never mind about the limits; is the general process correct?
While the integral over the whole positive domain is , the function is NOT even. For this to be a density function, since .
Then again, there are infinitely many possible limits for and that will work. I am not 100% confident in my Prob/Stats knowledge though, so don't quote me on this.