One challenge is that rejecting, or even detecting that something is wrong is often not a simple decision to make. Parsing a CSV in an invalid encoding will usu

One challenge is that rejecting, or even detecting that something is wrong is often not a simple decision to make. Parsing a CSV in an invalid encoding will usually produce "valid" but nonsensical output. Determining that that has happened when your destination schema is somewhat freeform can be a challenge.