Consider using less general names in example schemas

I was looking at the example QueriesResolver and saw that it implemented Query from io.github.kobylynskyi. I wasn’t sure what this class was or why I’d need to implement it. Then I think I figured out that it’s probably being generated from the schema which has a query of type Query. Perhaps we could call this ProductQuery to make it clearer that it’s specific to the example schema and not something general that every resolver needs to implement?

[kobylynskyi]

• @benmccann generic Query class (having all queries) is generated because somebody can have a pretty small schema with 5-7 queries and it might make sense to implement Query instead of implement FirstQuery, SecondQuery, ThirdQuery, FourthQuery, etc. It is up to the client to decide which way to go, that’s why we are generating both.

• I think it is a pretty good idea to have Query class to be named QueryResolver and all AbcQuery classes to be named AbcQueryResolver classes. But with such minor “improvement” we could brake client implementation. So if we do this, I would agree with @joffrey-bion and introduce config options for configuring prefix/suffix for generated apps. Created #136 for this.

@benmccann, @joffrey-bion, Thanks for bringing this up. Closing the issue.

[joffrey-bion]

@benmccann no problem, you raise good points that are useful to consider 😉

I will let @kobylynskyi take a look at this and close as he pleases. He’s the maintainer of the repo and may have a different opinion about this.

the resolver interface generated for a regular type Xxx is named XxxResolver (see this line in the code). So now I would vote for renaming Query to QueryResolver for the sake of consistency 😄

In particular, what do you think of this @kobylynskyi ?