In article 131293140631@ese-31.sys.uea.ac.uk, cjp@sys.uea.ac.uk (C. Jeremy Pye) writes:
> I have run code on several compilers, and the following
> inconsitent error appears for the various compilers.
> Please reason this out for me ..
>
> function() {
>
>   for (int j=0;j<10;j++)
>     for (int i=0;i<10;i++) {
>       do something ...
>     }
>
> Is 'i' declared/in scope at this point ?
> Simply, can I write
>
>   for (int i=0;i<10;i++)
>     do something ..
>
> or should it be
>
>   for (i=0;i<10;i++)
>     do something ..
>
>   return ...
> }
>
> I can't find/understand anything I have read
> in Annotated ref ... is it clear.
>
> please reply by email. Thanks in advance.
> .........................................................
> C. Jeremy Pye
> Nonlinear Image Processing Group   Tel: +44 (603) 593220
> School Of Information Systems      Fax: +44 (603) 507720
> UEA, Norwich, NR4 7TJ, UK       Email: cjp@sys.uea.ac.uk


You cannot do :
 for (int j = 0; j<10; j++)
  for (int i = 0; i<10; i++)
   do something...

as when j = 1, i.e. the second cycle of the outer loop, i has already been
declared, so there is a second declaration.

the way to cure this is to use teh second method.  i.e

int i,j;

 for (j = 0; j<10; j++)
  for (i = 0; i<10; i++)
   do something...

Hope this helps.

Steve.

--
A wasted youth is better by far than | The views expressed are my own and
a wise and productive old age.       | are not always correct.
                                     |
       Meatloaf - Bat out of Hell II | email : sjg@cs.bham.ac.uk

cjp@sys.uea.ac.uk (C. Jeremy Pye) writes:

>I have run code on several compilers, and the following
>inconsitent error appears for the various compilers.
>Please reason this out for me ..
>
>function() {
>
>  for (int j=0;j<10;j++)
>    for (int i=0;i<10;i++) {
>      do something ...
>    }
>
>Is 'i' declared/in scope at this point ?

No.  The outermost `for' introduces an implicit scope.  It's as if
you had written the following:
   for (int j=0;j<10;j++)
     { // start implicit scope
       for (int i=0;i<10;i++) {
         do something ...
       }
     } // end implicit scope
But this is a relatively recent decision by the ANSI/ISO committee.
See ARM chapter 19, section 6.4.  (If your ARM doesn't list it,
try a more recent one - mine is "Reprinted with corrections, May 1992.")
Some compilers may not implement the new rules yet.

--
Fergus Henderson                     fjh@munta.cs.mu.OZ.AU

In article <CI0qz3.qF@cs.bham.ac.uk> S.J.Gallacher-SE2@cs.bham.ac.uk
(Steven J Gallacher) writes:

|> In article 131293140631@ese-31.sys.uea.ac.uk, cjp@sys.uea.ac.uk (C. Jeremy Pye) writes:
|> > I have run code on several compilers, and the following
|> > inconsitent error appears for the various compilers.
|> > Please reason this out for me ..

|> > function() {

|> >   for (int j=0;j<10;j++)
|> >     for (int i=0;i<10;i++) {
|> >       do something ...
|> >     }

|> > Is 'i' declared/in scope at this point ?
|> > Simply, can I write

|> >   for (int i=0;i<10;i++)
|> >     do something ..

|> > or should it be

|> >   for (i=0;i<10;i++)
|> >     do something ..

|> >   return ...
|> > }

|> > I can't find/understand anything I have read
|> > in Annotated ref ... is it clear.

I don't think that this is in the ARM (but I may be wrong), but there
is a generally accepted rule which has been adapted by the standards
committee: the scope of the inner declaration terminates at the end of
the enclosing 'for'.  More generally, *all* declarations controlled by
a conditional expression (while, the second part of the for, if) have
a scope which ends at the end of the condition, thus:

 if ( x )
     for ( int i = ...
or
 while ( x )
     for ( int i = ...

In both cases, the 'i' goes out of scope at the end of the enclosing
if/while statement.

Note that of the compilers I have tried, none of those which did not
already implement this rule did anything that could possibly be
considered legal in the presence of a type with constructors and
destructors.

|> You cannot do :
|>  for (int j = 0; j<10; j++)
|>   for (int i = 0; i<10; i++)
|>    do something...

|> as when j = 1, i.e. the second cycle of the outer loop, i has already been
|> declared, so there is a second declaration.

As pointed out above, this is not correct.  However, the reason given
is precisely the rationale for the special rule.  Try changing the
type of 'i' to a type which has both a constructor and a destructor,
and see when each would otherwise get called.

|> the way to cure this is to use teh second method.  i.e

|> int i,j;

|>  for (j = 0; j<10; j++)
|>   for (i = 0; i<10; i++)
|>    do something...

This is of course legal, but goes against the general rule of always
initializing in the declaration.
--
James Kanze                             email: kanze@us-es.sel.de
GABI Software, Sarl., 8 rue du Faisan, F-67000 Strasbourg, France
Conseils en informatique industrielle --
                   -- Beratung in industrieller Datenverarbeitung

cjp@sys.uea.ac.uk (C. Jeremy Pye) writes:
>
>I have run code on several compilers, and the following
>inconsitent error appears for the various compilers.
>Please reason this out for me ..
>
>function() {
>
>  for (int j=0;j<10;j++)
>    for (int i=0;i<10;i++) {
>      do something ...
>    }
>
>Is 'i' declared/in scope at this point ?
>
No, but 'j' is. The outer loop introduces a new scope for its body, even
though there are no braces.

>I can't find/understand anything I have read
>in Annotated ref ... is it clear.
>
Look at the for/while equivalence rule at the top of ARM 6.5.3. This says
that the above code is equivalent to:

 int j=0;
 while(j<10) {
  int i=0;
  while(i<10) {
   do something ...
   i++;
  }
  j++
 }

All of the above braces are a result of the re-writing rule, and make it
clear that a new scope is introduced for the loop body.

Steve.
--
UUCP:  ...!uknet!warwick!steve Internet: steve@maths.warwick.ac.uk
JANET:  steve@uk.ac.warwick.maths PHONE:  +44 203 524657

In article <cjp-131293140631@ese-31.sys.uea.ac.uk> cjp@sys.uea.ac.uk (C. Jeremy Pye) writes:
>I have run code on several compilers, and the following
>inconsitent error appears for the various compilers.
>Please reason this out for me ..

 From what?
>
>function() {
>
>  for (int j=0;j<10;j++)
>    for (int i=0;i<10;i++) {
>      do something ...
>    }
>
>Is 'i' declared/in scope at this point ?

 As it currently stands (in at least one place in the Working Paper),
the condition and body of a for statement are nested
(in an implied inner scope), whereas the for-init
statement is not. Thus, i is not in scope but j is.
(another place was not updated correctly, and should reflect the
above rules)

 There is a proposal to change this so that
the j is not in scope where you ask either.

 Dont ask whether this is legal:

 if(int i=1){int i=1; i++;}

[technically, it is, as far as I can determine, but its
subject to interpretation and may not have been the
committee's intent.]


--
        JOHN (MAX) SKALLER,         INTERNET:maxtal@suphys.physics.su.oz.au
 Maxtal Pty Ltd,      CSERVE:10236.1703
        6 MacKay St ASHFIELD,     Mem: SA IT/9/22,SC22/WG21
        NSW 2131, AUSTRALIA

I have run code on several compilers, and the following
inconsitent error appears for the various compilers.
Please reason this out for me ..

function() {

  for (int j=0;j<10;j++)
    for (int i=0;i<10;i++) {
      do something ...
    }

Is 'i' declared/in scope at this point ?
Simply, can I write

  for (int i=0;i<10;i++)
    do something ..

or should it be

  for (i=0;i<10;i++)
    do something ..

  return ...
}

I can't find/understand anything I have read
in Annotated ref ... is it clear.

please reply by email. Thanks in advance.
........................................................
C. Jeremy Pye
Nonlinear Image Processing Group   Tel: +44 (603) 593220
School Of Information Systems      Fax: +44 (603) 507720
UEA, Norwich, NR4 7TJ, UK       Email: cjp@sys.uea.ac.uk