Adjust to Pandas breaking change, bump Python version (#101)

Pandas 1.1 has broken Record Mover's usage of the read_csv() function by adding error checking in cases where a certain argument would be unused.

Details of the Pandas change:

* pandas-dev/pandas#27394
* pandas-dev/pandas#31383

See Records Mover test failures here: 
* https://app.circleci.com/pipelines/github/bluelabsio/records-mover/1089/workflows/e62f1cf0-f8d0-4e22-9652-112df72b02b8/jobs/9439

Author: vinceatbluelabs

.python-version

@@ -1 +1 @@
-records-mover-3.8.3
+records-mover-3.8.5

deps.sh

@@ -7,7 +7,7 @@ export SLUGIFY_USES_TEXT_UNIDECODE=yes
 brew update && ( brew upgrade pyenv || true )
 pyenv rehash  # needed if pyenv is updated
 
-python_version=3.8.3
+python_version=3.8.5
 # zipimport.ZipImportError: can't decompress data; zlib not available:
 #    You may need `xcode-select --install` on OS X
 #    https://github.com/pyenv/pyenv/issues/451#issuecomment-151336786

metrics/coverage_high_water_mark

@@ -1 +1 @@
-93.7100
+93.7200

metrics/flake8_high_water_mark

@@ -1 +1 @@
-170
+169

records_mover/records/delimited/csv_streamer.py

@@ -41,10 +41,21 @@ def stream_csv(filepath_or_buffer: Union[str, IO[bytes]],
         'header': header,
         'compression': pandas_compression_from_hint[compression_hint],
         'escapechar': hints.get('escape'),
-        'prefix': 'untitled_',
         'iterator': True,
         'engine': 'python'
     }
+    if header is None:
+        # Pandas only accepts the prefix argument (which makes for
+        # tidier column names when otherwise not provided) when the
+        # header is explicitly marked as missing, not when it's
+        # available or even when we ask Pandas to infer it.  Bummer,
+        # as this means that when Pandas infers that there's no
+        # header, the column names will end up different than folks
+        # explicitly tell records mover that there is no header.
+        #
+        # https://github.com/pandas-dev/pandas/issues/27394
+        # https://github.com/pandas-dev/pandas/pull/31383
+        kwargs['prefix'] = 'untitled_'
     if 'quoting' in hints:
         quoting = hints['quoting']
         kwargs['quoting'] = pandas_quoting_from_hint[quoting]

records_mover/records/mover.py

@@ -17,7 +17,7 @@
 
 def move(records_source: RecordsSource,
          records_target: RecordsTarget,
-         processing_instructions: ProcessingInstructions=ProcessingInstructions()) -> MoveResult:
+         processing_instructions: ProcessingInstructions = ProcessingInstructions()) -> MoveResult:
     """Copy records from one location to another.  Applies a sequence of
     possible techniques to do this in an efficient way and respects
     the preferences set in records_source, records_target and

records_mover/records/pandas/read_csv_options.py

@@ -161,7 +161,13 @@ def pandas_read_csv_options(records_format: DelimitedRecordsFormat,
     # Leaving this in case a future version of Pandas behaves
     # better.
     #
-    pandas_options['prefix'] = 'untitled_'
+    if pandas_options['header'] is None:
+        # Pandas only accepts the prefix argument when the
+        # header is marked as missing.
+        #
+        # https://github.com/pandas-dev/pandas/issues/27394
+        # https://github.com/pandas-dev/pandas/pull/31383
+        pandas_options['prefix'] = 'untitled_'
 
     #
     # mangle_dupe_cols : bool, default True

tests/unit/records/test_csv_streamer.py

@@ -24,7 +24,6 @@ def test_stream_csv_no_compression(self,
                                              engine='python',
                                              escapechar=None,
                                              header='infer',
-                                             prefix='untitled_',
                                              quoting=0,
                                              iterator=True,
                                              sep=mock_field_delimiter)
@@ -49,7 +48,6 @@ def test_stream_csv_gzip(self,
                                              engine='python',
                                              escapechar=None,
                                              header='infer',
-                                             prefix='untitled_',
                                              iterator=True,
                                              sep=mock_field_delimiter)
             mock_io.TextIOWrapper.assert_not_called()
@@ -72,7 +70,6 @@ def test_stream_filename(self,
                                              engine='python',
                                              escapechar=None,
                                              header='infer',
-                                             prefix='untitled_',
                                              iterator=True,
                                              sep=mock_field_delimiter)
             mock_io.TextIOWrapper.assert_not_called()

tests/unit/records/test_pandas_read_csv_options.py

@@ -118,7 +118,6 @@ def test_pandas_read_csv_options_csv(self):
             'engine': 'python',
             'error_bad_lines': True,
             'header': 0,
-            'prefix': 'untitled_',
             'quotechar': '"',
             'quoting': 0,
             'warn_bad_lines': True,