Re: Nested definitions

On 2025-07-03 22:46, minforth wrote:

On Thu, 3 Jul 2025 17:42:02 +0000, peter wrote:

[...]

>
I am not fond of mixing definitions inside others.
>

 Section 3.4.5 of the standard document is very clear on this.

<https://forth-standard.org/standard/usage#subsection.3.4.5>
This section applies to standard *programs*, not systems.
A standard system may provide facilities for nested compilation, and a system-dependent program may use these facilities.

On the other hand, quotations do not fall under this verdict
because neither namespace nor search order management is
required.

Yes. There are three categories whose members are prohibited from being executed in a standard program when the current definition exists:
   - a definition that allocates data space,
   - a defining word,
   - `:NONAME` (note that this word is not a defining word).
And the word `[:` does not fall into any of these categories.
NB: "current definition" is a formal term,
see at <https://forth-standard.org/standard/notation#notation:terms>

My <: func .. ;> emulation uses quotations.

I think this should be treated as a nested definition.

No big deal.
I find it useful every now and then.

--
Ruvim