Re: neos Universal Compiler and FORTRAN IV

On 2025-03-30 14:00:21 +0000, Mr Flibble said:

On Sun, 30 Mar 2025 12:04:27 +0300, Mikko wrote:
 

On 2025-03-29 15:47:33 +0000, Mr Flibble said:
 

On Sat, 29 Mar 2025 10:11:13 +0000, Richard Heathfield wrote:
 

On 29/03/2025 09:51, Mikko wrote:

On 2025-03-28 14:14:10 +0000, Mr Flibble said:
 

<snip>

 If a native FORTRAN IV compiler can parse it then neos can parse it
as long as it is somehow expressable in the neos attribute grammar.

 If you don't know how to express it in the neos attribute grammar
then nobody knows and neos is not universal, at lieast in the
practical sense.

  You don't understand, Mikko. You probably think that "neos can parse
it"
means that neos can parse it. What it really means is that /in Mr
Flibble's head/ neos can parse it, and anything in Mr Flibble's head
is right by definition, and /therefore/ neos can parse FORTRAN IV even
if it can't.

 I see no reason why neos can't parse FORTRAN IV, FACT.

 I can see a reason why it can't be used for that purpose: you have not
told how it can be done. If you can't tell how to use neos then nobody
is going to use it. If you can't solve the small problem of tokenizing
the example lines
 REALITY IS, AS I MAY LOGICALLY SAY, REAL LOGICAL = WHAT I SAY
 then nobody will try the bigger problem of tokenizing all of FORTRAN IV
or any later Fortran and parsing and compiling it.
 It does not really matter what can be done if nobody is going to try.

 I already indicated that a tokenisation stage is entirely optional in neos.

It can be merged with the parser but the parser must be able to detect and
identify tokens or at least identifiers. If you can't tell how to do that
then your neos is not unversal.
--
Mikko