TST: add test for read_csv with unicode bug

see pandas-dev#13549

Author: nateGeorge

pandas/io/tests/parser/common.py

@@ -8,6 +8,7 @@
 import re
 import sys
 from datetime import datetime
+from io import BytesIO
 
 import nose
 import numpy as np
@@ -1583,3 +1584,13 @@ def test_temporary_file(self):
         new_file.close()
         expected = DataFrame([[0, 0]])
         tm.assert_frame_equal(result, expected)
+
+    def test_read_csv_utf_aliases(self):
+         # see gh issue 13549
+         expected = pd.DataFrame({'mb_num': [4.8], 'multibyte': ['test']})
+         for byte in [8, 16]:
+             for fmt in ['utf-{0}', 'utf_{0}', 'UTF-{0}', 'UTF_{0}']:
+                 encoding = fmt.format(byte)
+                 data = 'mb_num,multibyte\n4.8,test'.encode(encoding)
+                result = self.read_csv(BytesIO(data), encoding=encoding)
+                tm.assert_frame_equal(result, expected)