30 years after my first exposure to C, I'm finally doing it.

A bit of a monologue. I've always been fascinated by my experience that those that love assembler programming, dislike C, and vice versa.

I'm programmed since 1977 in HS and have worked in 4 different hardware (assembler) platforms (and a dozen languges). Every instruction set I've run across has been pretty intuitive (6502 the best of micros and the Four-Phase 32-bit mini the best instruction set ever). That is, until I looked at my first assembly listed for my first C program. What an uuuuuuuuuuuuuuuuugly instruction set. It's almost like a C-language designer designed the instruction set. However, I will say I'm vastly impressed by the quality of code generated (quite impressed actually).

Anyhoo, when I saw C in college, I couldn't understand how anyone would design a language like that when they can design one from scratch.

But, here I am. I wrote my first .bat script ever a month ago (2700 lines) and thought I'd be able to make it shorter by using C. Wrongo.


Questions:


It seems char* whatever and char *whatever are identical?

When you have a pointer, are you allowed to increment it by one and keep using it? It seems the compiler will "know" it's a pointer and increment it by 4 (i.e. it's a pointer to a 4-byte INT) for you.

I see a lot of "while (joevariable)" of "if (joevariable)' but it doesn't have an == or != condition of any kind. It seems that that's the same as joevariable != 0)?

Unicode. Looks like it's simply 2-bytes per "ascii" character. I played around with printf to see if I could get a handle on what occurs and interestingly, it's like incrementing the subscript for the string increments by 2 if wchar vs. 1 for char. Is it a good habit to do things a certain way to allow unicode stuff????

Yes, the placement of the pointer symbol doesn't matter. The only issue is when you get a situation such as this:

Code:

---

char *e, f;

---

, which could be confusing (as F isn't a pointer).

The pointer increment will take into account the size of the type, and automatically increment by that much.

You're pretty much right, but just to clarify:

Code:

---

while (i)            /* basic */
while (i != 0)    /* literal translation, but not necessary */

printf("%d", (i == 0));  /* will be either 1 or 0, depending if "i" is zero or not */
printf("%d", (i != 3));  /* will be either 1 (i is not 3) or 0 (i is 3) */

/* ^ shows that "while" and "if" operate with boolean logic, which is why it works */

while ( (i != 0) != 1) /* now thats just silly, but it works */

---

As for Unicode, I'm not really sure, but these might help:
> wchar.h man page
> wprintf(3) - Linux man page

Quote:

---

Originally Posted by jlewand

A bit of a monologue. I've always been fascinated by my experience that those that love assembler programming, dislike C, and vice versa.

---

Why would it have to be one or the other? There's absolutely room for both. It's not really about what one likes, it's about using the right tool for the job. These days assembly is far more often than not, the wrong tool for the job, but I'll use it if I need to.

Quote:

---

When you have a pointer, are you allowed to increment it by one and keep using it? It seems the compiler will "know" it's a pointer and increment it by 4 (i.e. it's a pointer to a 4-byte INT) for you.

---

Yes it will. It's an abstraction. You don't have to know how big it is, it you want to move 3 items over then you just add 3.
You're protected from the lower level details that the memory address has say 3*4=12 added to it in such a case. It's simply not as productive or flexible to have to deal with manually incrementing by a multiple of specific size. Heaven forbid you get it wrong and access the 2.5th item in an array and cause a bus error or slower misaligned read.

Quote:

---

I see a lot of "while (joevariable)" of "if (joevariable)' but it doesn't have an == or != condition of any kind. It seems that that's the same as joevariable != 0)?

---

Yes and it's largely just personal style. It's just like how in assembly you can use a branch-if-zero instruction or you can fill a register with zero and use a compare instruction then branch based on the flags that the comparison instruction set. The difference being though that a typical C compiler is smart enough to turn an if (joevariable != 0) into a simple branch-if-zero instruction as well.

Quote:

---

Unicode. Looks like it's simply 2-bytes per "ascii" character. I played around with printf to see if I could get a handle on what occurs and interestingly, it's like incrementing the subscript for the string increments by 2 if wchar vs. 1 for char. Is it a good habit to do things a certain way to allow unicode stuff????

---

In many circumstances yes, but not necessarily. It probably makes no sense to try and cater for either in an assembly language program.

1. Yes it's identical, the positon of the indirection symbol doesn't matter, as long as it isn't at the back of the variable.
2. This depends, there are 2 kinds of pointers: constant pointers and variable pointers, the name of an array for example is a constan pointer. You can't increment these, variable pointers you can increment.
3. This are so called booleans, since there isn't a real boolean-type in C, we use integers that can have the value 1 or 0. 1 stands for TRUE, 0 for FALSE, how that if (1) works, you can figure out yourself.
4. I'm not sure of this, all I can say is that a char (ASCII) variable is 1 byte.

Quote:

---

Originally Posted by kevinstrijbos

the name of an array for example is a constan pointer.

---

Bad example because it is not true :p

Uh? It's true... Well yeah, if you don't think to deep... The name of an array is the name of an adress, much like a pointer...

Complete your understanding here -> Arrays and Pointers