the deformation of contours

Here is a practical superpower that follows from Cauchy's theorem: you may bend, slide, and reshape a contour freely without changing the integral, as long as the part of the plane you sweep across is one where f stays holomorphic. The integral cares only about the homotopy class of the path — what it encloses — not its exact shape. This turns ugly contours into convenient ones.

Concretely, the deformation principle says: if two contours (with the same endpoints, or two closed loops) can be continuously slid into one another through a region where f is holomorphic, their integrals are equal. The proof idea is to look at the thin band swept between the old and new contour; that band lies in a hole-free, holomorphic patch, so by Cauchy's theorem the integral over its closed boundary is zero, which forces the two contour integrals to match. For closed loops the upshot is a license to replace a complicated loop by, say, a tiny circle around the one singularity it encloses.

This is the everyday workhorse of integral evaluation. Want the integral of f around an awkward closed curve that encircles a single pole at z0? Shrink the curve down to a small circle tightly hugging z0 — the value is unchanged because the region you swept is singularity-free — and the small circle is easy to compute. The only rule you must respect is that you may not drag the contour across a point where f fails to be holomorphic; a singularity is a wall the deformation cannot pass through, and trying to cross it is exactly where the integral would jump.