(EDITED on Manmay's request)
Go by the restrictions imposed by different functions we have...oi shob restriction er intersection nile tumi domain paabe.
R1 : Denominator cannot be zero
[x] + 2 ≠0
=> [x] ≠-2
Now [x] = -2 for -2 ≤ x < -1
So [x] ≠-2 for x < -2 U x ≥ -1
Let me explain. For [x] = 2, we are given that x ≥ -2 AND x < -1. Let (-2, ∞) = A and (-∞, -1] = B. So our expression becomes (A ∩ B) where ∩ is intersection/AND. Now when we have to find for [x] ≠2, we complement that set.
By DeMorgan's law, (A ∩ B)' = A' U B' which is the set I found.
So we have x → (-∞, -2) U [-1, ∞)
R2 : Square root term must be non negative
4 - x²[x] + 2 ≥ 0
We have two subcases here.
CASE I : Nr ≥ 0 AND Dr > 0
So x² ≤ 4 AND [x] > -2
Here you must know about the Greatest Integer inequalities. I posted them here : http://www.targetiit.com/iit-jee-forum/posts/greatest-integer-function-exhaustive-collection-pl-13636.html
[x] > I implies x ≥ I + 1
So [x] > -2 implies x ≥ -1
So |x| ≤ 2 AND x ≥ -1
=> -2 ≤ x ≤ 2 AND x ≥ -1
Therefore x → [-1, 2]
or we have CASE II : Nr ≤ 0 AND Dr < 0
|x| ≥ 2 AND [x] < -2
=> x ≤ -2 U x ≥ 2 AND x < -2 (as [x] < I implies x < I)
=> x → (-∞, -2)
So our final restriction(2nd) is x→ (-∞, -2) U [-1, 2] (UNION of both cases)
Now our domain of function is obtained by :
Df = R1 ∩ R2
= x → (-∞, -2) U [-1, ∞) ∩ (-∞, -2) U [-1, 2]
So our final domain is x → (-∞, -2) U [-1, 2]