As the data are from a continuous distribution, this will never happen.
The observations are assigned the mean rank for their positions in the ordered list.
The observations are thrown out and not considered in the analysis.
The equal values are arbitrarily assigned an appropriate rank in sequence.
The observations are assigned the mean rank for their positions in the ordered list.