World: r3wp

We are aiming for a purely syntactic description. So semantic checks 
at a later stage are out of the picture for now (no action blocks 
in general, and here: no loading of digits in particular).

But basic accuracy is important :)

Yes, 1000.1000.1000 is recognized as a tuple. The fact that it is 
"too big" is another check.

So ".0." should be recognised as valid, that's important. ".1234." 
being invalid, not so much.

(Same for integer and decimal overflows, as Steeve already remarked.)

tuple-syntax: [0 3 digit #"." 1 3 digit #"." 0 7 [1 3 digit #"."] 
and termination]

I know there is still the need to check the overflow but I think 
it's slightly better than using some or any digits

No need to check overflow, as stated above.

But it has to match syntactically valid tuples :)

The above would fail for "0000..", no?

no, but it would fail another cases. I forgot the last number

tuple-syntax: [0 3 digit #"." 1 3 digit #"." 0 7 [1 3 digit #"."] 
0 3 digit and termination]

Still fails for {0000..}.

to get rid of the 3-digit enforcement:

tuple-syntax: [any digit #"." some digit #"." 0 7 [some digit #"."] 
any digit and termination]

wait...

0..0
 :)

yep

replace some by any

But but... it's not only the leading and trailing zeros that can 
be removed.
It takes me time......

So the initial implementation of ladislav was right from the start

adding the termination is needed, though

let me try again :-)

the only requested digit is the second  one.
.0.... is valid

or/and the first one

0..... is valid
.0..... is valid

aha, did not check the second one, needs a correction, then

tuple-syntax: [[some digit 2 9 [#"." any digit] | #"." some digit 
1 8 [#"." any digit] ] and termination]

should be better

OK

Fiouuuu.... that one was hard :-)

Well.. may be silly question, but can I use those parse magics in 
my projects? I didn't see license information.

A very good question. No license attached, yet. We'll fix that.

Ok, thanks.

Andreas and Ladislav, already off ?

I'm around.

Still miss some datatypes ;-)

Have a go at adding some more, then :)

I don't want a collision. I thought your or Ladislav would finish 
it with the others making comments

Can do a merge ...

The slash words are weird. Didn't know that specific syntax.
It works with R2 not with R3 anymore.

Maybe it was a mistake to allow such syntax in R2, so it's why it 
has been deprecated by Carl.

Should we keep it ?

(See the last update of Ladislav)

Not working though (the first part is a valid refinement in R2 ):
>> [///x//]
== [///x //]

Definitivly, I don't think we should  keep  such syntax even in R2. 
It makes no sense to use such (it was an error corrected in R3).

I think it was mainly meant as a way to make / based ops   things 
like //  and  ///  .  I don't see why it should be removed.  it can 
only contain  "/"  characters.  otherwise its a refinement.

there's no way [///x] should be a valid refinement. It will fail 
when evaluated

But yeah it's another problem

I mixed things, but there is something wrong with the both datatypes 
in R2 (it"s why it has been corrected in R3)

yeah, except when we use refinements as such within dialects, then 
it doesn't matter, since they are not being used as paths.

but its ugly so,  ///x  shouldn't be "promoted"  IMHO.   if its an 
un-intended loophole in the lexical analysis, then it shouldn't be 
within the syntax.r code.

about the more-less-word rule:
>> and << are allowed