How do I merge rows in a pandas DataFrame while ignoring specified values and checking for conflicts?

Question

I have dataframes that look like this:

  id src col1 col2 col3
0  a   x    5    x    1
1  b   a    5    n    3
2  c   a