The Base-1 Number System

As we know, in an n-based number system, there are n different types of digits. In this way, a 1-based number system has only 1 type of digit, the "0". Here are the rules to interpret 1-based numbers. Each number consists of some space separated blocks of 0. A block may have 1, 2 or more 0s. There is a "flag" variable associated with each number

• A block with a single 0 sets "flag" variable to 1
• A block with two 0s sets the "flag" to 0
• If there are n (n > 2) 0s in a block, n–2 binary digits with the current value of flag is appended to your number.

Note that, the first block of every number will have at most 20s. For example, the 1-base number 0 0000 00 000 0 0000 is equivalent to binary 11011.

• 1st block sets the flag to 1
• 2nd block has 40s. So append flag(=1) 4–2 = 2 times (11).
• 3rd block has 20s. Set the flag to 0
• 4th block has 30s. Append flag(=0) 3-2 = 1 time (110).
• 5th block has a single 0. Set flag = 1
• 6th and block has 40s. Append flag(=0) 4-2 = 2 times (11011).

The final binary number won’t have more than 30 digits. Once, you’ve completed the process, convert the binary value to decimal & print, you’re done!

Input

Input will have at most 2010 test cases. Each case consists of a 1-based number as described above. A number may be spanned to multiple lines but a single block will always be in a single line. Termination of a case will be indicated by a single "#" char which will be space-separated from the last digit of your input number. The last case in the input is followed by a "~" character indicating, end of input.

Output

For each test case, output a single line with the decimal equivalent value of your given 1-based number.