This patch tries to enable mock_Tk.Text._decode to handle the following patterns
insert linestart
insert lineend
insert wordstart
insert wordend
insert +x chars
insert -x chars

These additions are required for testing AutoExpand and are written keeping the same in mind. Also, adds respective tests for test_decode in test_text.py.

I would like to know if my approach is acceptable or whether it needs changes.

issue18292 is about adding unittests for AutoExpand.

(A 2.7 patch will be submitted once the above changes become acceptable)