Merge pull request #441 from Simon-Chenzw/master

Support pandas StringArray and ArrowStringArray

clickhouse_driver/numpy/helpers.py

@@ -1,13 +1,16 @@
 import numpy as np
 import pandas as pd
+from pandas.core.arrays import ExtensionArray
 
 
 def column_chunks(columns, n):
     for column in columns:
-        if not isinstance(column, (np.ndarray, pd.DatetimeIndex)):
+        if not isinstance(
+            column, (np.ndarray, pd.DatetimeIndex, ExtensionArray)
+        ):
             raise TypeError(
                 'Unsupported column type: {}. '
-                'ndarray/DatetimeIndex is expected.'
+                'ndarray/DatetimeIndex/ExtensionArray is expected.'
                 .format(type(column))
             )
 

tests/numpy/columns/test_string.py

@@ -3,6 +3,11 @@
 except ImportError:
     np = None
 
+try:
+    import pandas as pd
+except ImportError:
+    pd = None
+
 from tests.numpy.testcase import NumpyBaseTestCase
 
 
@@ -40,6 +45,17 @@ def test_nullable(self):
             )
             self.assertEqual(inserted[0].dtype, object)
 
+    def test_insert_pandas_string(self):
+        with self.create_table('a String'):
+            df = pd.DataFrame({'a': ['a', 'b', 'c']}, dtype='string')
+            self.client.insert_dataframe(
+                'INSERT INTO test VALUES', dataframe=df
+            )
+
+            query = 'SELECT * FROM test'
+            inserted = self.emit_cli(query)
+            self.assertEqual(inserted, 'a\nb\nc\n')
+
 
 class ByteStringTestCase(NumpyBaseTestCase):
     client_kwargs = {'settings': {'strings_as_bytes': True, 'use_numpy': True}}