Urukhayy2015-03-13 11:52:06
Urukhayy, 2015-03-13 11:52:06

A couple of questions. Is it worth it to use functions like this? What's the best way to terminate a function?

1) Let's say there are 3 sections in the code that require the same code, consisting of 5-10 lines. Moreover, the result of executing this code is a number.
What is better, copy the same fragment to 3 places, or create a function and return a number? What if there are a lot of such functions?
2) Is it possible to complete the execution of a function with returns?

if(a == b) return print("test");
// код, если a != b
return 1;

// vs

if(a != b)
// действие
else print("test");

And if there are many such investments?

Answer the question

In order to leave comments, you need to log in

4 answer(s)
Armenian Radio, 2015-03-13

1) It's better to create a function. Even better is to study the architecture of the problem well.
2) return - regular way to exit the function. This is slightly better than kilometers of nested conditions.

Anton Mudrenok, 2015-03-13

I advise you to ask a question more specifically.
1. If you adhere to the OOP paradigm, then copying the text is in any case worthless. Especially if the result of the function does not depend on class variables. Write a static class, implement the method you need there, and call it anywhere without creating instances.
2. Also not entirely clear. In the first case, your function returns a function, in the second, nothing is returned to you, but just a set of instructions. It is logical to compare these 2 options:

public string Func(int a, int b) {
   if (a == b) return "test";
   return "error";

public string Func2(int a, int b) {
   string text = (a == b) ? text = "test" : text = "eroor";
   return text;

I am for option 2, because you don’t need to look for the code where the exit from the function was, but this is a matter of taste

Igor Kalashnikov, 2015-03-13

If you want to get "self-documenting" code, move complex and non-obvious checks into functions.

function checkUser(user) {
    return (user != null && user.name != null && user.role == 'admin')

function checkSetupConfig(params) {
    var isValid = false;
    if (checkUser(params.user) || params.installationPath != null) {
        isValid = true;
    /// тут могут быть еще проверки ... 
    return isValid;

function setup(params)
    if (checkSetupConfig(params)) {

write 1 return at the end, or place them inside if-s too - pure taste and depends on the command.
For example, when I was on a project with a huge amount of legacy, and there were class methods for 100-300 lines (which is bad design, but still), then 1 return at the end made assessment easier. You can immediately look down and understand what is thrown out of the method.
This is especially true for untyped languages ​​like JS. When in different places, return can return anything, a number, an object with an error, a string.
And yet, returning the completion code is a bad form, if in your code "1" --- carries some semantic meaning, then it's better to make an enum, or collect an object like:
var validationResult = {
    isValid : fale,
    errorMessage :  'User not logged in',
    // errorCode :  'UE001',  // лучше коды делать строками, и читабельными, но можно и 1
    errorCode : 1

Mrrl, 2015-03-13

If a and b are global, then it's better to make one global function that will look at them:
and put it next to the variables. Then drag and drop other appeals to them there. You look, and a class with some meaningful properties will be assembled.
return is a good thing, but if you return a value, then it should make sense. For the sake of reducing a pair of parentheses, it is not worth inventing a return type. So that

if(a == b){
// код, если a != b

Didn't find what you were looking for?

Ask your question

Ask a Question

731 491 924 answers to any question