from pandas.core.arrays import PandasArray
 from pandas.core.construction import extract_array
 from pandas.core.missing import isna
@@ -82,7 +83,7 @@ def __from_arrow__(self, array):
         return StringArray._concat_same_type(results)
 
 
-class StringArray(PandasArray):
+class StringArray(NAFormatterMixin, PandasArray):
     """
     Extension array for string data.
 

pandas/core/arrays/timedeltas.py

@@ -457,7 +457,7 @@ def median(
     # ----------------------------------------------------------------
     # Rendering Methods
 
-    def _formatter(self, boxed=False):
+    def _formatter(self, boxed=False, terminal=False):
         from pandas.io.formats.format import _get_format_timedelta64
 
         return _get_format_timedelta64(self, box=True)

pandas/core/frame.py

@@ -655,7 +655,7 @@ def _info_repr(self) -> bool:
             self._repr_fits_horizontal_() and self._repr_fits_vertical_()
         )
 
-    def __repr__(self) -> str:
+    def _repr_base(self, is_repr: bool = True) -> str:
         """
         Return a string representation for a particular DataFrame.
         """
@@ -681,10 +681,17 @@ def __repr__(self) -> str:
             line_width=width,
             max_colwidth=max_colwidth,
             show_dimensions=show_dimensions,
+            terminal=is_repr,
         )
 
         return buf.getvalue()
 
+    def __repr__(self) -> str:
+        return self._repr_base(is_repr=True)
+
+    def __str__(self) -> str:
+        return self._repr_base(is_repr=False)
+
     def _repr_html_(self) -> Optional[str]:
         """
         Return a html representation for a particular DataFrame.
@@ -761,6 +768,7 @@ def to_string(
         line_width: Optional[int] = None,
         max_colwidth: Optional[int] = None,
         encoding: Optional[str] = None,
+        terminal: bool = False,
     ) -> Optional[str]:
         """
         Render a DataFrame to a console-friendly tabular output.
@@ -812,6 +820,7 @@ def to_string(
                 show_dimensions=show_dimensions,
                 decimal=decimal,
                 line_width=line_width,
+                terminal=terminal,
             )
             return formatter.to_string(buf=buf, encoding=encoding)
 

pandas/core/series.py

@@ -1292,8 +1292,7 @@ def reset_index(self, level=None, drop=False, name=None, inplace=False):
 
     # ----------------------------------------------------------------------
     # Rendering Methods
-
-    def __repr__(self) -> str:
+    def _repr_base(self, is_repr: bool = True):
         """
         Return a string representation for a particular Series.
         """
@@ -1318,11 +1317,18 @@ def __repr__(self) -> str:
             min_rows=min_rows,
             max_rows=max_rows,
             length=show_dimensions,
+            terminal=is_repr,
         )
         result = buf.getvalue()
 
         return result
 
+    def __repr__(self) -> str:
+        return self._repr_base(is_repr=True)
+
+    def __str__(self) -> str:
+        return self._repr_base(is_repr=False)
+
     def to_string(
         self,
         buf=None,
@@ -1335,6 +1341,7 @@ def to_string(
         name=False,
         max_rows=None,
         min_rows=None,
+        terminal: bool = False,
     ):
         """
         Render a string representation of the Series.
@@ -1382,6 +1389,7 @@ def to_string(
             float_format=float_format,
             min_rows=min_rows,
             max_rows=max_rows,
+            terminal=terminal,
         )
         result = formatter.to_string()