Re: [PATCH olang v1 2/2] docs: spec: add variables and constants specification

[Johnny Richard <johnny@johnnyrichard.com>]

On Thu, Mar 28, 2024 at 11:11:38AM -0300, Carlos Maniero wrote:
> On Wed Mar 27, 2024 at 6:37 PM -03, Johnny Richard wrote:
> > On Wed, Mar 27, 2024 at 12:21:28AM -0300, Carlos Maniero wrote:
> > > This commit introduces the specification for variables and constants. A
> > > valid program under this specification is as follows:
> > > 
> > >   var x: u32 = 1
> >
> > Since this patch adds support to assignments, lets also add support to
> > all assignment operators like "-=" "+=" "<<=" and so on.
> >
> > >   const y: u32 = 2
> >
> > This patch lacks support to the following valid assignment expression
> > (which I think adds flexibility to the language):
> >
> >     var x: u32 = a = b = c = 1
> >     var y: u32
> >     y = a = b = c = 1
> >
> > > -<global-statements>   ::= <function-definition>
> > > +<global-statements>   ::= <function-definition> | <variable-definition> | <variable-reassign> | <const-definition>
> >
> > nitpick: maybe _external-declaration_ sounds better since definitions are
> > not statements?
> 
> What do you think about *file-declarations* a file declaration? Does not
> mean that it is external. At least while we try to define a file.

hmm.. It sounds weird mention file. Feels like you are going to declare
files... But I don't want to spend a lot of time on it.  It was a
nitpick... If you think *external-declaration* is not good enough we can
keep *global-statement* (no blocking here).

> > > +
> > > +(* Variables *)
> > > +<variable-definition> ::= 'var' <ws> <variable-name> <ows> ':' <ows> <type> <ows> <variable-assign>?
> > > +<const-definition>    ::= 'const' <ws> <variable-name> <ows> ':' <ows> <type> <ows> <variable-assign>
> >
> > We don't need to define twice the variable-definition for "var" and
> > "const", let's combine both in a single rule.
> 
> I don't know if you noticed, but variables has an optional assignment
> while const has required assignment.

Sure, I noticed.  I took a look to how C solve this problem and seems
like it allows defining a const without assignment  (I don't know why...).

I would do the same to keep the gramma simple. And if we really want to
advice the developer about weird declaration, we could warning message
during semantics checking phase.

> > +<variable-definition>  ::= <variable-qualifier> <ws> <variable-name> <ows> ':' <ows> <type> (<ows> <assign-operator> <ows> <expression>)?
>
> This is invalid for const unless we want to allow unassigned constants.
> It also can be interpreted as a semantic error, in that case I believe
> the spec should not check for semantics.

Let's handle it during semantics analysis.