REF: dataframe formatters/outputs (pandas-dev#36510)

Author: ivanovmg

pandas/_typing.py

@@ -38,6 +38,8 @@
     from pandas.core.indexes.base import Index
     from pandas.core.series import Series
 
+    from pandas.io.formats.format import EngFormatter
+
 # array-like
 
 AnyArrayLike = TypeVar("AnyArrayLike", "ExtensionArray", "Index", "Series", np.ndarray)
@@ -127,6 +129,10 @@
 EncodingVar = TypeVar("EncodingVar", str, None, Optional[str])
 
 
+# type of float formatter in DataFrameFormatter
+FloatFormatType = Union[str, Callable, "EngFormatter"]
+
+
 @dataclass
 class IOargs(Generic[ModeVar, EncodingVar]):
     """

pandas/core/frame.py

@@ -788,10 +788,8 @@ def _repr_html_(self) -> Optional[str]:
                 max_cols=max_cols,
                 show_dimensions=show_dimensions,
                 decimal=".",
-                table_id=None,
-                render_links=False,
             )
-            return formatter.to_html(notebook=True)
+            return fmt.DataFrameRenderer(formatter).to_html(notebook=True)
         else:
             return None
 
@@ -874,9 +872,12 @@ def to_string(
                 max_cols=max_cols,
                 show_dimensions=show_dimensions,
                 decimal=decimal,
+            )
+            return fmt.DataFrameRenderer(formatter).to_string(
+                buf=buf,
+                encoding=encoding,
                 line_width=line_width,
             )
-            return formatter.to_string(buf=buf, encoding=encoding)
 
     # ----------------------------------------------------------------------
 
@@ -2476,29 +2477,29 @@ def to_html(
             columns=columns,
             col_space=col_space,
             na_rep=na_rep,
+            header=header,
+            index=index,
             formatters=formatters,
             float_format=float_format,
+            bold_rows=bold_rows,
             sparsify=sparsify,
             justify=justify,
             index_names=index_names,
-            header=header,
-            index=index,
-            bold_rows=bold_rows,
             escape=escape,
+            decimal=decimal,
             max_rows=max_rows,
             max_cols=max_cols,
             show_dimensions=show_dimensions,
-            decimal=decimal,
-            table_id=table_id,
-            render_links=render_links,
         )
         # TODO: a generic formatter wld b in DataFrameFormatter
-        return formatter.to_html(
+        return fmt.DataFrameRenderer(formatter).to_html(
             buf=buf,
             classes=classes,
             notebook=notebook,
             border=border,
             encoding=encoding,
+            table_id=table_id,
+            render_links=render_links,
         )
 
     # ----------------------------------------------------------------------

pandas/core/generic.py

@@ -4,7 +4,6 @@
 from datetime import timedelta
 import functools
 import gc
-from io import StringIO
 import json
 import operator
 import pickle
@@ -109,7 +108,11 @@
 from pandas.core.window import Expanding, ExponentialMovingWindow, Rolling, Window
 
 from pandas.io.formats import format as fmt
-from pandas.io.formats.format import DataFrameFormatter, format_percentiles
+from pandas.io.formats.format import (
+    DataFrameFormatter,
+    DataFrameRenderer,
+    format_percentiles,
+)
 from pandas.io.formats.printing import pprint_thing
 
 if TYPE_CHECKING:
@@ -3149,7 +3152,7 @@ def to_latex(
             escape=escape,
             decimal=decimal,
         )
-        return formatter.to_latex(
+        return DataFrameRenderer(formatter).to_latex(
             buf=buf,
             column_format=column_format,
             longtable=longtable,
@@ -3182,7 +3185,7 @@ def to_csv(
         date_format: Optional[str] = None,
         doublequote: bool_t = True,
         escapechar: Optional[str] = None,
-        decimal: Optional[str] = ".",
+        decimal: str = ".",
         errors: str = "strict",
         storage_options: StorageOptions = None,
     ) -> Optional[str]:
@@ -3340,39 +3343,33 @@ def to_csv(
         """
         df = self if isinstance(self, ABCDataFrame) else self.to_frame()
 
-        from pandas.io.formats.csvs import CSVFormatter
+        formatter = DataFrameFormatter(
+            frame=df,
+            header=header,
+            index=index,
+            na_rep=na_rep,
+            float_format=float_format,
+            decimal=decimal,
+        )
 
-        formatter = CSVFormatter(
-            df,
+        return DataFrameRenderer(formatter).to_csv(
             path_or_buf,
             line_terminator=line_terminator,
             sep=sep,
             encoding=encoding,
             errors=errors,
             compression=compression,
             quoting=quoting,
-            na_rep=na_rep,
-            float_format=float_format,
-            cols=columns,
-            header=header,
-            index=index,
+            columns=columns,
             index_label=index_label,
             mode=mode,
             chunksize=chunksize,
             quotechar=quotechar,
             date_format=date_format,
             doublequote=doublequote,
             escapechar=escapechar,
-            decimal=decimal,
             storage_options=storage_options,
         )
-        formatter.save()
-
-        if path_or_buf is None:
-            assert isinstance(formatter.path_or_buf, StringIO)
-            return formatter.path_or_buf.getvalue()
-
-        return None
 
     # ----------------------------------------------------------------------
     # Lookup Caching

pandas/io/formats/csvs.py

@@ -5,14 +5,15 @@
 import csv as csvlib
 from io import StringIO, TextIOWrapper
 import os
-from typing import Any, Dict, Hashable, Iterator, List, Optional, Sequence, Union
+from typing import TYPE_CHECKING, Any, Dict, Iterator, List, Optional, Sequence, Union
 
 import numpy as np
 
 from pandas._libs import writers as libwriters
 from pandas._typing import (
     CompressionOptions,
     FilePathOrBuffer,
+    FloatFormatType,
     IndexLabel,
     Label,
     StorageOptions,
@@ -30,18 +31,17 @@
 
 from pandas.io.common import get_filepath_or_buffer, get_handle
 
+if TYPE_CHECKING:
+    from pandas.io.formats.format import DataFrameFormatter
+
 
 class CSVFormatter:
     def __init__(
         self,
-        obj,
+        formatter: "DataFrameFormatter",
         path_or_buf: Optional[FilePathOrBuffer[str]] = None,
         sep: str = ",",
-        na_rep: str = "",
-        float_format: Optional[str] = None,
         cols: Optional[Sequence[Label]] = None,
-        header: Union[bool, Sequence[Hashable]] = True,
-        index: bool = True,
         index_label: Optional[IndexLabel] = None,
         mode: str = "w",
         encoding: Optional[str] = None,
@@ -54,10 +54,11 @@ def __init__(
         date_format: Optional[str] = None,
         doublequote: bool = True,
         escapechar: Optional[str] = None,
-        decimal=".",
         storage_options: StorageOptions = None,
     ):
-        self.obj = obj
+        self.fmt = formatter
+
+        self.obj = self.fmt.frame
 
         self.encoding = encoding or "utf-8"
 
@@ -79,35 +80,45 @@ def __init__(
         self.mode = ioargs.mode
 
         self.sep = sep
-        self.na_rep = na_rep
-        self.float_format = float_format
-        self.decimal = decimal
-        self.header = header
-        self.index = index
-        self.index_label = index_label
+        self.index_label = self._initialize_index_label(index_label)
         self.errors = errors
         self.quoting = quoting or csvlib.QUOTE_MINIMAL
-        self.quotechar = quotechar
+        self.quotechar = self._initialize_quotechar(quotechar)
         self.doublequote = doublequote
         self.escapechar = escapechar
         self.line_terminator = line_terminator or os.linesep
         self.date_format = date_format
-        self.cols = cols  # type: ignore[assignment]
-        self.chunksize = chunksize  # type: ignore[assignment]
+        self.cols = self._initialize_columns(cols)
+        self.chunksize = self._initialize_chunksize(chunksize)
+
+    @property
+    def na_rep(self) -> str:
+        return self.fmt.na_rep
+
+    @property
+    def float_format(self) -> Optional["FloatFormatType"]:
+        return self.fmt.float_format
 
     @property
-    def index_label(self) -> IndexLabel:
-        return self._index_label
+    def decimal(self) -> str:
+        return self.fmt.decimal
 
-    @index_label.setter
-    def index_label(self, index_label: Optional[IndexLabel]) -> None:
+    @property
+    def header(self) -> Union[bool, Sequence[str]]:
+        return self.fmt.header
+
+    @property
+    def index(self) -> bool:
+        return self.fmt.index
+
+    def _initialize_index_label(self, index_label: Optional[IndexLabel]) -> IndexLabel:
         if index_label is not False:
             if index_label is None:
-                index_label = self._get_index_label_from_obj()
+                return self._get_index_label_from_obj()
             elif not isinstance(index_label, (list, tuple, np.ndarray, ABCIndexClass)):
                 # given a string for a DF with Index
-                index_label = [index_label]
-        self._index_label = index_label
+                return [index_label]
+        return index_label
 
     def _get_index_label_from_obj(self) -> List[str]:
         if isinstance(self.obj.index, ABCMultiIndex):
@@ -122,30 +133,17 @@ def _get_index_label_flat(self) -> List[str]:
         index_label = self.obj.index.name
         return [""] if index_label is None else [index_label]
 
-    @property
-    def quotechar(self) -> Optional[str]:
+    def _initialize_quotechar(self, quotechar: Optional[str]) -> Optional[str]:
         if self.quoting != csvlib.QUOTE_NONE:
             # prevents crash in _csv
-            return self._quotechar
+            return quotechar
         return None
 
-    @quotechar.setter
-    def quotechar(self, quotechar: Optional[str]) -> None:
-        self._quotechar = quotechar
-
     @property
     def has_mi_columns(self) -> bool:
         return bool(isinstance(self.obj.columns, ABCMultiIndex))
 
-    @property
-    def cols(self) -> Sequence[Label]:
-        return self._cols
-
-    @cols.setter
-    def cols(self, cols: Optional[Sequence[Label]]) -> None:
-        self._cols = self._refine_cols(cols)
-
-    def _refine_cols(self, cols: Optional[Sequence[Label]]) -> Sequence[Label]:
+    def _initialize_columns(self, cols: Optional[Sequence[Label]]) -> Sequence[Label]:
         # validate mi options
         if self.has_mi_columns:
             if cols is not None:
@@ -161,12 +159,16 @@ def _refine_cols(self, cols: Optional[Sequence[Label]]) -> Sequence[Label]:
 
         # update columns to include possible multiplicity of dupes
         # and make sure sure cols is just a list of labels
-        cols = self.obj.columns
-        if isinstance(cols, ABCIndexClass):
-            return cols._format_native_types(**self._number_format)
+        new_cols = self.obj.columns
+        if isinstance(new_cols, ABCIndexClass):
+            return new_cols._format_native_types(**self._number_format)
         else:
-            assert isinstance(cols, Sequence)
-            return list(cols)
+            return list(new_cols)
+
+    def _initialize_chunksize(self, chunksize: Optional[int]) -> int:
+        if chunksize is None:
+            return (100000 // (len(self.cols) or 1)) or 1
+        return int(chunksize)
 
     @property
     def _number_format(self) -> Dict[str, Any]:
@@ -179,17 +181,6 @@ def _number_format(self) -> Dict[str, Any]:
             decimal=self.decimal,
         )
 
-    @property
-    def chunksize(self) -> int:
-        return self._chunksize
-
-    @chunksize.setter
-    def chunksize(self, chunksize: Optional[int]) -> None:
-        if chunksize is None:
-            chunksize = (100000 // (len(self.cols) or 1)) or 1
-        assert chunksize is not None
-        self._chunksize = int(chunksize)
-
     @property
     def data_index(self) -> Index:
         data_index = self.obj.index