Simple String Copy Question

[Canadian0469]

Hi guys,

Okay, I'm a 2nd year computer science student and I'm taking my first C++ course. I have some background in program and have used C++ for graphics programming but I'm reading the first few pages of my profs notes and am already stumped!

Here's the code snippet that's causing me the distress:

int i = 0;
do
{
s[i] = t[i];
i = i + 1;
} while (s[i] != '\0');

The purpose of this code is to copy a string. I understand that \0 is an indicator of the end of a string in C++ but the condition of the while loop is check after incrementing i. So, from what I can tell, isn't s[i] referring to a still untouched index of the array?

Thanks,

Canadian0469

[master5001]

int i = 0;
do
{
  s[i] = t[i];
  i = i + 1;
} while (s[i] != '\0');

Means

Starting with a number i who has the value of 0
Copy the i element of the array "t" to the i element of the array "s"
Increment i by one
Until the i element of "s" is equal to zero.

Do I see any issues with this code? Sure, what if t is not large enough to hold all of the s values inside of it? And no, keep in mind that your code will do all of that stuff then execute the while condition and check to see what is going on. A loop designed like this even copies over the '\0' (null terminator).

[cpjust]

Yuk!
Looks a little too dangerous for my taste.
I'd probably do something more like this:

int i = 0;
do
{
    s[i] = t[i];
} while ( s[i++] != '\0' );

[master5001]

Which still assumes a lot of "t." What specifically is the assignment here, Canadian0469?

[Daved]

That code assumes that s is not an empty string (among other things). There's no reason to assume that when cpjust's version is equally effective and doesn't make the assumption.

It also assumes that t is at least as large as s, which is probably acceptable for the exercise.

>> I understand that \0 is an indicator of the end of a string in C++ <<

Actually, it indicates the end of a string in C. It's just that you can also use C style strings in C++ and some out-of-date books/tutorials/instructors still teach them. (Note that in this case the point of the code is to look at the algorithm, so it's not a problem to be using C style strings for this exercise).

[Canadian0469]

Alright I'm not looking at alternative ways to write this piece of code, I'm only trying to understand the example.

Let me explain what is confusing me. I understand that it's trying to copy the string stored in array t to array s. Look at how I would trace the example where array t holds the string "a\0", i.e. t[0]='a' and t[1]='\0':

At first we have:
i=0
s[0]=0
s[1]=0
t[0]='a'
t[1]=\0

Then, first iteration of loop:
s[0]='a'
s[1]=0
i=1
checks is s[1] != '\0'? Yes, it's still uninitialized so continue...

Then, second iteration of loop:
s[1]=\0
i=2
checks is s[2] != '\0'? Yes, it's still unitialized so continue...

This is the part where I get confused. I thought at this point it should be seeing that s[1] has been changed to \0, instead, because i was incremented in the loop, it doesn't even see this change as its always checking the index of the array that is 1 index following the one that just got changed... Am I wrong?

[cpjust]

It also assumes that t is at least as large as s, which is probably acceptable for the exercise.

Don't you mean that s is at least as big as t?

The way the teacher wrote it, the size of t would need to be at least strlen( t ) + 2, since it's always copying 1 past the NUL terminator; otherwise it could be accessing memory on a new page boundary that it doesn't own. And obviously s would also need to be at least that size.

EDIT: Oh and there's another bug. Since the teacher's while loop increments i and then checks if s[i] != '\0', what if your s string starts out as "a\0" and t is "xyz\0"? After copying x to s[0], it will check s[1] and see that it's already '\0' and not bother copying the rest of t.
Maybe you should tell your teacher about those bugs in the code.

[cpjust]

This is the part where I get confused. I thought at this point it should be seeing that s[1] has been changed to \0, instead, because i was incremented in the loop, it doesn't even see this change as its always checking the index of the array that is 1 index following the one that just got changed... Am I wrong?

Yup, that's the big problem with that code.

[Canadian0469]

What a great way to start the semester... That was seriously my first example.

Anyway, thanks for the help guys! I just wanted to make sure I was tracing it correctly and that it was actually an error.

Cheers!

Canadian0469

[cpjust]

What a great way to start the semester... That was seriously my first example.

Anyway, thanks for the help guys! I just wanted to make sure I was tracing it correctly and that it was actually an error.

Cheers!

Canadian0469

Hopefully the teacher isn't doing anything else dumb like void main()...

[Daved]

>> Don't you mean that s is at least as big as t?
Yes, I missed the most important thing - that s was the target of the copy and t was the source.

[CornedBee]

that s was the target of the copy and t was the source.

Oh dear!

I do hope the example was given to you as an error search example, and not something serious. There's so much wrong with it ...