## RLM - Run-Length Mathematics

no tags

Run-length encoding of a number replaces a run of digits (that is, a sequence of consecutive equivalent digits) with the number of digits followed by the digit itself. For example, 44455 would become 3425 (three fours, two fives). Note that run-length encoding does not necessarily shorten the length of the data: 11 becomes 21, and 42 becomes 1412. If a number has more than nine consecutive digits of the same type, the encoding is done greedily: each run grabs as many digits as it can, so 111111111111111 is encoded as 9161.

Implement an integer arithmetic calculator that takes operands and gives results in run-length format. You should support addition, subtraction, multiplication, and division. You won't have to divide by zero or deal with negative numbers.

### Input/Output

The input will consist of several test cases, one per line. For each test case, compute the run-length mathematics expression and output the original expression and the result, as shown in the examples. The (decimal) representation of all operands and results will fit in signed 64-bit integers.

### Example

```Input:
11 + 11
988726 - 978625
12 * 41
1124 / 1112
13 * 33
15 / 16

Output:
11 + 11 = 12
988726 - 978625 = 919111
12 * 41 = 42
1124 / 1112 = 1112
13 * 33 = 39
15 / 16 = 10
``` vengatesh15: 2017-09-07 13:20:44 easy one .. hodobox: 2015-09-24 23:35:33 used long long -> WA used unsigned long long -> AC Himanshu: 2013-06-17 06:05:59 a tricky case:: i/p: 9292 - 92 o/p: 9292 - 92 = 9290 this cost me 2 WA.:) Last edit: 2013-06-17 07:24:28 (^@_^@): 2012-06-22 06:36:21 learned some disastrous things about string data type for input till endoffile use while (scanf("%s %[+-*/] %s",s1,ch,s2) !=EOF) dont use while(!cin.eof()) with s1 s2 as string data type take as character array cause me many wa Miorel-Lucian Palii: 2009-11-26 07:17:47 Generally I would agree. Unfortunately many contests do not provide such information. This problem is meant to prepare you for such contests ;) And for this particular problem, the time limit doesn't seem to be an issue, as I write this there are nearly 300 submissions and only 4 TLEs. Seshadri R: 2009-11-04 16:46:02 Instead of saying several test cases, it would be helpful if an approximate number is specified. This will enable us to prepare test data of the stated magnitude and test whether our submission would suffer TLE or not