Long range problem

#1

i am getting wrong answer for large test cases
instead of using long in my code

#2

bro this code has an efficient solution which works in O( N ) time

code is attached below

explanation
An efficient solution is to consider pattern of binary values from L to R. We can see that first bit from L to R either changes from 0 to 1 or it stays 1 i.e. if we take the XOR of any two numbers for maximum value their first bit will be fixed which will be same as first bit of XOR of L and R itself.
After observing the technique to get first bit, we can see that if we XOR L and R, the most significant bit of this XOR will tell us the maximum value we can achieve i.e. let XOR of L and R is 1xxx where x can be 0 or 1 then maximum XOR value we can get is 1111 because from L to R we have all possible combination of xxx and it is always possible to choose these bits in such a way from two numbers such that their XOR becomes all 1. It is explained below with some examples,

eg:
L = 8 ( 1000 ) R = 20(10100)

xor would give
11100
msb of L^R is 10000 ( 5th position )
this implies that
a combination occurs b/w l & r which would fetch us the value with all 11111 ( => 15 & 16 )

#3

with this approach my code is running

#4

but my code is giving me wrong answer for large test cases

#5

10000000000 20000000000 this test case is giving me wrong answer instead of using long

#6
#7
#8

Sir u are getting all correct answer

and

yeh test case banega he nhi
l <= r <= 1000
are the given constraints