Discussion about new syntax for logic elements

[AndreasKrall]

Currently logic elements are defined using annotations with a base annotation and annotations depending on the base annotation, e.g.:

[size : 16]              // dependent annotation
[ageordered]             // dependent annotation 
[reservation station]    // base annotation
logic intQueue

Another syntax should be developed where the base annotation is integrated into the definition of the logic element. Possible examples are:

reservation station intQueue
reservation-station intQueue
logic reservation station intQueue
logic [reservation station] intQueue
logic (reservation station) intQueue
logic {reservation station} intQueue
logic intQueue : reservation station
logic intQueue [ reservation station ]
logic intQueue ( reservation station )
logic intQueue { reservation station }
logic intQueue is reservation station
logic intQueue as reservation station
logic intQueue = reservation station
logic intQueue implements reservation station

The advantage of the keyword logic is that no additional keywords have to be added as reservation and station are identifiers.

The result of the discussion (in the comments) is described here:

[AndreasKrall]

I prefer the variant:

logic [reservation station] intQueue

It is close to the old syntax and does not give problems in the grammar. We do not consume new keywords which cannot be used as identifiers. But if we rework the allowedIdentifierKeywords they keywords in the MiA should not raise any problems.

The syntax would be logic "[" identifier [identifier] "]" identifier.

[jaraonthe-dot-net]

I wonder if using square brackets leads to confusion with annotations. I.e. when looking at it quickly, it looks like an annotation, but in a weird place (after instead of before the entity's keyword). Maybe using a different type of braces would be better.

Other than that, I like this order in the syntax. Reminds me of Generics in OOP.