I was watching a video of a professor who said that $\int \frac{1}{1-x^2} dx$ can be done using the method of integration by parts. Though he skipped the solution and simply gave the answer, I tried solving it myself but I'm having difficulty moving forward. What I did was:
Let $u = \frac{1}{1-x^2}, du = \frac{2x}{(1-x^2)^2} dx , dv = dx, v=x$.
Then, $$\int \frac{dx}{1-x^2} = \frac{x}{1-x^2} - \int \frac{2x^2}{(1-x^2)^2} dx$$
However, the new integral seems problematic. I tried doing the method again but the equation just ends up to be $0=0$.
Is integration by parts not a viable way of solving this integral?
P.S.: I already know the answer using a different method (partial fraction decomposition). I just want to know how to solve the integral in a different way.
Partial fraction here is the best way, because $$\frac{1}{1-x^2}=\frac{-1/2}{1-x}+\frac{1/2}{1+x}.$$ Then we have linear factor and it is not difficult to evaluate.
If you want a different method, you can consider write $$\frac{1}{1-x^2}=\frac{1}{(\sqrt{1-x^2})^2}$$ and then the trigonometric substitution $u(x)=\sin x$.
The way using integration by parts as you noticed seems really impractical, if you work with the second antiderivative you eventually you might be interested in $\int \frac{x^4}{(1-x^2)^3}\, dx$ and you don't get out of that growth of powers. At some point we might decide to stop you and use another method, but that wouldn't make much sense.
I guess they tried to say "partial fractions" in the video as the question suggested by the professor Hans in the comments.