writing

Some semantic analogies to syntactic constructs in Jevko

The Standard Jevko Grammar is specified in a way which is amenable to representation in terms of common programming language structures.

In particular:

Jevko = *Subjevko Suffix

says that a Jevko is a sequence of zero or more Subjevkos followed by a Suffix.

Syntactic sequences usually convey collections of some sort.

A particular kind of collection is a collection of key-value pairs. Which brings us to the definition of Subjevko:

Subjevko = Prefix "[" Jevko "]"

A Subjevko is a Prefix followed by an [, followed by a Jevko, followed by a ].

If we ignore the delimiting brackets, a Subjevko is then a Prefix-Jevko pair. A mapping from Subjevkos into key-value pairs is straightforward.

An example:

key1 [value1]
key2 [value2]

That’s a Jevko made of 2 Subjevkos. That is analogous to a collection of 2 key-value pairs. From here, translating a Jevko parse tree to a proper collection is trivial.

Another basic kind of collection is an ordered collection of values. Which could be seen as a special case of a collection of key-value pairs, where keys are implied by the syntactic order of values.

This can be represented with Subjevkos with empty Prefixes:

[value1]
[value2]

would represent an orederd collection of 2 simple values.