Description

Customers of the Bank of Kalii handle their banking transactions similar to the way they handle their taxes: be as terse as possible. As a result, when a customer writes a check or fills out a deposit or withdrawal form, they leave off the year on any date they write down. So, instead of writing: 09/20/2005, they would write: 9/20 and be done with it. In general, the year can be inferred since it will be relatively close to the date the transaction is actually processed by the bank.

Without going into the intricate details of how the Bank of Kalii calculates interest and banking fees (that is a problem for another time...), suffice to say the bank must determine the actual date the customer wrote on the check or form, and calculate the number of days prior (or in the future) the document is dated. You see, Kaliian bankers, like their government officials, are overworked, so they may not get around to processing transactions for up to a week. The customers know this, so they often date their checks and forms a several days in the future - this complicates the bankers' duties as well.

Your job is to write a program to compare a date written on a check or form with the date the transaction is being processed, and, determine the full date the customer meant as well as how many days prior (or in the future) the document is dated.

Input

The first line of input contains an integer N which is the number of datasets that follow (1 ≤ N ≤ 1000). Each dataset consists of a single line containing two dates: the transaction date and the document date; there is a single space between them. The transaction date is of the form M/D/Y where M is the month number (1 ≤ M ≤ 12), D is the day of month (1 ≤ D ≤ md₁) and Y is the year (2000 ≤ Y ≤ 2200). The document date is of the form m/d where m is the month number (1 ≤ m ≤ 12) and d is the day of month (1 ≤ d ≤ md₂). The values of md₁ and md₂ will not exceed the number of days in the respective months M and m.

Output

For each dataset print out the dataset number followed by a space followed by the result of the date comparison as shown in the table below:

7
11/20/2005 11/21
11/20/2005 11/17
11/20/2005 11/20
11/20/2005 11/13
11/20/2005 11/28
1/2/2005 12/30
12/31/2100 1/3

1 11/21/2005 IS 1 DAY AFTER
2 11/17/2005 IS 3 DAYS PRIOR
3 SAME DAY
4 11/13/2005 IS 7 DAYS PRIOR
5 OUT OF RANGE
6 12/30/2004 IS 3 DAYS PRIOR
7 1/3/2101 IS 3 DAYS AFTER

Source

Greater New York 2005