For maxdiff, you can do an export and one of the files is the .position which is a record of the position of the chosen items. You can look for respondents then that have answers like 1,2,1,2,1,2. Cbc this is easy as all answers are already position based. In 8.3, Acbc is now easier to spot this. Screeners would look like 1,1,1,1 or 0,0,0,0. Choice tasks and calibrations are just like Cbc.
Other than that I don't know of any automated way to perform this. If you want the entire survey checked for straight liners, you would have to do this manually for now.