Do not look for implicits when into.transform set up flags

[MateuszKubuszok]

One of the (unanswered) questions from Gitter mentioned this code:

implicit val childBtoChildBC: Transformer[ChildB, ChildBC] = (childb: ChildB) => {
      childb.into[ChildBC]
        .enableBeanSetters
        .enableBeanGetters
        .transform
    }

While the question was about forward reference extends over definition of value (self reference in the implicit -> cyclical dependency) it macro would not find implicit itself if it skip on looking for an implicit. For our users it seems intuitive to assume that if we put anything between .into and .transform, then implicit will not be looked for. In 0.8.0 we disabled looking for implicit it override is set (withField*, withCoproduct*), but not for flags (enable*, disable*).

It was for a good reason: we are using implicit TransformerConfiguration to share several flags. Some instance of it is always present. If we disable looking for implicit if any flag is set:

• settings any flag with implicit would disable all implicit lookups which probably isn't what people would look for
• if we set a flag by implicit and disable it locally, implicit would be looked up again - even though we just defined a customization

So probably the most intuitive approach would be to disable implicit lookup if any flag is set for instance but not if flags are set globally (via implicit) - if user want to use the implicit Transformer they would not set any customization themselves, if they want to force inline derivation they can repeat any flag (this should be documented).