OpenEuphoria

1. procedures calling procedure calling procedure...

• Posted by rubis at fem.unicamp.br Oct 14, 2002
• 489 views

Hi people !

How do I call a procedure that is after the call ?

for example:

--begin
procedure a()
?1
end procedure

procedure b()
a()
c()
end procedure

procedure c()
?3
end procedure

--end

I'm doing this using call_proc(c(), {}), but something tells me that there 
is a more easy or intellignet way to do this...

thanks
Rubens

3. Re: procedures calling procedure calling procedure...

• Posted by francis at gmx.co.uk Oct 14, 2002
• 445 views

Cunning? Well, you can define routine ID constants/variables at the top of the 
file, reference them in the middle and set them at the end. Even in C, it is 
ordinary programming practice to "#define function(parameters);" if it is called
before it is instantiated. It would be slightly more convenient to be able to 
call *before* it is defined, ut we are talking about an interpreted language! In
terms of coding, I think is is not too much work around at all.

>Sorry but there isn't. RDS believes this is the best way to do it. I strongly 
disagree with RDS on 
>that issue.
>
>The RDS philosophy seems to be that good programming practice is to only refer 
to things that have 
>already been "seen" by the one-pass interpreter. Thus you cannot refer to any 
identifier that is 
>defined lower down in the file. So yes, the physical layout of indentifier 
definitions in a source 
>code file has been deemed to be vitally significant by RDS.
>
>This means that you as a programmer have to invent cunning ways to get around 
this limitation in the 
>language (as implemented by RDS). Of course, RDS has also recognised that there
>
are legitimate 
>reasons for coders to do this, so Euphoria has been provided with routine_id(),
>
call_proc/func() 
>routines.
>
>My suspicions are that it is now too hard for RDS to change the way the 
interpreter works to allow 
>for forward referencing, so we are stuck with this limitation for now.
>
>Whenever I get a spare 12-months, I'll write a new implementation of the 
interpreter 
>
>-------------
>Derek.
>
>15/10/2002 7:21:59 AM, rubis at fem.unicamp.br wrote:
>
>>
>>Hi people !
>>
>>How do I call a procedure that is after the call ?
>>
>>for example:
>>
>>--begin
>>procedure a()
>>?1
>>end procedure
>>
>>procedure b()
>>a()
>>c()
>>end procedure
>>
>>procedure c()
>>?3
>>end procedure
>>
>>--end
>>
>>I'm doing this using call_proc(c(), {}), but something tells me that there 
>>is a more easy or intellignet way to do this...
>>
>>thanks
>>Rubens
>>
>>
>---------
>Cheers,
>Derek Parnell 
>ICQ# 7647806
>
>
>
>

5. Re: procedures calling procedure calling procedure...

• Posted by "Carl W." <euphoria at cyreksoft.yorks.com> Oct 15, 2002
• 451 views

<rubis at fem.unicamp.br> wrote:

> How do I call a procedure that is after the call ?
>
> for example:
>
> --begin
> procedure a()
> ?1
> end procedure
>
> procedure b()
> a()
> c()
> end procedure
>
> procedure c()
> ?3
> end procedure
>
> --end
>
> I'm doing this using call_proc(c(), {}), but something tells me that there
> is a more easy or intellignet way to do this...

Hi.

Like others have said, there's no predeclaration syntax in Euphoria, so
you're stuck with using call_proc(), call_func(), and routine_id(). This
doesn;t mean you can't make your code look tidy though:

procedure a()
    puts(1,"a()\n")
end procedure

integer c_id -- predeclare c()

procedure b()
    a()
    call_proc(c_id,{})
end procedure

procedure c()
    puts(1,"c()\n")
end procedure
c_id = routine_id("c") -- assign something to the predeclaration var

You can even have consitent style if you don't mind being a bit more
verbose:

type predeclare(object sub) -- Using a verb as a type name :)
    if  integer(sub)
    and sub >= 0
    then   return 1
    end if return 0
end type

predeclare a_ -- Using underscores to distinguish the routine_id from
predeclare c_ -- the actual routine name. Eu will complain otherwise.

procedure a()
    puts(1,"a()\n")
end procedure
a_ = routine_id("a")

procedure b()
    call_proc(a_,{}) -- Consistent syntax in b()
    call_proc(c_,{})
end procedure

procedure c()
    puts(1,"c()\n")
end procedure
c_ = routine_id("c")

HTH,
Carl

7. Re: procedures calling procedure calling procedure...

• Posted by petelomax at blueyonder.co.uk Oct 15, 2002
• 459 views

{{{ On Wed, 16 Oct 2002 07:28:55 +1000, Derek Parnell <ddparnell at bigpond.com> wrote:

I've bit my tongue on this, apparently too long.

Just what, exactly, is *SO* hard about

global integer closeCompareRID ... call_proc(closeCompareRID, {) ... procedure closeCOMPARE() closeCompareRID routine_id("closeCOMPARE")

over & above:

closeCOMPARE() ... procedure closeCOMPARE()

????? Yeah, two extra lines and some syntax, SO? Just how often do you actually use forward references anyway?

I, for one, possibly lone voice, appreciate that the former clearly signals to me, the human reader, that closeCOMPARE() has not yet been defined; I haven't missed it; it may not even be in this source. (which is true anyway for global routines of include files)

Derek, I am frankly amazed that you bear such a grudge on this issue as surely if this had been in the language from the get-go then routine_id() would not exist and your work on win32lib would basically be nigh impossible?!

Pete

PS Rob, routine_id("rubbish") currently returns -1 (assuming of course that no procedure or function rubbish has been defined).

What chances of an error/warning message instead?

I feel a minor spelling mistake can cause major headaches in this case.

9. Re: procedures calling procedure calling procedure...

• Posted by Robert Craig <rds at RapidEuphoria.com> Oct 15, 2002
• 450 views

Derek Parnell writes:

> For example, even though I actively campaign 
> against the use of GOTO, I would allow it in a language 
> so that people could choose NOT to use it. Of
> course many would use it, but that is a problem of their own making.

I'd rather live in a world in which cigarettes
don't exist, than a world where they are introduced
and then actively campaigned against. 

With GOTO's in the language, you'd have to be
more careful when reading someone else's code.
An apparently familiar control-flow structure,
could in fact be something hideous. To be sure,
you'd have to scan every line of code looking for GOTO's.

> Robert, why is forward referencing not included in Euphoria?

Define-it-before-you-use-it is slightly easier to implement.
Otherwise, you might have to fill in dummy information and 
"back patch" it when you find the real information.
But this is a relatively minor issue.

At first (many years ago, pre-1.0) I was prepared to 
allow define-it-anywhere, and I had a workable design 
for it, but then I started to think about a couple of languages
I had used previously that imposed define-it-before-you-use-it, 
and I remembered that after a bit of initial frustration,
I grew to like the imposed, machine-checked discipline.

I'm happy the way Euphoria is, in this regard.

I like the fact that when I'm scrolling around in an
editor, I know which direction to move in to find 
the definition of a routine. I like the fact that I can pick
up anyone's program, and see the structure, at least
in terms of the ordering of high-level versus low-level
routines. I can also be assured that the routines in
a file included at the top of the main file, are used
by the main file, but cannot call routines in the main file.
(except via obvious uses of call_func/call_proc, and only
by permission of the author of the main file who has to
pass in a routine_id)
This gives me some confidence in understanding the 
overall structure of the program.

If people could lay out their routines in random order
you'd lose all this. Of course people would arrange their
routines in some way that means something to them,
but probably not to anyone else, and their arrangement
would not be machine verified, so it would not be trustworthy.

Also, over the years I've added code that assumes
define-it-before-you-use-it. I can think of
various optimizations in the interpreter and the translator.

Regards,
   Rob Craig
   Rapid Deployment Software
   http://www.RapidEuphoria.com

11. Re: procedures calling procedure calling procedure...

• Posted by Derek Parnell <ddparnell at bigpond.com> Oct 15, 2002
• 457 views

Thanks Robert.

>> Robert, why is forward referencing not included in Euphoria?
>
>Define-it-before-you-use-it is ...[BIG SNIP]
---------
Cheers,
Derek Parnell