If more than one censored keyword combination is present, these
algorithms isolate the censored keyword combination whose rightmost
component is leftmost in the censored text.  The existing algorithms
isolate the censored keyword combination whose leftmost component is
rightmost in the censored text.