Show that the most general transformation fixing the origin and preserving distances is either a rotation, or a rotation followed by a reflection in the real axis, for a transformation $f: \mathbb{C} \to \mathbb{C}$.
I was thinking of first assuming that $f$ is analytic, then the Schwarz' Lemma will give us that $f(z)=az$ for $|a|=1$, showing that $f$ is indeed a rotation. However, how can one then show that for a general transformation $f$ will be a rotation followed by a reflection in the real axis?
Consider a transformation $f$ that preserves distances and fixes $0$. It follows that $\vert f(1)\vert=1$, that is $f(1)=e^{it}$ for some real $t$. Thus the function $g:\mathbb{C}\to\mathbb{C}$ defined by $g(z)=e^{-it}f(z)$ preserves distances and also satisfies $g(0)=0$, $g(1)=1$.
Now, let $w=e^{i\pi/3}$, the points $0$, $1$ and $w$ form an equilateral triangle, so their images $g(0)=0$, $g(1)=1$, and $g(w)$ also form an equilateral triangle. Thus, we have two cases, either $g(w)=w$ or $g(w)=\bar{w}$.
Case 1. $g(w)=w$. Let $z\in\mathbb{C}$. Since $g$ preserves the two circles $C(0,\vert z\vert)$ and $C( 1,\vert z-1\vert)$, so it preserves their intersection which is $\{z,\bar{z}\}$, that is $g(z)\in\{z,\bar{z}\}$. Now, if $z=\bar{z}$ we get $g(z)=z$. So, assume that $z\ne\bar{z}$, if $g(z)=\bar{z}$ we get $\vert z - w\vert=\vert \bar{z} - w\vert =\vert z-\bar{w} \vert$, and this implies that $z=\bar{z}$, which is absurd. Therefore we must have $g(z)=z$ in this case. Consequently, $f(z)=e^{it} z$, and $f$ is the rotation of angle $t$ around the origin $O$.
Case 2. $g(w)=\bar{w}$. We consider $h:\mathbb{C}\to\mathbb{C}$ defined by $h(z)=\overline{g(z)}$. Clearly, $h$ preserves distances, and fixes the points $0$, $1$ and $w$, so, according to the previous case $h(z)=z$ for every $z$. That is $f(z)=e^{it}\overline{z}=\overline{e^{-it}z}$ for every $z$. Which is the desired conclusion.