EHN: to_csv compression accepts file-like object (pandas-dev#21249)

Author: minggli

doc/source/whatsnew/v0.24.0.txt

@@ -13,7 +13,7 @@ New features
 Other Enhancements
 ^^^^^^^^^^^^^^^^^^
 - :func:`to_datetime` now supports the ``%Z`` and ``%z`` directive when passed into ``format`` (:issue:`13486`)
--
+- :func:`to_csv` now supports ``compression`` keyword when a file handle is passed. (:issue:`21227`)
 -
 
 .. _whatsnew_0240.api_breaking:
@@ -184,4 +184,3 @@ Other
 -
 -
 -
-

pandas/conftest.py

@@ -105,6 +105,16 @@ def compression(request):
     return request.param
 
 
+@pytest.fixture(params=['gzip', 'bz2', 'zip',
+                        pytest.param('xz', marks=td.skip_if_no_lzma)])
+def compression_only(request):
+    """
+    Fixture for trying common compression types in compression tests excluding
+    uncompressed case
+    """
+    return request.param
+
+
 @pytest.fixture(scope='module')
 def datetime_tz_utc():
     from datetime import timezone

pandas/core/frame.py

@@ -1689,8 +1689,7 @@ def to_csv(self, path_or_buf=None, sep=",", na_rep='', float_format=None,
             defaults to 'ascii' on Python 2 and 'utf-8' on Python 3.
         compression : string, optional
             A string representing the compression to use in the output file.
-            Allowed values are 'gzip', 'bz2', 'zip', 'xz'. This input is only
-            used when the first argument is a filename.
+            Allowed values are 'gzip', 'bz2', 'zip', 'xz'.
         line_terminator : string, default ``'\n'``
             The newline character or character sequence to use in the output
             file

pandas/core/series.py

@@ -3761,8 +3761,7 @@ def to_csv(self, path=None, index=True, sep=",", na_rep='',
             non-ascii, for python versions prior to 3
         compression : string, optional
             A string representing the compression to use in the output file.
-            Allowed values are 'gzip', 'bz2', 'zip', 'xz'. This input is only
-            used when the first argument is a filename.
+            Allowed values are 'gzip', 'bz2', 'zip', 'xz'.
         date_format: string, default None
             Format string for datetime objects.
         decimal: string, default '.'

pandas/io/formats/csvs.py

@@ -154,9 +154,9 @@ def save(self):
             # GH 17778 handles compression for byte strings.
             if not close and self.compression:
                 f.close()
-                with open(self.path_or_buf, 'r') as f:
+                with open(f.name, 'r') as f:
                     data = f.read()
-                f, handles = _get_handle(self.path_or_buf, self.mode,
+                f, handles = _get_handle(f.name, self.mode,
                                          encoding=encoding,
                                          compression=self.compression)
                 f.write(data)

pandas/tests/test_common.py

@@ -231,13 +231,33 @@ def test_standardize_mapping():
               columns=['X', 'Y', 'Z']),
     Series(100 * [0.123456, 0.234567, 0.567567], name='X')])
 @pytest.mark.parametrize('method', ['to_pickle', 'to_json', 'to_csv'])
-def test_compression_size(obj, method, compression):
-    if not compression:
-        pytest.skip("only test compression case.")
+def test_compression_size(obj, method, compression_only):
 
     with tm.ensure_clean() as filename:
-        getattr(obj, method)(filename, compression=compression)
+        getattr(obj, method)(filename, compression=compression_only)
         compressed = os.path.getsize(filename)
         getattr(obj, method)(filename, compression=None)
         uncompressed = os.path.getsize(filename)
         assert uncompressed > compressed
+
+
+@pytest.mark.parametrize('obj', [
+    DataFrame(100 * [[0.123456, 0.234567, 0.567567],
+                     [12.32112, 123123.2, 321321.2]],
+              columns=['X', 'Y', 'Z']),
+    Series(100 * [0.123456, 0.234567, 0.567567], name='X')])
+@pytest.mark.parametrize('method', ['to_csv'])
+def test_compression_size_fh(obj, method, compression_only):
+
+    with tm.ensure_clean() as filename:
+        with open(filename, 'w') as fh:
+            getattr(obj, method)(fh, compression=compression_only)
+            # GH 17778
+            assert fh.closed
+        compressed = os.path.getsize(filename)
+    with tm.ensure_clean() as filename:
+        with open(filename, 'w') as fh:
+            getattr(obj, method)(fh, compression=None)
+            assert not fh.closed
+        uncompressed = os.path.getsize(filename)
+        assert uncompressed > compressed