Yeah but int x, y, z would do the same and everyone cam read it

why not int *a, *b ? you know its pointer and would check for NULL for example. also you know this is first rank

Both are fine to me. The advantages I see in having the typedef - no useless discussions on where to put the * - can't "forget" the * - can't have multiple variables of different types in the same line

goto is widely used(check linux kernel source). however its better to avoid, cause hard to read code. You should jump while looking through a code. For several variables in one line no strong arguments. I read lots of book, articles - only today I`ve got this knowledge from you

that`s what I meant. Pavel wrote each var should be declared on separate line

1.weak argument..2. if you forget * most likely you would get compilation error. 3. if its the problem(however I don`t understand why) just don`t do it.

1. yes and no. Having consistency is not a wear argument (imo) and by not having a typedef you either need to manually validate every change for consistency or you need to setup some infrastructure that automates that part. Using a typedef provides this consistency for "free" 2. I still think you should prevent mistakes and errors as soon as possible.

1. you achieve this consistency by adding new entity (typedef) which has drawbacks. So more errors and harder to read a code

Is turbo c+ a good app for coding C language?

Is turbo c+ a good app for coding C language?

I think no, cause its out of the C standard

if I read int *a I know how to work with. If I face with pint a, I should find in source what is pint. And I could deal with it like non ptr type(cause I don`t what is pint).

Then suggest me a better application :)

Visual Studio or Visual Studio Code + GCC

this is true for every type, so you could only use primitive types in C. Decent editors or IDEs will be able to help you. Additionally this is a one-off cost, whereas you would need to pay attention for the placement every time you use it.

1. there is special treatment for ptr type. for example we usually validate it for NULL if its function argument. 2. there are no other options for struct, or union. in this cause you can simply use int *a. If you have to pass pint to function would you use pint or *pint? you should check what is pint first. If you have int* and don`t need to change address you just you *int

it is still a "ptr type", so you can still check it for NULL #include <stdlib.h> typedef int* pint; void Foo(pint a) { if (a == NULL) { return; } // ... }

lets work with huge code source. We have hundreds of pint, pfloat, pdouble, pcustomtype and so on. some of them are pointers, some of them not like parent . do we have arent type? also pint could be defined in third party lib, so you have multiple definition. So now you should find out what is pint in source jungles. for hello world project almost any decision is ok

I do work with a huge code basis (> 10 million of lines) and it works great

Actually code readability is usually the most important today. More code - less readability. Decrease it, to avoid potential discussions int *p vs int* p - doubtful idea

1. there is special treatment for ptr type. for example we usually validate it for NULL if its function argument. 2. there are no other options for struct, or union. in this cause you can simply use int *a. If you have to pass pint to function would you use pint or *pint? you should check what is pint first. If you have int* and don`t need to change address you just you *int

On my job we too have typedef to pointer type to make it easier to use. It's more readable than using *. Proper IDE shows what this type actually is. And you don't even need to know that most of the time anyway. Method gets pint - I'm sending pint.

how do you know how to pass pint to method: by val or ref? how do you know if to check pint inside method with nullptr?

Whatever this method expects. It will not compile otherwise.

what about creating your own method?

What about it? IDE shows me definition if I really need to know. If type is common I know already what type it is. How should that affect my decisions anyway?

you need IDE with indexing, you should relay on it, you should check what is pint. what if mulpiple definitions of pint? I would simply use int*, less entities - easier to maintain code

you need IDE with indexing, you should relay on it, you should check what is pint. what if mulpiple definitions of pint? I would simply use int*, less entities - easier to maintain code

you need IDE with indexing, you should relay on it, you should check what is pint. what if mulpiple definitions of pint? I would simply use int*, less entities - easier to maintain code

Do think the same about CI, clang-format, linters, sanitisers etc.? "Who needs tools when you can do this manually?"

no. these tools help to maintain code. the only drawback - you should devote more time to write in the right way.

#include <stdlib.h> int main() { int n,count=0; scanf("%d",n); while(n>0) { count=n/10; count++; } printf("count %d",count); }

Hey guys, working on a long code so I am using multiple files. How do I include those files as headers without the use of file directory in let's say main.cpp

#include <stdlib.h> int main() { int n,count=0; scanf("%d",n); while(n>0) { count=n/10; count++; } printf("count %d",count); }

You don't modify n inside your loop body?

The loop does not work

Hey guys, working on a long code so I am using multiple files. How do I include those files as headers without the use of file directory in let's say main.cpp

Every compiler has switch to add search path. For example https://www.rapidtables.com/code/linux/gcc/gcc-i.html

What's the solution?

What's the solution?

Solution is to fix your code so it does what you want it to do. I don't know what is expected behavior in this case.

The question is count number of digits in a number

Replace count=n/10; With n=n/10;

#include <stdlib.h> #include<stdio.h> int main() { int n,count=0,rem; scanf("%d",&n); while(n>0) { rem=n%10; n=n/10; count++; } printf("count %d",count); }

It does not change

What does not change? Why do you think people can read your mind?

How do I fix a multiple definition of function error message?

I only used one definition; calling functions defined in non main cpp file and some how getting that error message

leave one definition, use namespace

Was it defined in header by any chance?

Multiple files here, can I send them to use privately?

Yes

no. send here

https://pastebin.pl/view/2642c9d3

You're doing it wrong. You're not supposed to include cpp. You should include headers to cpp,

Please how do you mean?

Ok, reading them

1) We have a legacy C library that is indispensable in a project. The code content of this library is not to be modified, but the library offers the ability to register callbacks that are preferably coded in C++ by the developer utilizing the library. How can such a callback mechanism be possible to construct, and what is an important detail concerning the C library and the signatures of the registered callback functions? Explain. 2) Similar to the C library in the previous question, a C++ library allows callback functions to be registered for utilization as well. Would you expect the C++ library to make use of the same callback mechanism as suggested in the previous question, or are there better alternatives that the C++ library possibly can offer? Explain.