Logical binding aliases or binding renaming

[agentgt]

With the exception of POJO getters JStachio uses the exact name to find bindings (variables/sections).

Thus if you want a variable named in kebab style like "inner-message" which is not a valid Java identifier you cannot do it.

This is a problem for folks coming from JSON based models where a such a naming style is more typical.

An obvious solution would be to offer an annotation that you can put on a field or method similar to Jackson's @JsonProperty this seems rather laborious and in some cases a model cannot be annotated.

So careful design considerations need to be made and more feedback.

If you would like to see some solution to this problem please thumb-up!

[dsyer]

I would be OK with just converting kebab-case to camelCase and failing if it's not available. It happens at compile time right, so it's easy to spot?

[agentgt]

It happens at compile time right, so it's easy to spot?

That is correct.

[agentgt]

I have some hesitations that I can't really formulate yet into words about defaulting to converting kebab-case to camelCase.

Partly because kebab-case just happens to not be a valid Java identifier so it happens to be one of the only naming conventions that will work. Also I think it is the least likely used and probably what people will want is snake_case particularly if they have some JSON REST API where they have chosen snake_case.

I will wait to see someone complain about it some more to figure out use cases.