Thank you, Bitbucket. This question is very nice.

Initially I picked 6 and then changed to 5, thinking of the S in A has 6 unique value and then R in A has only 5 unique values and was expecting it will work like LEFT JOIN and was wrong.

Also I was expecting the data will be dirty, but seeing the dataset each row has its unique representation with the combination of A-R and B-R.

SRS

AlphaD12341001

AlphaD12342001

betaA11223001

betaA11224001

bravoC13425001

charlieD12341001

charlieD12342001

harryA11223001

harryA11224001