CLN: remove py2-legacy UnicodeReader, UnicodeWriter

pandas/io/common.py

@@ -2,7 +2,6 @@
 
 import bz2
 import codecs
-import csv
 import gzip
 from io import BufferedIOBase, BytesIO
 import mmap
@@ -17,9 +16,7 @@
     List,
     Mapping,
     Optional,
-    TextIO,
     Tuple,
-    Type,
     Union,
 )
 from urllib.parse import (  # noqa
@@ -597,16 +594,3 @@ def next(self) -> bytes:
 
     def close(self):
         self.reader.close()
-
-
-# Keeping these class for now because it provides a necessary convenience
-# for "dropping" the "encoding" argument from our I/O arguments when
-# creating a Unicode I/O object.
-def UnicodeReader(f, dialect=csv.excel, encoding="utf-8", **kwds):
-    return csv.reader(f, dialect=dialect, **kwds)
-
-
-def UnicodeWriter(
-    f: TextIO, dialect: Type[csv.Dialect] = csv.excel, encoding: str = "utf-8", **kwds
-):
-    return csv.writer(f, dialect=dialect, **kwds)

pandas/io/formats/csvs.py

@@ -22,7 +22,6 @@
 from pandas.core.dtypes.missing import notna
 
 from pandas.io.common import (
-    UnicodeWriter,
     _get_compression_method,
     _get_handle,
     _infer_compression,
@@ -196,10 +195,8 @@ def save(self):
                 escapechar=self.escapechar,
                 quotechar=self.quotechar,
             )
-            if self.encoding == "ascii":
-                self.writer = csvlib.writer(f, **writer_kwargs)
-            else:
-                self.writer = UnicodeWriter(f, encoding=self.encoding, **writer_kwargs)
+            # Note: self.encoding is irrelevant here
+            self.writer = csvlib.writer(f, **writer_kwargs)
 
             self._save()
 

pandas/io/parsers.py

@@ -34,6 +34,7 @@
     is_categorical_dtype,
     is_dtype_equal,
     is_extension_array_dtype,
+    is_file_like,
     is_float,
     is_integer,
     is_integer_dtype,
@@ -62,13 +63,11 @@
 from pandas.io.common import (
     _NA_VALUES,
     BaseIterator,
-    UnicodeReader,
     UTF8Recoder,
     _get_handle,
     _infer_compression,
     _validate_header_arg,
     get_filepath_or_buffer,
-    is_file_like,
 )
 from pandas.io.date_converters import generic_parser
 
@@ -2431,23 +2430,13 @@ class MyDialect(csv.Dialect):
                 self.line_pos += 1
                 sniffed = csv.Sniffer().sniff(line)
                 dia.delimiter = sniffed.delimiter
-                if self.encoding is not None:
-                    self.buf.extend(
-                        list(
-                            UnicodeReader(
-                                StringIO(line), dialect=dia, encoding=self.encoding
-                            )
-                        )
-                    )
-                else:
-                    self.buf.extend(list(csv.reader(StringIO(line), dialect=dia)))
 
-            if self.encoding is not None:
-                reader = UnicodeReader(
-                    f, dialect=dia, encoding=self.encoding, strict=True
-                )
-            else:
-                reader = csv.reader(f, dialect=dia, strict=True)
+                # Note: self.encoding is irrelevant here
+                line_rdr = csv.reader(StringIO(line), dialect=dia)
+                self.buf.extend(list(line_rdr))
+
+            # Note: self.encoding is irrelevant here
+            reader = csv.reader(f, dialect=dia, strict=True)
 
         else: