BUG: to_clipboard fails to format output for Excel (pandas-dev#21111)

Author: david-liu-brattle-1

doc/source/whatsnew/v0.23.2.txt

@@ -57,6 +57,7 @@ Fixed Regressions
 - Bug in both :meth:`DataFrame.first_valid_index` and :meth:`Series.first_valid_index` raised for a row index having duplicate values (:issue:`21441`)
 - Fixed regression in unary negative operations with object dtype (:issue:`21380`)
 - Bug in :meth:`Timestamp.ceil` and :meth:`Timestamp.floor` when timestamp is a multiple of the rounding frequency (:issue:`21262`)
+- Fixed regression in :func:`to_clipboard` that defaulted to copying dataframes with space delimited instead of tab delimited (:issue:`21104`)
 
 .. _whatsnew_0232.performance:
 

pandas/io/clipboards.py

@@ -1,6 +1,7 @@
 """ io on the clipboard """
 from pandas import compat, get_option, option_context, DataFrame
-from pandas.compat import StringIO, PY2
+from pandas.compat import StringIO, PY2, PY3
+import warnings
 
 
 def read_clipboard(sep=r'\s+', **kwargs): # pragma: no cover
@@ -32,7 +33,7 @@ def read_clipboard(sep=r'\s+', **kwargs): # pragma: no cover
 
  # try to decode (if needed on PY3)
  # Strange. linux py33 doesn't complain, win py33 does
- if compat.PY3:
+ if PY3:
  try:
  text = compat.bytes_to_str(
  text, encoding=(kwargs.get('encoding') or
@@ -55,11 +56,27 @@ def read_clipboard(sep=r'\s+', **kwargs): # pragma: no cover
 
  counts = {x.lstrip().count('\t') for x in lines}
  if len(lines) > 1 and len(counts) == 1 and counts.pop() != 0:
- sep = r'\t'
+ sep = '\t'
 
+ # Edge case where sep is specified to be None, return to default
  if sep is None and kwargs.get('delim_whitespace') is None:
  sep = r'\s+'
 
+ # Regex separator currently only works with python engine.
+ # Default to python if separator is multi-character (regex)
+ if len(sep) > 1 and kwargs.get('engine') is None:
+ kwargs['engine'] = 'python'
+ elif len(sep) > 1 and kwargs.get('engine') == 'c':
+ warnings.warn('read_clipboard with regex separator does not work'
+ ' properly with c engine')
+
+ # In PY2, the c table reader first encodes text with UTF-8 but Python
+ # table reader uses the format of the passed string. For consistency,
+ # encode strings for python engine so that output from python and c
+ # engines produce consistent results
+ if kwargs.get('engine') == 'python' and PY2:
+ text = text.encode('utf-8')
+
  return read_table(StringIO(text), sep=sep, **kwargs)
 
 
@@ -99,7 +116,7 @@ def to_clipboard(obj, excel=True, sep=None, **kwargs): # pragma: no cover
  if excel:
  try:
  if sep is None:
- sep = r'\t'
+ sep = '\t'
  buf = StringIO()
  # clipboard_set (pyperclip) expects unicode
  obj.to_csv(buf, sep=sep, encoding='utf-8', **kwargs)
@@ -108,8 +125,11 @@ def to_clipboard(obj, excel=True, sep=None, **kwargs): # pragma: no cover
  text = text.decode('utf-8')
  clipboard_set(text)
  return
- except:
- pass
+ except TypeError:
+ warnings.warn('to_clipboard in excel mode requires a single '
+ 'character separator.')
+ elif sep is not None:
+ warnings.warn('to_clipboard with excel=False ignores the sep argument')
 
  if isinstance(obj, DataFrame):
  # str(df) has various unhelpful defaults, like truncation

pandas/tests/io/test_clipboard.py

@@ -88,8 +88,6 @@ def check_round_trip_frame(self, data, excel=None, sep=None,
  tm.assert_frame_equal(data, result, check_dtype=False)
 
  # Test that default arguments copy as tab delimited
- @pytest.mark.xfail(reason='to_clipboard defaults to space delim. '
- 'Issue in #21104, Fixed in #21111')
  def test_round_trip_frame(self, df):
  self.check_round_trip_frame(df)
 
@@ -99,10 +97,6 @@ def test_round_trip_frame_sep(self, df, sep):
  self.check_round_trip_frame(df, sep=sep)
 
  # Test white space separator
- @pytest.mark.xfail(reason="Fails on 'delims' df because quote escapes "
- "aren't handled correctly in default c engine. Fixed "
- "in #21111 by defaulting to python engine for "
- "whitespace separator")
  def test_round_trip_frame_string(self, df):
  df.to_clipboard(excel=False, sep=None)
  result = read_clipboard()
@@ -111,21 +105,17 @@ def test_round_trip_frame_string(self, df):
 
  # Two character separator is not supported in to_clipboard
  # Test that multi-character separators are not silently passed
- @pytest.mark.xfail(reason="Not yet implemented. Fixed in #21111")
  def test_excel_sep_warning(self, df):
  with tm.assert_produces_warning():
  df.to_clipboard(excel=True, sep=r'\t')
 
  # Separator is ignored when excel=False and should produce a warning
- @pytest.mark.xfail(reason="Not yet implemented. Fixed in #21111")
  def test_copy_delim_warning(self, df):
  with tm.assert_produces_warning():
  df.to_clipboard(excel=False, sep='\t')
 
  # Tests that the default behavior of to_clipboard is tab
  # delimited and excel="True"
- @pytest.mark.xfail(reason="to_clipboard defaults to space delim. Issue in "
- "#21104, Fixed in #21111")
  @pytest.mark.parametrize('sep', ['\t', None, 'default'])
  @pytest.mark.parametrize('excel', [True, None, 'default'])
  def test_clipboard_copy_tabs_default(self, sep, excel, df):
@@ -139,10 +129,6 @@ def test_clipboard_copy_tabs_default(self, sep, excel, df):
  assert clipboard_get() == df.to_csv(sep='\t')
 
  # Tests reading of white space separated tables
- @pytest.mark.xfail(reason="Fails on 'delims' df because quote escapes "
- "aren't handled correctly. in default c engine. Fixed "
- "in #21111 by defaulting to python engine for "
- "whitespace separator")
  @pytest.mark.parametrize('sep', [None, 'default'])
  @pytest.mark.parametrize('excel', [False])
  def test_clipboard_copy_strings(self, sep, excel, df):
@@ -193,8 +179,6 @@ def test_invalid_encoding(self, df):
  with pytest.raises(NotImplementedError):
  pd.read_clipboard(encoding='ascii')
 
- @pytest.mark.xfail(reason='to_clipboard defaults to space delim. '
- 'Issue in #21104, Fixed in #21111')
  @pytest.mark.parametrize('enc', ['UTF-8', 'utf-8', 'utf8'])
  def test_round_trip_valid_encodings(self, enc, df):
  self.check_round_trip_frame(df, encoding=enc)