Use of nested loops.

[Neptune]

Neptune, in many cases you use multiple nested loops without even knowing
it.

Let's look at Keith's a modified example. Open a text file and read and
print
each line. You need a loop in which to read a line, test you have
successfully
read it and then print it. Inside that loop, you would probably use fgets
for
reading and puts or printf for printing. Each of these functions contains at
least one loop, possibly more than one. So, you may get nested loops simply
by
calling a function in a loop. Isn't programming fun

It is fun - that's the attraction (once I get my head around some basic
constructs that is ). I like your example Peter - it makes sense and
touches on something that I will be coming to in due course, but was
already curious about: readin in text files, so this will pre-'arm' me
so to speak.

To answer your original question, nested loops are used whenever an
algorithm
asks for it. Processing more than one-dimensional entities is the most
common
example. My example is one of them: the line number could be considered the
vertical and the position of a character in a line the horizontal dimension.

It appears then that (theoretically speaking anyway), there are no
limits to the number of loops one can nest? I cannot think of any
example to illustrate this, but was wondering if one can nest one or two
or five loops, is there a maximum to which one can go before it
discombobulates entirely. I can imagine that there would be logistical
nightmares (indentation running off of the page or line wrapping
endlessly, trying to keep the iterations straight in one's head, even
finding a value to engage in that kind of exercise, etc), so the query
really is theoretical.

Anyway, thanks for your thoughts.


--
"Today a young man on acid realised that all matter was really energy
condensed to a slow vibration, that we are all one consciousness
experiencing itself subjectively, there's no such thing as death,
life is only a dream, and we're the imaginations of ourselves.
Here's Tom with the weather ..." - Bill Hicks.

 

[Neptune]

Ben Pfaff wrote:

int bars[] = {1, 6, 2, 4, 9};
int i;
for (i = 0; i < sizeof bars / sizeof *bars; i++) {

That was helpful. Gives me an idea of its usage outside of text book
example. I haven't come across "sizeof" before. I presume that it is
some standard way of referencing the length of the array?

sizeof yields the number of bytes in its operand. `sizeof bars'
is the number of bytes in array bars[]; `sizeof *bars' is the
number of bytes in a single element of bars[]. Thus, the
quotient of those two expressions is the number of elements in
bars[]; in this case, 5.

This is, of course, quite correct.

One thing to watch out for is the distinction between arrays and
pointers. They are *not* the same thing (though some people might try
to tell you they are), but there are some contexts in which a pointer
name and an array name can be used in the same way. There are times
when you have to be very careful to know whether you're applying
sizeof to an array object or to a pointer.

Section 6 of the C FAQ at <http://www.eskimo.com/~scs/C-faq/top.html>
covers this well.

Thanks for the link Keith. I've added this link to my list of things to
check into over the next couple of days. Is it true that pointers have a
certain unpopularity in terms of the trouble they pose new students of
C? Is that a justified reputation?

--
"Today a young man on acid realised that all matter was really energy
condensed to a slow vibration, that we are all one consciousness
experiencing itself subjectively, there's no such thing as death,
life is only a dream, and we're the imaginations of ourselves.
Here's Tom with the weather ..." - Bill Hicks.

 

[Neptune]

A simpler more concrete example. Consider an othello game
<http://www.ugateways.com/bof4.html>

You have a board which is made up of 8 by 8 squares. You decide
to represent the state of a squares as an integer.

#define UNOCCUPIED -1
#define BLACK 0
#define WHITE 1

Bonus question, why is it a good (and is it good?) idea to make
black 0 and white 1 instead of any other two numbers?

This is an interesting example Thomas. I'd hazard a guess that one would
use 1 and 0 to represent true/false values, but this is only a guess?

Now you need the board

#define SIZE 8

int board[SIZE][SIZE];

To start of the game you need some simple code
int i;
int j;

for(i = 0;i < SIZE;i++)
{
for(j = 0;j < SIZE;j++)
{
board[j] = UNNOCCUPIED;
}
}

board[3][3] = WHITE;
board[3][4] = BLACK;
board[4][3] = WHITE;
board[4][4] = BLACK;

Ok, You are now ready to start the game. (And that you can do
yourself

Thanks for the vote of confidence Thanks for the example - nice one.

<OT>
I saw your signature. Tool fan? Or "just" Bill Hicks?
IMNSVHO they both rock!
</OT>

Both - and Tool still rocks and Bill died too damn young!!! I'm now
waiting for Tool's next release. As it so happens it was Tool that
steered me in the direction of Hicks's work courtesy of the Aenema album.

--
"Today a young man on acid realised that all matter was really energy
condensed to a slow vibration, that we are all one consciousness
experiencing itself subjectively, there's no such thing as death,
life is only a dream, and we're the imaginations of ourselves.
Here's Tom with the weather ..." - Bill Hicks.

 

[Neptune]

Hello.

I am working my way through Zhang's "Teach yourself C in 24 hrs (2e)"
(Sam's series), and for nested loops, he writes (p116) "It's often
necessary to create a loop even when you are already in a loop." Then he
goes on to portray a contrived example that doesn't tell me under what
conditions a nested loop might be favoured as a solution? i.e. what are
nested loops useful for? What kinds of algorithms are served by nested
loops? etc. Is any of this making sense?

Anyway - thoughts welcomed.

- Andy

One example you may come across in your book is a basic bubble-sort which
will sort an array of elements, e.g.

for (i=0; i<MAX-1; i++)
for (j=0; j<MAX-1-i; j++)
if (ELEMENT[j+1]>ELEMENT[j])
{
/*swap the elements*/
TEMP = ELEMENT[j+1];
ELEMENT[j+1] = ELEMENT[j];
ELEMENT[j] = TEMP;
}

HTH
Allan

Thanks Allan. I have heard about bubble sort but have not yet
encountered it. That pleasure yet awaits me!!! The code you have
listed here, would this be the standard algorithm for the bubble-sort
method, or are there also other ways of tackling this problem?

--
"Today a young man on acid realised that all matter was really energy
condensed to a slow vibration, that we are all one consciousness
experiencing itself subjectively, there's no such thing as death,
life is only a dream, and we're the imaginations of ourselves.
Here's Tom with the weather ..." - Bill Hicks.

 

[Mark Gordon]

Thomas Stegen wrote:


This is an interesting example Thomas. I'd hazard a guess that one
would use 1 and 0 to represent true/false values, but this is only a
guess?

Close. Any non-0 value is true. However, (!0 == 1) && (!1 == 0) allowing
you to pass either to a function then simply do checks and assignment
like

if (cell == yours) {
/* cell is mine */
cell = !yours; /* you don't own it no more */
}

 

[pete]

Hello.

I am working my way through Zhang's "Teach yourself C in 24 hrs (2e)"
(Sam's series), and for nested loops, he writes (p116) "It's often
necessary to create a loop even when you are already in a loop." Then he
goes on to portray a contrived example that doesn't tell me under what
conditions a nested loop might be favoured as a solution? i.e. what are
nested loops useful for? What kinds of algorithms are served by nested
loops? etc. Is any of this making sense?

Anyway - thoughts welcomed.

- Andy

One example you may come across in your book is a basic bubble-sort which
will sort an array of elements, e.g.

for (i=0; i<MAX-1; i++)
for (j=0; j<MAX-1-i; j++)
if (ELEMENT[j+1]>ELEMENT[j])
{
/*swap the elements*/
TEMP = ELEMENT[j+1];
ELEMENT[j+1] = ELEMENT[j];
ELEMENT[j] = TEMP;
}

HTH
Allan

Thanks Allan. I have heard about bubble sort but have not yet
encountered it. That pleasure yet awaits me!!! The code you have
listed here, would this be the standard algorithm for the bubble-sort
method, or are there also other ways of tackling this problem?

It's the simplest form of bubblesort.
A more sophisticated form,
records the location of the last swapped pair,
and uses that information to tighten the loops.
There is also a "cocktail shaker" variation.

But you can only optimize the number of comparrisons made.

All sorting functions which work by swapping pairs of
adjacent out of order elements, will sort any given array order,
with the same number of swaps.

If you modify bubblesort to the point where it does something
else besides swap adjacent out of order elements,
then it wouldn't be bubblesort any more.

 

[Irrwahn Grausewitz]

It appears then that (theoretically speaking anyway), there are no
limits to the number of loops one can nest? I cannot think of any
example to illustrate this, but was wondering if one can nest one or two
or five loops, is there a maximum to which one can go before it
discombobulates entirely. I can imagine that there would be logistical
nightmares (indentation running off of the page or line wrapping
endlessly, trying to keep the iterations straight in one's head, even
finding a value to engage in that kind of exercise, etc), so the query
really is theoretical.

Theoretically it's of course possible to nest loops the deep you like,
but for common programming tasks two or three levels of loop nesting
should suffice. If you are exceeding five levels of nesting you should
rethink your algorithm and/or code structure.
[ I recently wrote a tiny test-suite for an interpreter for a language
with C-like syntax, and in the loop-test part I got a serious headache
at level nine. ]
Note that real-world implementations have limitations on how many levels
of loop nesting are possible, but usually the limits are far beyond what
a sane programmer would actually use.

Regards

 

[pete]

Irrwahn Grausewitz wrote:

Theoretically it's of course possible to nest loops the deep you like,

My copy of the C89 last public draft, has limits for that:

2.2.4.1 Translation limits
* 15 nesting levels of compound statements, iteration control
structures, and selection control structures

But I don't see anything like that, in N869.

 

[Keith Thompson]

Thanks for the link Keith. I've added this link to my list of things
to check into over the next couple of days. Is it true that pointers
have a certain unpopularity in terms of the trouble they pose new
students of C? Is that a justified reputation?

Yes, pointers can be pretty confusing to newbies. It can be far too
easy to get misleading ideas about how pointers work.

Suggestion: Read the FAQ from start to finish. You probably won't
understand a lot of it; that tells you where you need further study.
Later on, read it from start to finish again; it will make a lot more
sense.

Having said that, the FAQ is not intended as a tutorial; it's designed
more to correct misconceptions than to present the underlying ideas in
the first place. You might be better off diving into the FAQ after
you've finished whatever text book or tutorial you're using (though
you should probably read 18.10, "What's a good book for learning C?",
right away.)

 

[Keith Thompson]

It appears then that (theoretically speaking anyway), there are no
limits to the number of loops one can nest? I cannot think of any
example to illustrate this, but was wondering if one can nest one or
two or five loops, is there a maximum to which one can go before it
discombobulates entirely. I can imagine that there would be logistical
nightmares (indentation running off of the page or line wrapping
endlessly, trying to keep the iterations straight in one's head, even
finding a value to engage in that kind of exercise, etc), so the query
really is theoretical.

Quick answer: Yes.

The standard requires an implementation to handle at least 127 nesting
levels of blocks (the actual statement is more complicated than that,
but the details aren't important for this discussion). Realistically,
most compilers don't use fixed-size data structures internally, so the
number of nested loops is going to be limited by the compiler running
out of memory, not by any fixed upper bound.

The compiler will let you have far more levels of nested loops than
you should.

Note that nested loops don't have to be physically nested. It's very
common for a function to be called from within a loop, and for the
function itself to execute a loop, and so on to arbitrarily many
levels. Done properly, this can avoid the discombobulation you're
quite rightly concerned about.

 

[Irrwahn Grausewitz]

My copy of the C89 last public draft, has limits for that:
2.2.4.1 Translation limits
* 15 nesting levels of compound statements, iteration control
structures, and selection control structures

That's the very reason why I wrote:

But I don't see anything like that, in N869.

The closest I can come up with is this:

ISO/IEC 9899:1999TC1
5.2.4.1 Translation limits
[...]
— 127 nesting levels of blocks
[...]

Regards

 

[Peter Nilsson]

My copy of the C89 last public draft, has limits for that:

2.2.4.1 Translation limits
* 15 nesting levels of compound statements, iteration control
structures, and selection control structures

But I don't see anything like that, in N869.

-- 127 nesting levels of blocks

Would seem to be applicable.

 

[pete]

-- 127 nesting levels of blocks

Would seem to be applicable.

Except that loops don't always have blocks.

 

[Glen Herrmannsfeldt]

(snip)

It is fun - that's the attraction (once I get my head around some basic
constructs that is ). I like your example Peter - it makes sense and
touches on something that I will be coming to in due course, but was
already curious about: readin in text files, so this will pre-'arm' me
so to speak.
(snip)

It appears then that (theoretically speaking anyway), there are no
limits to the number of loops one can nest? I cannot think of any
example to illustrate this, but was wondering if one can nest one or two
or five loops, is there a maximum to which one can go before it
discombobulates entirely. I can imagine that there would be logistical
nightmares (indentation running off of the page or line wrapping
endlessly, trying to keep the iterations straight in one's head, even
finding a value to engage in that kind of exercise, etc), so the query
really is theoretical.

Compilers may have stack limits, register limits, or otherwise have a reason
to limit it. The limit should be large enough that ordinary programs don't
run into it.

"Today a young man on acid realised that all matter was really energy
condensed to a slow vibration, that we are all one consciousness
experiencing itself subjectively, there's no such thing as death,
life is only a dream, and we're the imaginations of ourselves.
Here's Tom with the weather ..." - Bill Hicks.

This sounds way too much like superstring theory. Did anyone watch NOVA
last week or this week?

Otherwise, I don't think quantum mechanics excludes it as a possibility.

-- glen

 

[Glen Herrmannsfeldt]

My copy of the C89 last public draft, has limits for that:

2.2.4.1 Translation limits
* 15 nesting levels of compound statements, iteration control
structures, and selection control structures

These are minimum requirements, though the compilers may exceed them.

I believe that I have done more then that. I once had them nested pretty
deep to solve a cross number puzzle. Each loop had a continue statement in
it, so the resulting program ran very fast.

-- glen

 

[Keith Thompson]

Except that loops don't always have blocks.

A block isn't necessarily surrounded by curly braces.

C99 6.8.5 p5:

An iteration statement is a block whose scope is a strict subset
of the scope of its enclosing block. The loop body is also a block
whose scope is a strict subset of the scope of the iteration
statement.

So:

while (1) printf("I will not write infinite loops\n");

actually has two blocks.

 

[pete]

A block isn't necessarily surrounded by curly braces.

C99 6.8.5 p5:

An iteration statement is a block whose scope is a strict subset
of the scope of its enclosing block. The loop body is also a block
whose scope is a strict subset of the scope of the iteration
statement.

So:

while (1) printf("I will not write infinite loops\n");

actually has two blocks.

Thank you.

 

[lawrence.jones]

Except that loops don't always have blocks.

They do in C99.

-Larry Jones

Ha! Wild zontars couldn't drag that information out of me! Do your worst!
-- Calvin

 

[Morris Dovey]

They do in C99.

Hmmm. How about:

#include <stdio.h>
int main(void)
{ int i,j,k;
for (i=0; i<10; i++)
for (j=0; j<10; j++)
for (k=0; k<10; k++)
puts("Where's the blocks?");
return 0;
}

?

 

[Alex]

Hmmm. How about:

#include <stdio.h>
int main(void)
{ int i,j,k;
for (i=0; i<10; i++) <- 1

<- 2

for (j=0; j<10; j++) <- 3

<- 4

for (k=0; k<10; k++) <- 5
puts("Where's the blocks?"); <- 6
return 0;
}

I count 6 block scopes. In C99 you could have declared i, j, and k
within the initialization statement of the for loop. IIRC, the body
of a loop opens a new block scope. This does not necessitate the
use of curly brackets.

Alex