This follows similar code that we have a few copies of already (Kaidan, Qrca), but updates it to newer ZXing API to use zero-copy scanning at least on all video formats that ZXing can consume directly.
There are a number of apps wanting/needing this (Keysmith, Vakzination, Itinerary) beyond the ones already having similar code, so having it in Frameworks makes sense.
Qt6 changes the way how to integrate video analysis into QtMultimedia video feeds significantly, which is another reason to focus on a single implementation here.