The question is to evaluate $$ \int_0^\infty \frac{dx}{(1+x^2)(1+x^c)} $$ for arbitrary $c\geq0$. Here are my attempts:
(The methods behave somewhat differently for $c=0$ but that case is trivial so I will assume $c>0$)
Integrate by parts: \begin{align*} \int_0^\infty \frac{dx}{(1+x^2)(1+x^c)}&=\arctan(x)\frac{1}{1+x^c}\Big|_0^\infty+\int_0^\infty\arctan(x)\frac{cx^{c-1}}{(1+x^c)^2}dx\\ &=0+\int_0^\infty\arctan(x)\frac{cx^{c-1}}{(1+x^c)^2}dx\\ &=\int_0^\infty\frac{\arctan(x^{1/c})}{(1+x)^2}dx. \end{align*} This looks just as bad, if not worse, as before...
Partial fractions: good for small values of $c$, but seems impossible as $c$ gets large.
Along the same line as partial fractions, but I tried the Ostragodski method, which only helped simplify things for $c=2$.
Try and show it doesn't depend on $c$:
For $c=1$, $c=2$, and $c=3$, the integral evaluates to $\pi/4$. My gut tells me this isn't a coincidence, so define $$ f(c)=\int_0^\infty\frac{dx}{(1+x^2)(1+x^c)} $$ so that $$ f'(c)=\int_0^\infty\frac{-x^c\log(x)}{(1+x^2)(1+x^c)^2}dx. $$ It would be great if this evaluated to $0$ for arbitrary $c$, but I don't see a nice way to show that it does.
Any hints would be greatly appreciated. Thanks in advance.