Here we go again: when to use Malloc and when not?

**django** · 08-08-2011

Are you serious, const is in there.

**tabstop** · 08-08-2011

Then either they forgot to change that example from really-old-C to old-C or they're just toying with you (or they mention in all the words around the example that they're doing something not-quite-right for illustrative purposes). I don't have the book so I don't know which.

**breimer273** · 08-08-2011

In my experience, using gcc, a statement like

Code:

char *p = "this is a string";

was always acceptable. Trying to do anything with this string, however, is nearly impossible and I can't think of a situation where having a string literal would be very useful in a program, except for using for comparison. Dynamically allocating space for a string and then using strcpy was always taught to me as a more standard practice.

**django** · 08-08-2011

Hm, kinda stupid, the answer is just there:
"..pmessage is a pointer, initialized to point to a string constant; the pointer may subsequently be modified to point elsewhere..." (same page, my underlining)

**AndrewHunter** · 08-08-2011

There is no need for discussion on this topic really:

Originally Posted by C-99 6.7.8.32

EXAMPLE 8 The declaration

Code:

char s[] = "abc", t[3] = "abc";

defines ‘‘plain’’ char array objects s and t whose elements are initialized with character string literals. This declaration is identical to

Code:

char s[] = { 'a', 'b', 'c', '\0' },
t[] = { 'a', 'b', 'c' };

The contents of the arrays are modifiable. On the other hand, the declaration

Code:

char *p = "abc";

defines p with type ‘‘pointer to char’’ and initializes it to point to an object with type ‘‘array of char’’ with length 4 whose elements are initialized with a character string literal. If an attempt is made to use p to modify the contents of the array, the behavior is undefined.

**AndrewHunter** · 08-08-2011

Originally Posted by django

Hm, kinda stupid, the answer is just there:
"..pmessage is a pointer, initialized to point to a string constant; the pointer may subsequently be modified to point elsewhere..." (same page, my underlining)

I don't think anyone told you that you couldn't do it. In fact I specifically remember telling you that you could.

**django** · 08-08-2011

With all due respect Andrew, you are missing the point. My question is not about changing the string literal (which causes undefined behaviour) but changing where the pointer points to. Anyway, as far as I am concerned,t he matter is resolved. I would like to know however, which of my code samples would have caused your compiler to warn.

**django** · 08-08-2011

Ok, all cleared up. I will reread this post carefully tomorrow and learn. Thanks again for the article about pointers, it seems I need to reread that too

**TheBigH** · 08-08-2011

Originally Posted by breimer273

I can't think of a situation where having a string literal would be very useful in a program, except for using for comparison.

Well, there's obviously the first argument to printf- that's a string literal.

**quzah** · 08-08-2011

Originally Posted by AndrewHunter

You cannot modify string literals, as you have been told over and over again.

Originally Posted by CommonTater

That's why I asked about his compiler... It should be complaining it's butt off...

Adak used to argue that Turbo C let him modify string literals. I don't know if it warned or not.

Quzah.

**_Mike** · 08-08-2011

Originally Posted by TheBigH

Well, there's obviously the first argument to printf- that's a string literal.

It could be. Doesn't have to be though.

**quzah** · 08-08-2011

Originally Posted by _Mike

It could be. Doesn't have to be though.

I was trying to find a post I made a long time ago that modified the first argument to printf to print some stars or something, but I can't.

It doesn't help that the search here sucks.

Edit: HAH! help with nested loops

Quzah.

**~~CommonTater~~** · 08-08-2011

Originally Posted by django

Could you point out a line in my sample code were your compiler would complain it's bud off?

Code:

char *test = "This is a test";
strcpy(test,"And now it's over");

That is modifying a string literal and you can't do that. The reason you can't do that is that test is pointing to the read only section of the program's PE Image.

**Matticus** · 08-08-2011

Originally Posted by django

LCC compiler for windows

For the record, I used to run an old version of LCC because it was all I knew at the time. It worked fine for years, but there are definitely some major issues and bugs I've come across. The most frustrating is how it seems to default the character data type to either unsigned or signed, depending on where the characters are declared (it appears to be unsigned by default in "main()" but signed by default in a custom function).

It works well for learning, but I'd recommend updating to something nicer. Though I haven't actually seen anyone here reference LCC, so I'd be interested in hearing if there were other opinions on it.

**~~CommonTater~~** · 08-08-2011

Originally Posted by Matticus

For the record, I used to run an old version of LCC because it was all I knew at the time. It worked fine for years, but there are definitely some major issues and bugs I've come across. The most frustrating is how it seems to default the character data type to either unsigned or signed, depending on where the characters are declared (it appears to be unsigned by default in "main()" but signed by default in a custom function).

It works well for learning, but I'd recommend updating to something nicer. Though I haven't actually seen anyone here reference LCC, so I'd be interested in hearing if there were other opinions on it.

Well, FWIW... Pelles C is a "much modified" derivative of LCC that seems to have overcome many of the original shortcomings... Might make a good update for our OP friend...