Drop support for old versions of pandas

In January 2023 pandas 0.2.0 is 80 months old:
https://pypi.org/project/pandas/#history

Before this change pytest displayed a warning:

    qgrid/grid.py:30
    qgrid/grid.py:30
      /home/maxwell-k/github.com/maxwell-k/qgridtrusted/qgrid/grid.py:30: DeprecationWarning: distutils Version classes are deprecated. Use packaging.version instead.
        if LooseVersion(pd.__version__) > LooseVersion('0.20.0'):

After this change pytest displayed one fewer warning.

Author: maxwell-k

qgrid/grid.py

@@ -22,16 +22,6 @@
 from uuid import uuid4
 from six import string_types
 
-# versions of pandas prior to version 0.20.0 don't support the orient='table'
-# when calling the 'to_json' function on DataFrames.  to get around this we
-# have our own copy of the panda's 0.20.0 implementation that we use for old
-# versions of pandas.
-from distutils.version import LooseVersion
-if LooseVersion(pd.__version__) > LooseVersion('0.20.0'):
-    import pandas.io.json as pd_json
-else:
-    from . import pd_json
-
 
 class _DefaultSettings(object):
 
@@ -957,10 +947,12 @@ def should_be_stringified(col_series):
         else:
             self._row_styles = {}
 
-        df_json = pd_json.to_json(None, df,
-                                  orient='table',
-                                  date_format='iso',
-                                  double_precision=self.precision)
+        df_json = df.to_json(
+            None,
+            orient='table',
+            date_format='iso',
+            double_precision=self.precision,
+        )
 
         if update_columns:
             self._interval_columns = []
@@ -1037,10 +1029,10 @@ def should_be_stringified(col_series):
         # and then call 'to_json' again to get a new version of the table
         # json that has interval columns replaced with text columns
         if len(self._interval_columns) > 0 or len(self._period_columns) > 0:
-            df_json = pd_json.to_json(None, df,
-                                      orient='table',
-                                      date_format='iso',
-                                      double_precision=self.precision)
+            df_json = df.to_json(None,
+                                 orient='table',
+                                 date_format='iso',
+                                 double_precision=self.precision)
 
         self._df_json = df_json