Litusiano was pooing when he came up with the following problem:
Given $$$l$$$ and $$$r$$$, you have to count how many dibonacci numbers are in the range $$$[l, r]$$$.
A dibonacci number is an integer whose digits are the sum modulo $$$10$$$ of the two previous digits.
For example, $$$123$$$ is a dibonacci number, because $$$(1 + 2)$$$ mod $$$10 = 3$$$, $$$471$$$ is also a dibonacci number because $$$(4 + 7)$$$ mod $$$10 = 1$$$ and $$$2169$$$ is not a dibonacci number because $$$(1 + 6)$$$ mod $$$10 = 7$$$, which is different from $$$9$$$.
Note that all the positive integers $$$\leq 99$$$ are also dibonacci numbers.
Each test contains multiple test cases. The first line contains the number of test cases $$$t$$$ $$$(1 \leq t \leq 69 \cdot 10^{4})$$$. The description of the test cases follows.
For each test case:
There is a line consisting of two integers $$$l, r$$$ $$$(1 \leq l \leq r \leq 10^{18})$$$ — the asked range.
For each test case, output how many dibonacci numbers are in the given range.
52 71 1246 9777337773 36969100000000 1000000000
6 102 452 49 90
In the first test case, the answer is $$$6$$$ because all the numbers in the range $$$[2, 7]$$$ are dibonacci numbers.
In the second case, there are $$$102$$$ dibonacci numbers in the range $$$[1, 124]$$$, like $$$101$$$ or $$$69$$$.