SetUID Error Handling Refactor

[AlexBVolcy]

I noticed in the /setuid endpoint, that the error handling seemed repetitive and could be consolidated. So I wrote a function to simplify the NewSetUidEndpoint() functions error handling.

[VeronikaSolovei9]

Local testing with Prometheus metrics looks good. Code makes sense.

[guscarreon]

Why did we remove this line? In order for the refactored version to be completely equivalent, shouldn't we keep it?

121         if shouldReturn, status, body := preventSyncsGDPR(gdprRequestInfo, gdprPermsBuilder, tcf2Cfg); shouldReturn {
    +           w.Write([]byte(body))
122             var metricValue metrics.SetUidStatus
123             switch status {
124             case http.StatusBadRequest:
125                 metricValue = metrics.SetUidBadRequest
126             case http.StatusUnavailableForLegalReasons:
127                 metricValue = metrics.SetUidGDPRHostCookieBlocked
128             }
129             handleBadStatus(w, status, metricValue, errors.New(body), metricsEngine, &so)
130             return
131         }
endpoints/setuid.go

[AlexBVolcy]

So after looking into it more, this body value that preventSyncsGDPR is essentially a custom error message. So I don't think there's much of a difference between w.Write([]byte(body)) and w.Write([]byte(errors.New(body)) which is how it's written in handleBadStatus.

So I'd advocate for keeping the code how it is, and keeping it a bit cleaner. However, it is a very small difference, so I'm not totally against changing it either. What do you think?

[VeronikaSolovei9]

I'm ok to approve it again. Gus, what do you think it should be?

[guscarreon]

LGTM