Multi-encoding support in playhouse DataSet

[JarneVerhaeghe]

It appears the DataSet module from playhouse does not support other encodings such as 'latin-1'.

I locally solved it by adjusting the open_file, thaw, and freeze function to have an extra argument:
Line 21-22

def open_file(f, mode,encoding_='utf8'):
        return open(f, mode, encoding=encoding_)

Line 152-156

def freeze(self, query, format='csv', filename=None, file_obj=None, encoding='utf8',
               **kwargs):
        self._check_arguments(filename, file_obj, format, self._export_formats)
        if filename:
            file_obj = open_file(filename, 'w',encoding)

Line 164-168

 def thaw(self, table, format='csv', filename=None, file_obj=None,
           strict=False, encoding='utf8', **kwargs):
      self._check_arguments(filename, file_obj, format, self._export_formats)
      if filename:
          file_obj = open_file(filename, 'r',encoding)

This issue is more of a suggestion for multi-encoding support. I did not want to open a pull request for this.