XCode Tutorial Book
XCode Tutorial Book
License
Copyright notice
Copyright 2008 by Bert Altenburg, Alex Clarke and Philippe Mougin. Version 1.15 Released under a Creative Commons License: 3.0 Attribution Non-commercial http://creativecommons.org/licenses/by/3.0/ Attribution: The licensors, Bert Altenburg, Alex Clarke and Philippe Mougin, permit others to copy, modify and distribute the work. In return, the licensees must give the original authors credit. Non-commercial: The licensors permit others to copy, modify and distribute the work and use the work in paid-for and free courses. In return, licensees may not sell the work itself, although it may accompany other work that is sold.
CocoaLab
CocoaLab makes this book freely available online in the form of a wiki or a pdf document in English, Chinese and Arabic, from http://www.cocoalab.com.
Become An Xcoder
Contents
License . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Copyright notice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2 CocoaLab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
How to use this book . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
Become An Xcoder Bugging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Our first Application . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Debugging . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 05:5 05:6 05:7 05:8
Become An Xcoder Pointers again . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . The @ symbol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . A new kind of string . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . NSMutableString . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Exercise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . More pointers again! . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12:1 12:1 12:1 12:2 12:2 12:2 12:4
Table of contents
iii
Become An Xcoder
Introduction
Apple provides you with all the tools you need to create great Cocoa applications, for free. This set of tools, known under the name Xcode, comes with Mac OS X, or you can download it from the developer section on Apple's website. Several good books on programming for the Mac exist, but they assume that you already have some programming experience. This book doesn't. It teaches you the basics of programming, in particular Objective-C programming, using Xcode. After some 5 chapters, you will be able to create a basic program without a Graphical User Interface (GUI). After a few more chapters, you will know how to create simple programs with a GUI. When you have finished this booklet, you will be ready for the above-mentioned more advanced books. You will have to study those too, because there is a lot to learn. For now though, don't worry because this book takes it easy.
Programming is not a simple job. For your part, it requires some perseverance and actually trying all the stuff taught in this book yourself. You cannot learn how to play the piano or drive a car solely by reading books. The same goes for learning how to program. This book is in an electronic format, so you do not have any excuse not to switch to Xcode frequently. Therefore, as of chapter 5, we suggest you go through each chapter three times. The second time, try the examples for real, and then make small modifications to the code to explore how things work.
Introduction
Become An Xcoder
Before we start
Become An Xcoder
In secondary school, dots were out of fashion and variables called x and y (and a new fancy word, "algebra") were all the hype. Looking back, you may wonder why people felt so intimidated by this very small change in notation.
2 + 6 = x y = 3 * 4
Variables
Objective-C uses variables too. Variables are nothing more than convenient names to refer to a specific piece of data, such as a number. Here is an Objective-C statement, i.e. a line of code, where a variable is given a particular value.
//[1] x = 4;
The semicolon
The variable named x is given a value of 4. You will note there is a semi-colon at the end of the statement. That is because the semi-colon is required at the end of every statement. Why? Well, the code snippet of example [1] may look geeky to you, but a computer does not know what to do with it at all. A special program, called a compiler, is necessary to convert the text you typed into the necessary zeros and ones your Mac understands. Reading and understanding the text a human typed is very hard for a compiler, so you need to give it certain clues, for example where a particular statement ends. Which is what you do by using the semi-colon. If you forget a single semi-colon in your code, the code cannot be compiled, that is, it cannot be turned into a program your Mac can execute. Don't worry too much about that, because the compiler will complain if it can't compile your code. As we will see in a future chapter, it will try to help you find out what is wrong.
Naming variables
While variable names themselves have no special meaning to the compiler, descriptive variable names can make a program much easier for humans to read and hence easier to understand. That is a big bonus if you need to track down an error in your code. Errors in programs are traditionally called bugs. Finding and fixing them is called debugging. Hence, in real code we avoid using non-descriptive variable names like x. For example, the variable name for the width of a picture could be called pictureWidth [2].
//[2] pictureWidth = 8;
01:1
Become An Xcoder
From the big issue a compiler makes out of forgetting a semi-colon, you will understand that programming is all about details. One of those details to pay attention to is the fact that code is case-sensitive. That is, it matters whether you use capitals or not. The variable name pictureWidth is not the same as pictureWIDTH, or PictureWidth. In accordance with general conventions, I make my variable names up by fusing several words, the first without capital, and all other words making up the variable name starting with a capital, just as you can see in example [2]. This style is often called camelCase. By sticking to this scheme, I reduce the chance of programming mistakes due to case-sensitivity tremendously. Please note that a variable name always consists of a single word (or single character, at a pinch). While you have plenty freedom choosing variable names, there are several rules which a variable name has to conform with. While I could spell them all out, that would be boring at this point. The prime rule you must obey is that your variable name may not be a reserved word of Objective-C (i.e., a word that have a special meaning to Objective-C). By composing a variable name as contracted words, like pictureWidth, you are always safe. To keep the variable name readable, the use of capitals within the variable name is recommended. If you stick to this scheme, you'll have fewer bugs in your programs. If you insist on learning a couple of rules, finish this paragraph. Apart from letters, the use of digits is allowed, but a variable name is not allowed to start with a digit. Also allowed is the underscore character: "_". Here are a few examples of variable names. Good variable names: door8k do8or do_or Not allowed: door 8 (contains a space) 8door (starts with digit) Not recommended: Door8 (starts with capital)
Surprisingly, the compiler doesn't nitpick about spaces (except within variable names, keywords etc.!). To make the code easier on the eyes, we can use spaces.
//[4] pictureWidth = 8; pictureHeight = 6; pictureSurfaceArea = pictureWidth * pictureHeight;
01:2
Become An Xcoder
Numbers in general can be distinguished into two types: integers (whole numbers) and fractional numbers. You can see an example of each in the statements [5.1] and [5.2], respectively. Integers are used for counting, which is something we will do when we have to repeat a series of instructions a specified number of times (see chapter 7). You know fractional or floating-point numbers, for example, from baseball hitting averages. The code of example [5] will not work. The problem is that the compiler wants you to tell it in advance what variable names you are going to use in your program, and what type of data they are referring to, i.e. integers or floating point numbers. In geek-speak, this is called "to declare a variable".
//[6] int pictureWidth; float pictureHeight, pictureSurfaceArea; pictureWidth = 8; pictureHeight = 4.5; pictureSurfaceArea = pictureWidth * pictureHeight;
In line [6.1], int indicates that the variable pictureWidth is an integer. In the next line, we declare two variables in one go, by separating the variable names with a comma. More specifically, statement [6.2] says that both variables are of type float, i.e. numbers that contain fractional parts. In this case it is a bit silly that pictureWidth is of a different type than the other two variables. But what you can see is that if you multiply an int with a float, the result of the calculation is a float, which is why you should declare the variable pictureSurfaceArea as a float [6.2]. Why does the compiler want to know whether a variable represents an integer or a number with a fractional part? Well, a computer program needs part of the computer's memory. The compiler reserves memory (bytes) for each variable it encounters. Because different types of data, in this case int and float, require different amounts of memory and a different representation, the compiler needs to reserve the correct amount of memory and to use the correct representation. What if we are working with very big numbers or very high precision decimal numbers? They wouldn't fit in the few bytes reserved by the compiler, would they? That's right. There are two answers to this: first, both int and float have counterparts that can store bigger numbers (or numbers with higher precision). On most systems they are long long and double, respectively. But even these can fill up, which bring us to the second answer: as a programmer, it will be your job to be on the watch for problems. In any case, it is not a problem to be discussed in the first chapter of an introductory book. By the way, both integers and decimal numbers can be negative, as you know for example from your bank account. If you know that the value of a variable is never negative, you can stretch the range of values that fit in the bytes available.
//[7] unsigned int chocolateBarsInStock;
There is no such thing as a negative number of chocolate bars in stock, so an unsigned unsigned int type represents whole numbers greater than or equal to zero.
int
Declaring a variable
It is possible to declare a variable and assign it a value in one go [8].
//[8] int x = 10; float y= 3.5, z = 42;
01:3
Become An Xcoder
Data Types
As we have just seen, the data stored in a variable can be one of several specific types, for example an int or a float. In Objective-C, simple data types such as these are also known as scalar data. Here is a list of the common scalar data types available in Objective-C: Name void int unsigned float double char BOOL Type Void Integer Unsigned integer Floating point number Double precision floating point number Character Boolean Example Nothing ...-1, 0, 1, 2... 0, 1, 2... -0.333, 0.5, 1.223, 202.85556 0.52525252333234093890324592 793021 hello 0, 1; TRUE, FALSE; YES, NO.
Mathematical operations
In the previous examples, we performed a multiplication operation. Use the following symbols, officially known as operators, for doing basic mathematical calculations.
+ / * for addition for subtraction for division for multiplication
Using the operators, we can perform a wide range of calculations. If you take a look at the code of professional Objective-C programmers, you will come across a couple of peculiarities, probably because they're lazy typists. Instead of writing x
//[9] x++; //[10] ++x; = x + 1;
In either case this means: increase x by one. Under some circumstances it is important whether the ++ is before or after the variable name. Check out the following examples [11] and [12].
//[11] x = 10; y = 2 * (x++); //[12] x = 10; y = 2 * (++x);
In example [11], when all is said and done, y equals 20 and x equals 11. In contrast, in statement [12.2], x is incremented by one before the multiplication by 2 takes place. So, in the end, x equals 11 and y equals 22. The code of example [12] is equivalent with example [13].
//[13] x = 10; x++; y = 2 * x;
So, the programmer has actually merged two statements into one. Personally, I think this makes a program harder to read. If you take the shortcut that is fine but be aware that a bug may be lurking there.
01:4
Become An Xcoder
Parentheses
It will be old hat for you if you managed to pass high school, but parentheses can be used to determine the order in which operations are performed. Ordinarily * and / take precedence over + and -. So 2 * 3 + 4 equals 10. By using parenthesis, you can force the lowly addition to be performed first: 2 * (3 + 4) equals 14.
Division
The division operator deserves some special attention, because it makes quite a difference whether it is used with integers or floats. Take a look at the following examples [14, 15].
//[14] int x = 5, y = 12, ratio; ratio = //[15] float x = 5, y = 12, ratio; ratio = y / x; y / x;
In the first case [14], the result is 2. Only in the second case [15], the result is what you'd probably expect: 2.4.
Booleans
A Boolean is a simple logical true or false value. 1 and 0 stand for true and false are often used interchangably, and can be considered equivalent: True 1 False 0
They are frequently used in evaluating whether to perfom some action depending upon the boolean value of some variable or function.
Modulus
An operator you're probably unfamiliar with is % (modulus). It doesn't work as you might expect: the modulus operator is not a percentage calculation. The result of the % operator is the remainder from the integer division of the first operand by the second (if the value of the second operand is zero, the behavior of % is undefined).
//[16] int x = 13, y = 5, remainder; remainder = x % y;
Now the result is that remainder is equal to 3, because x is equal to 2*y + 3. Here are a few more examples of modulus:
21 % 7 is equal to 0 22 % 7 is equal to 1 23 % 7 is equal to 2 24 % 7 is equal to 3 27 % 7 is equal to 6 30 % 2 is equal to 0 31 % 2 is equal to 1 32 % 2 is equal to 0 33 % 2 is equal to 1 34 % 2 is equal to 0 50 % 9 is equal to 5 60 % 29 is equal to 2
It can come in handy at times, but note that it only works with integers.
A program is a series of instructions 01:5
Become An Xcoder
One common use for modulus is to determine if an integer is odd or even. If it is even, then a modulus of two will equal zero. Otherwise it will equal another value. For example:
//[17] int anInt; //Some code that sets the value of anInt if ((anInt % 2) == 0) { NSLog(@"anInt is even"); } else { NSLog(@"anInt is odd"); }
01:6
Become An Xcoder
So far our code examples have only been a few statements long, but even very simple programs can quickly grow to hundreds or thousands of lines. When you revisit your code after a few weeks or months it can be difficult to remember the reason for your programming choices. This is where comments come in. Comments help you quickly understand what a particular part of your code does and why it's there in the first place. Some programmers even go so far as to always begin coding a class as comments, which helps them organize their thinking and avoid coding themselves into a corner. You are advised to take some time commenting your code. We can assure you that you will gain back manyfold the time spent in the future. Also, if you share your code with someone else, your comments will help them adapt it to their own needs more quickly.
Making a comment
To create a comment, start the comment with two forward slashes.
// This is a comment
In Xcode comments are shown in green. If a comment is long, and spans multiple lines, put it between /*
/* This is a comment extending over two lines */
*/.
Outcommenting
We will discuss debugging a program shortly, as Xcode has great facilities for that. One way to debug the old-fashioned way is called outcommenting. By placing part of your code between /* */, you can temporarily disable ("outcomment") that part of the code, to see if the rest works as expected. This allows you to hunt down a bug. If the outcommented part should result in, for example, a value for a particular variable, you can include a temporary line where you set the variable to a value suitable for testing the remainder of your code.
Why comment?
The importance of comments cannot be overstated. It is often useful to add an explanation in plain English about what goes on in a long series of statements. That is because you don't have to deduce what the code does, and you can immediately see if the problem you are experiencing is in that part of the code. You should also use comments to express things that are difficult, or impossible to deduce from the code. For instance, if you program a mathematical function using a specific model described in details somewhere in a book, you would put a bibliographical reference in a comment associated with your actual code. Sometimes it is useful to write some comments before writing the actual code. It will help you to structure your thoughts and programming will be easier as a result. The code examples in this book do not contain as many comments as we would ordinarily have written in, because they are already surrounded by explanations.
Comments
02:1
Become An Xcoder
03: Functions
Introduction
The longest piece of code that we have seen so far had only five statements. Programs of many thousands of lines may seem a long way off, but because of the nature of Objective-C, we have to discuss the way programs are organized at an early stage. If a program were to consist of a long, continuous succession of statements, it would be hard to find and fix bugs. Besides, a particular series of statements may appear in your program in several places. If there is a bug, you must fix the same bug at several places. A nightmare, because it is easy to forget one (or two)! So, people have thought of a way to organize the code, making it easier to fix bugs. The solution to this problem is to group the statements depending on their function. For example, you may have a set of statements that allows you to calculate the surface area of a circle. Once you've checked that this set of statements works reliably, you will never have to go through that code again to see if the bug is there. The set of statements, called a function, has a name, and you can call that set of statements by this name to have its code executed. This concept of using functions is so fundamental, that there is always at least one function in a program: the main() function. This main() function is what the compiler looks for, so it will know where execution of the code at runtime must start.
Statement [1.1] shows the name of the function, i.e. "main", followed by opening and closing parentheses. While "main" is a reserved word, and the main() function is required to be present, when you define your own functions, you can call them just about anything you like. The parentheses are there for a good reason, but we won't discuss that until later in this chapter. In the following lines [1.2,1.4], there are curly braces. We must put our code between those curly braces { }. Anything between the curly braces is called the body of the function. I took some code from the first chapter and put it where it belongs [2].
//[2] main() { // Variables are declared below float pictureWidth, pictureHeight, pictureSurfaceArea; // We initialize the variables (we give the variables a value) pictureWidth = 8.0; pictureHeight = 4.5; // Here the actual calculation is performed pictureSurfaceArea = pictureWidth * pictureHeight; }
Functions
03:1
Become An Xcoder
That was easy, but our custom function starting at statement [3.9] doesn't do anything yet. Note that the function specification is outside the body of the main() function. In other words, functions are not nested. Our new circleArea() function must be called from the main() function. Let's see how we can do that [4].
//[4] main() { float pictureWidth, pictureHeight, pictureSurfaceArea, circleRadius, circleSurfaceArea; pictureWidth circleRadius = 8.0; = 5.0; // [4.7] pictureWidth * pictureHeight; pictureHeight = 4.5; pictureSurfaceArea = // [4.4]
03:2
Functions
Become An Xcoder
Passing in arguments
We added a pair of variable names of type float [4.4], and we initialized the variable circleRadius, i.e. gave it a value [4.7]. Of most interest is line [4.11], where the circleArea() function is called. As you can see, the name of the variable circleRadius has been put between the parentheses. It is an argument of the circleArea() function. The value of the variable circleRadius is going to be passed to the function circleArea(). When the function circleArea() has done its job of performing the actual calculation, it must return the result. Let's modify the circleArea() function of [3] to reflect this [5]. Note: only the circleArea() function is shown.
//[5] circleArea(float theRadius) // [5.1] { float theArea; theArea = 3.1416 * theRadius * theRadius; return theArea; } // pi times r square [5.4]
In [5.1] we define that for the circleArea() function a value of type float is required as input. When received, this value is stored in a variable named theRadius. We use a second variable, i.e. theArea to store the result of the calculation [5.4] in, so we must declare it [5.3], in the same way we declared variables in the main() function [4.4]. You will note that the declaration of the variable theRadius is done within the parentheses [5.1]. Line [5.5] returns the result to the part of the program from which the function was called. As a result, in line [4.11], the variable circleSurfaceArea is set to this value. The function in example [5] is complete, except for one thing. We have not specified the type of data that the function will return. The compiler requires us to do that, so we have no choice but to obey and indicate it is of type float [6.1].
//[6] float circleArea(float theRadius) { float theArea; theArea = 3.1416 * theRadius * theRadius; return theArea; } //[6.1]
As the first word of line [6.1] indicates, the data returned by this function (i.e., the value of variable theArea) is of type float. As a programmer, you will have ensured that the variable circleSurfaceArea in the main() function [4.8] is of that data type too, so the compiler has no reason to nag us on this one. Not all functions require an argument. If there is none, the parentheses () are still required, even though they are empty.
//[7] int throwDice() { int noOfEyes; // Code to generate a random value from 1 to 6 return noOfEyes; }
Functions
03:3
Become An Xcoder
Returning values
Not all functions return a value. If a function does not return a value, it is of type "void". Thereturn statement is then optional. If you use it, the return keyword must not be followed by a value/variable name.
//[8] void beepXTimes(int x); { // Code to beep x times return; }
If a function has more than one argument, like the pictureSurfaceArea() function below, the arguments are separated by a comma.
//[9] float pictureSurfaceArea(float theWidth, float theHeight) { // Code to calculate surface area }
The main() function should, by convention, return an integer, and so yes, it does have a return statement too. It should return 0 (zero, [10.9]), to indicate that the function was executed without problems. As the main() function returns an integer, we must write "int" before main() [10.1]. Let's put all the code we have in one list.
//[10] int main() { float pictureWidth, pictureHeight, pictureSurfaceArea, circleRadius, circleSurfaceArea; pictureWidth = 8; pictureHeight = 4.5; circleRadius = 5.0; pictureSurfaceArea = return 0; } float circleArea(float theRadius) { float theArea; theArea = 3.1416 * theRadius * theRadius; return theArea; } // [10.12] pictureWidth * pictureHeight; // [10.8] circleSurfaceArea = circleArea(circleRadius); // [10.9]
03:4
Functions
Become An Xcoder
Note: the remainder of the program is not shown (see [10]). We will soon compile this program for real. First a couple of odds and ends. When writing programs, it is advisable to keep future reuse of code in mind. Our program could have a rectangleAfunction, as shown below [12], and this function could be called in our main() function. This is useful even if the code we put in a function is used only once. The main() function becomes easier to read. If you have to debug your code, it will be easier to find where the bug might be in your program. You might find that it is in a function. Instead of having to go through a long sequence of statements, you just have to check the statements of the function, which are easy to find, thanks to the opening and closing curly braces.
rea() //[12] float rectangleArea(float length, float width) { return (length * width); }
As you can see, in a simple case like this, it is possible to have a single statement [12.3] for both the calculation and returning the result. I used the superfluous variable theArea in [10.15] just to show you how to declare the variable in a function. While the functions we defined ourselves in this chapter are rather trivial, it is important to realize that you can modify a function without impact on the code that calls the function as long as you do not change the declaration of the function (i.e., its first line). For example, you can change the variable names in a function, and the function still works (and this will not disrupt the rest of the program either). Someone else could write the function, and you could use it without knowing what goes on inside the function. All you need to know is how to use the function. That means knowing: the function's name the number, order and type of the function's arguments what the function returns (the value of the surface area of the rectangle), and the type of the result In the example [12], these answers are, respectively: rectangleArea Two arguments, both floats, where the first represents the length, the second the width. Th e function returns something, and the result is of type float (as can be learned from the first word of statement [12.1]).
Functions
03:5
Become An Xcoder
Shielded variables
The code inside the function is shielded from the main program, and from other functions, for that matter. What this means is that the value of a variable within a function is by default not affected by any other variable in any other function, even if it has the same name. This is a most essential feature of Objective-C. In Chapter 5, we will discuss this behavior again. But first, we are going to start with Xcode and run the above program [10].
03:6
Functions
Become An Xcoder
Using NSLog
Let's look into how the NSLog() function is used.
//[1] int main() { NSLog(@"Julia is my favourite actress."); return 0; }
Upon execution, the statement of example [1] would result in the text "Julia is my favourite actress." being displayed. Such text between @" and " is called a string. In addition to the string itself, the NSLog() function prints various additional information, like the current date and the name of the application. For example, the complete output of the program [1] on my system is:
2005-12-22 17:39:23.084 test[399] Julia is my favourite actress.
A string can have a length of zero or more characters. Note: In the following examples only the interesting statements of the main() function are shown.
//[2] NSLog(@""); NSLog(@" ");
Statement [2.1] contains zero characters and is called an empty string (i.e., it has a length equal to zero). Statement [2.2] is not an empty string, despite how it looks. It contains a single space, so the length of that string is 1. Several special character sequences have a special meaning in a string. These special characters sequences are known as escape sequences. For instance, to force the last word of our sentence to begin printing on a new line, a special code must be included in statement [3.1]. This code is \n, short for a new line character.
//[3] NSLog(@"Julia is my favourite \nactress.");
Now the output looks like this (only the relevant output is shown):
Julia is my favourite actress.
The backslash in [3.1] is called an escape character, as it indicates to the NSLog() function that the next character is not an ordinary character to be printed to the screen, but a character that has a special meaning: in this case the "n" means "start a new line".
Printing on screen 04:1
Become An Xcoder
In the rare event that you want to print a backslash to the screen, it may seem you have a problem. If a character after a backslash has a special meaning, how is it possible to print a backslash? Well, we just put another backslash before (or indeed after) the backslash. This tells the NSLog() function that the (second) backslash, i.e. the one more to the right, is to be printed and that any special meaning should be ignored). Here is an example:
//[4] NSLog(@"Julia is my favourite actress.\\n");
Displaying variables
So far, we have displayed static strings only. Let's print the value obtained from a calculation to the screen.
//[5] int x, integerToDisplay; x = 1; integerToDisplay = 5 + x; NSLog(@"The value of the integer is %d.", integerToDisplay);
Please note that, between parentheses, we have a string, a comma and a variable name. The string contains something funny: %d. Like the backslash, the percentage character % has a special meaning. If followed by a d (short for decimal number), upon execution, at the position of %d the output value of what is after the comma, i.e. the current value of the variable integerToDisplay, will be inserted. Running example [5] results in
The value of the integer is 6.
It is up to you how many significant digits (the ones after the period) are displayed. To display two significant digits, you put .2 between % and f, like this:
//[7] float x, floatToDisplay; x = 12345.09876; floatToDisplay = x/3.1416; NSLog(@"The value of the float is %.2f.", floatToDisplay);
Later, when you know how to repeat calculations, you may want to create a table of values. Imagine a conversion table of Fahrenheit to Celsius. If you want to display the values nicely, you want the values in the two columns of data to have a fixed width. You can specify this width with an integer value between % and f (or % and d, for that matter). However, if the width you specify is less than the width of the number, the width of the number takes prevalence.
//[8] int x = 123456; NSLog(@"%2d", x); NSLog(@"%4d", x); NSLog(@"%6d", x); NSLog(@"%8d", x);
04:2
Printing on screen
Become An Xcoder
In the first two statements [8.2, 8.3] we actually claim too little space for the number to be displayed in full, but the space is taken anyway. Only statement [8.5] specifies a width wider than the value, so now we see the appearance of additional spaces, indicative of the width of the space reserved for the number. It is also possible to combine the specification of width and the number of decimal numbers to be displayed.
//[9] float x=1234.5678 NSLog(@"Reserve a space of 10, and show 2 significant digits."; NSLog(@"%10.2d", x);
Printing on screen
04:3
Become An Xcoder
Linking to Foundation
We are only one question and one answer away from executing our first program. So, how does our program know about this useful function NSLog()? Well, it doesn't, unless we tell it to. To do that, our program has to tell the compiler to import a library of goodies (that luckily comes free with every Mac), including the function NSLog(), using the statement:
#import <Foundation/Foundation.h>
This statement must be the first statement of our program. When we put together all that we have learned in this chapter, we get the following code, which we are going to run in the next chapter.
//[11] #import <foundation/foundation.h> float circleArea(float theRadius); float rectangleArea(float width, float height); int main() { float pictureWidth, pictureHeight, pictureSurfaceArea, circleRadius, circleSurfaceArea; pictureWidth circleRadius = 8.0; = 5.0; pictureHeight = 4.5; pictureSurfaceArea = rectangleArea(pictureWidth, pictureHeight); = circleArea(circleRadius); NSLog(@"Area of circle: %10.2f.", circleSurfaceArea); NSLog(@"Area of picture: %f. ", pictureSurfaceArea); return 0; } float circleArea(float theRadius) { float theArea; theArea = 3.1416 * theRadius * theRadius; return theArea; } float rectangleArea(float width, float height) // second custom function { return width*height; } // first custom function
04:4
Printing on screen
Become An Xcoder
Creating a project
Now, start Xcode, which you find in the Applications folder of the Developer folder. When you do that for the first time, it will ask you a couple of questions. Agree with the default suggestions, they are fine, and you can always change them in the Preferences later, should you want to. To really get started, select New Project from the File menu. A dialog window appears containing a list of possible project types.
We want to create a very simple program in Objective-C, without a GUI (Graphical User Interface), so scroll down and select Foundation Tool under the Command Line Utility section.
05:1
Become An Xcoder
Enter a name for your application, such as "justatry". Choose a location where you want to save your project, and click Finish. The project we are about to create can be run from the Terminal. If you want to be able to do that, and want to avoid some hassle, make sure the name of your project is just one word. Also, it is customary not to start the names of programs run from the Terminal with a capital letter. On the other hand, names of programs with a graphical user interface should start with a capital.
05:2
Become An Xcoder
Exploring Xcode
Now you are presented with a window that you as a programmer will see a lot. The window has two frames. At the left is the "Groups & Files" frame for accessing all the files that your program is made up of. Currently there aren't too many, but later when you are creating multilingual GUI programs, this is where the files for your GUI and for the various languages can be found. The files are grouped and kept within folders, but you will search for these folders on your hard disk in vain. Xcode offers these virtual folders ("Groups") for the purpose of organizing your stuff. In the frame at the left named Groups & Files, open the group justatry to go to the group that reads Source. In it is a file named justatry.m [1]. Remember that every program must contain a function named main()? Well, this is the file that contains this main() function. Later in this chapter we are going to modify it to include the code of our program. If you open justatry.m by double-clicking its icon, you are in for a pleasant surprise. Apple has already created the main() function for you.
//[1] #import <Foundation/Foundation.h> int main (int argc, const char * argv[]) // [1.3] { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; // [1.5] // insert code here... NSLog(@"Hello, World!"); [pool drain]; //[1.9] return 0; }
05:3
Become An Xcoder
Take a look at the program and look for things you recognize. You will see: The import statement required for functions such as NSLog(), starting with a pound sign. The main() function. The curly braces which are to contain the body of our program. A comment, which invites us to put our code there. An NSLog() statement for printing a string to the screen. The return
0;
statement.
There are also a couple of things you will not recognize: Two funny-looking arguments between the parentheses of the main() function [1.3] A statement starting with NSAutoreleasePool [1.5] Another statement containing the words pool and drain [1.9] Personally I'm not exactly happy when book authors present me, the reader, with code full of unfamiliar statements and promises that it will all become clear later. Yeah, sure. That is why I went out of my way to familiarize you with the concept of "functions" so you wouldn't be confronted with too many new concepts. You already do know that functions are a way to organize a program, that every program has a main() function, and what a function looks like. However, I have to admit that I can't fully explain everything you see in example [1] right now. I'm really sorry that I have to ask you to ignore these statements (i.e., [1.3, 1.5 and 1.9]) for the time being. There are other things about the Objective-C language that you need to become familiar with first, allowing you to write simple programs. The good thing is, that you have already made it past two difficult chapters, and the upcoming three chapters are pretty easy before we have to deal with some harder stuff again. If you really don't want to be left without any explanation, here is the executive summary.The arguments in the main() function are required for running the program from the Terminal. Your program takes up memory. Memory that other programs would like to use when you're done with it. As a programmer, it is your job to reserve the memory that you need. Of equal importance, you have to give the memory back when you're done. This is what the two statements with the word "pool" in them are for.
Build and Go
Let's run the program provided by Apple [1]. Press the second hammer icon labeled Build and Go to build (compile) and run the application.
The program is executed and the results are printed in the Run Log window, together with some additional information. The last sentence says that the program has exited (stopped) with return 0. There you see the value of zero that is returned by the main() function, as discussed in Chapter 3 [7.9]. So, our program made it to the last line and didn't stop prematurely. So far so good!
05:4
Become An Xcoder
Bugging
Let's go back to example [1] and see what happens if there is a bug in the program. For example, I've replaced the NSLog() statement with another one, but I "forgot" the semi-colon indicating the end of the statement.
//[2] #import <Foundation/Foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; // insert code here... NSLog(@"Julia is my favourite actress") [pool drain]; return 0; } //[2.9] //Whoops, forgot the semicolon!
To build the application, press the build icon in the toolbar. A red circle appears before statement [2.9].
If you click it, the line below the toolbar shows a brief description of the complaint:
error: syntax error before "drain".
Parsing is one of the first things a compiler does: It walks through the code and checks whether it can understand each and every line. To help it understand the meaning of various parts, it's up to you to provide clues. So, for the import statement [2.1], you have to provide a pound sign (#). To indicate the end of a statement [2.8], you have to provide a semi-colon. By the time the compiler is at line [2.9], it notices something is wrong. However, it doesn't realize that the problem occurred not in this line, but in the previous line where the semi-colon is missing. The important lesson here is that, while the compiler tries to give sensible feedback, that feedback is not necessarily an accurate description of the actual problem, nor is the position in the program necessarily the actual position of the error (although it will probably be very close). Fix the program by adding the semi-colon and run the program again to make sure it works fine.
05:5
Become An Xcoder
float rectangleArea(float width, float height); int main (int argc, const char * argv[]) {
// [3.6]
NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; float pictureWidth, pictureHeight, pictureSurfaceArea, circleRadius, circleSurfaceArea; pictureWidth = 8; pictureHeight = 4.5; circleRadius = 5.0; pictureSurfaceArea = pictureWidth * pictureHeight; Area of circle: %10.2f.", circleSurfaceArea = circleArea(circleRadius); NSLog(@"Area of picture: %f. [pool drain]; return 0; } float circleArea(float theRadius) { float theArea; theArea = 3.1416 * theRadius * theRadius; return theArea; } float rectangleArea(float width, float height) { return width*height; } // [3.27] // [3.20] pictureSurfaceArea, circleSurfaceArea);
Take your time to make sure you understand the structure of the program. We have the function headers [3.3, 3.4] of our custom functions circleArea() [3.21] and rectangleArea() [3.28] before the main() function [3.6], as they should be. Our custom functions are outside the curly braces of the main() function [3.5]. We put the code of the body of the main() function where Apple has told us to put it. When the code is executed, we get the following output:
Area of picture: 36.000000. Area of circle: 78.54.
05:6
Become An Xcoder
Debugging
When a program gets more complicated, it gets harder to debug. So sometimes you want to find out what is going on inside the program while it is running, Xcode makes this easy to do. Just click in the grey margin before the statements for which you want to know the values of the variables. Xcode will insert a "breakpoint" represented by a blue grey arrow icon.
Please note that you will see the values of the variables before that particular statement is executed, so often you'll need to put the breakpoint at the statement after the one you are interested in. Now, keep the mouse down while clicking the second hammer button in the toolbar, and a menu will pop-up.
Select Build and Go (Debug). You will see the following window.
05:7
Become An Xcoder
The Xcode debugger lets you execute the program step by step and look at variables.
The program will run until it reaches the first breakpoint. If you check the top right pane, you will be able to see the values of the various variables. Any values set or changed since the last breakpoint are displayed in red. To continue executing, use the Continue button. The debugger is a powerful tool. Play with it for a while to familiarize with it.
Conclusion
We have now all that is needed to write, debug and run simple programs for Mac OS X. If you do not wish to make Graphical User Interface programs, the only thing you have to do now is increase your knowledge of Objective-C to enable you to develop more sophisticated non-graphical programs. In the next few chapters we're going to do exactly that. After that, we will dive into GUI-based applications. Read on!
05:8
Become An Xcoder
Line [1.2] shows the if() instruction, also known as a conditional instruction. You will recognize the curly braces, which will contain all the code you want to execute provided the logical expression between parentheses evaluates to true. Here, if the condition age > 30 is met then the string [1.4] will be printed. Whether the condition is met or not, the string of line [1.6] will be printed, because it is outside the curly braces of the if() clause.
if() else()
We may also provide an alternative set of instructions if the condition is not met, using an if...else statement [2].
//[2] // age is an integer variable that stores the user's age if (age > 30) { NSLog(@"age is older than thirty."); } else { NSLog(@"age is younger than thirty."); } NSLog(@"Finished."); //[2.7] //[2.4]
The string in statement [2.7] would only be printed if the condition were not met, which is not the case here [2].
Comparisons
Apart from the greater than sign in statement [2.2], the following comparison operators for numbers are at your disposal.
== > < >= <= != equal to greater than less than greater than or equal to less than or equal to not equal to
Take particular note of the equality operator - it is two equals signs. It is all too easy to forget this and use merely one equal sign. Unfortunately that is the assignment operator, and would set the variable to a particular value. This is a common cause of confusion, and buggy code, for beginners. Now say it out loud: I will not forget to use two equal signs when testing for equality!
Conditional Statements
06:1
Become An Xcoder
Comparison operators are quite useful when you want to repeat a series of statements several times. That will be the topic of the next chapter. First, we will discuss some other aspects of if statements that may come in handy.
Exercise
Let's take a closer look at performing a comparison. A comparison operation results in one of only two possible outcomes: The result is either true or false. In Objective-C, true and false are represented as either 1 or 0 respectively. There even is a special data type, named BOOL that you can use to represent such values. To denote the "true" value, you can write either 1 or YES. To denote the "false" value, you can write either 0 or NO.
//[3] int x = 3; BOOL y; y = (x == 4); // y will be 0.
It is possible to check for more conditions. If more than one condition must be met, use a logical AND, represented by two ampersands: &&. If at least one of the conditions must be met, use a logical OR represented by two pipes: ||.
//[4] if ( (age >= 18) && (age < 65) ) { NSLog(@"Probably has to work for a living."); }
It is also possible to nest conditional statements. This is simply a matter of putting one conditional statement inside the curly brackets of another conditional statement. First the outermost condition will be evaluated, then, if it is met, the next statement inside, and so on:
//[5] if (age >= 18) { if (age < 65) { NSLog(@"Probably has to work for a living."); } }
06:2
Conditional Statements
Become An Xcoder
But even then, that would require repeating the call. At times, you will need to execute a one or more of statements several times. Like all programming languages, Objective-C offers several ways to achieve that.
for()
If you know the number of times the statement (or group of statements) has to be repeated, you may specify that by including that number in the for statement of example [2]. The number must be an integer, because you cannot repeat an operation, say, 2.7 times.
//[2] int x; for (x = 1; x <= 10; x++) { NSLog(@"Julia is my favourite actress."); } NSLog(@"The value of x is %d", x); //[2.4]
In example [2], the string [2.4] would be printed 10 times. First, x is assigned the value of 1. The computer then evaluates the condition with the formula we have put in place: x <= 10. This condition is met (since x is equal to 1), so the statement(s) between the curly braces are performed. Then, the value of x is increased, here by one, due to the expression x++. Subsequently, the resulting value of x, now 2, is compared with 10. As it is still smaller than and not equal to 10, the statements between the curly braces are executed again. As soon as x is 11, the condition x <= 10 is no longer met. The last statement [2.6] was included to prove to you that x is 11, not 10, after the loop has finished. At times, you will need to make larger steps than just a simple increment using x++. All you need to do is substitute the expression you need. The following example [2] converts degrees Fahrenheit to degrees Celsius.
//[3] float celsius, tempInFahrenheit; for (tempInFahrenheit = 0; tempInFahrenheit <= 200; tempInFahrenheit = tempInFahrenheit + 20) { celsius = (tempInFahrenheit - 32.0) * 5.0 / 9.0; NSLog(@"%10.2f -> %10.2f", tempInFahrenheit, celsius); }
07:1
Become An Xcoder
120.00 -> 48.89 140.00 -> 60.00 160.00 -> 71.11 180.00 -> 82.22 200.00 -> 93.33
while()
Objective-C has two other ways to repeat a set of statements:
while () { }
and
do {} while ()
The former is basically identical to the for-loop we discussed above. It starts by performing a condition evaluation. If the result of that evaluation is false, the statements of the loop are not executed.
//[4] int counter = 1; while (counter <= 10) { NSLog(@"Julia is my favourite actress.\n"); counter = counter + 1; } NSLog(@"The value of counter is %d", counter);
In this case, the value of counter is 11, should you need it later in your program! With the do
//[5] int counter = 1; do { NSLog(@"Julia is my favourite actress.\n"); counter = counter + 1; } while (counter <= 10); NSLog(@"The value of counter is %d", counter); {} while ()
instruction, the statements between the curly braces are executed at least once.
The counter's value at the end is 11. You have gained some more programming skills, so now to tackle a harder subject. In the next chapter, we are going to build our first program with a Graphical User Interface (GUI).
07:2
Become An Xcoder
Objects in action
As an example, let's take a look at how windows are handled within a program written in Objective-C, such as Safari. Take a look at an open Safari window of your Mac. At the top left, there are three buttons. The red one is the close button. So what happens if you close a window by clicking that red button? A message is sent to the window. In response to this message, the window executes some code in order to close itself.
The window is an object. You can drag it around. The three buttons are objects. You can click them. These objects have a visual representation on screen, but this is not true of all objects. For instance, the object that represents a connection between Safari and a given web site does not have a visual representation on screen.
08:1
Become An Xcoder
An object (e.g. the window) can contain other objects (e.g. the buttons)
Classes
You can have as many Safari windows as you want. Do you think that Apple's programmers: Programmed each of those windows in advance, using their massive brainpower to anticipate how many windows you might want to have, or Made a kind of template and let Safari create the window objects from it on the fly? Of course, the answer is 2. They created some code, called a class, which defines what a window is and how it should look and behave. When you create a new window, it is actually this class that creates the window for you. This class represents the concept of a window, and any particular window is actually an instance of that concept (in the same way that 76 is an instance of the concept of a number).
Instance Variables
The window you created is present at a certain location on the screen of your Mac. If you minimize the window to the Dock, and make it reappear, it will take exactly the same position on the screen that it had before. How does this work? The class defines variables suitable for remembering the position of the window on the screen. The instance of the class, i.e. the object, contains the actual values for these variables. So, each window object contains the values of certain variables, and different window objects will in general contain different values for those variables.
Methods
The class not only created the window object, but also gave it access to a series of actions it can perform. One of those actions is close. When you click the "close" button of a window, the button sends the message close, to that window object. The actions that can be performed by an object are called methods. As you will see, they resemble functions very closely, so you will not have much trouble learning to use them if you have followed us so far.
Objects in memory
When the class creates a window object for you, it reserves memory (RAM) to store the position of the window and some other information. However, it does not make a copy of the code to close the window. That would be a waste of computer memory because this code is the same for every window. The code to close a window needs to be present only once, but every window object has access to that code belonging to the window class. As before, the code you are about to see in this chapter contains some lines for reserving memory and releasing it back to the system. As indicated earlier, we will not discuss this advanced subject until much later. Sorry.
08:2
Become An Xcoder
Exercise
Our Application
We are going to create an application with two buttons and a text field. If you press one button, a value is entered into the text field. If you press the other button, another value is put into the text field. Think of it as a two-button calculator that can't do calculations. Of course, once you learn more you can figure out how to create a calculator for real, but I like small steps.
If one of the buttons of our app is pressed, it will send a message. This message contains the name of a method to be executed. This message is sent to, well, to what? In case of the window, the close message was sent to that window object, which was an instance of the window class. What we need now is an object that is capable of receiving a message from each of the two buttons, and can tell the text field object to display a value.
08:3
Become An Xcoder
Here is the general format of how to send messages in Objective-C, both without [1.1] and with [1.2] an argument:
//[1] [receiver message]; [receiver messageWithArgument:theArgument];
As you can see in each of these statements, the whole shebang is placed between square brackets and the eternal semicolon is present as the finishing touch. Between the brackets, the receiving object is mentioned first, followed by the name of one of its methods. If the method being called requires one or more values, they must be provided as well [1.2].
Double click on the MainMenu.nib file in Xcode 08:4 A Program With a GUI
Become An Xcoder
Dragging GUI objects from the palettes window to your application's window.
Behind the scenes, the action of dragging a button from the palettes window to your application's window creates a new button object and puts it in your window. The same goes on for the text field and any other objects you might drag to your window from the palettes window. Note that if you hold your cursor over an icon in the palettes window, a name will be displayed, such as NSButton or NSTextView. These are the names of classes provided by Apple. Later in this chapter we will see how we can find the methods of these classes, which we need to perform the necessary actions in our program. Arrange the objects you dragged onto the window "Window" nicely. Resize them as you see fit. Change the text of the button objects by double-clicking them, one at a time. I invite you to explore the palettes window after we have finished this project to get a grip on how to add other objects to your window.
08:5
Become An Xcoder
Class background
As promised above, we are going to create a class. But before we do that, let's look a bit deeper into how classes work. To save a lot of programming effort, it would be nice if you could build on what others have already built, instead of writing everything from scratch. If you, for example, wanted to create a window with special properties (capabilities), you would need to add just the code for these properties. You wouldn't need to write code for all other behavior, such as minimizing or closing a window. By building on what other programmers have done, you would inherit all these kinds of behavior for free. And this is what makes Objective-C so different from plain C. How is this done? Well, there is a window class (NSWindow), and you could write a class that inherits from that class. Suppose you add some behavior to your own window class. What happens if your special window receives a "close" message? You didn't write any code for that, and didn't copy such code into your class either. Simple, if the class of the special window doesn't contain the code for a particular method, the message is transferred automatically to the class from which the special window class inherits (its "superclass"). And if necessary, this goes on until the method is found (or it reaches the top of the hierarchy of inheritance.) If the method cannot be found, you sent a message that can't be handled. It is like requesting a garage to change the tires of your sleigh. Even the boss of the garage can't help you. In such cases Objective-C will signal an error.
08:6
Become An Xcoder
Custom classes
What if you want to implement your own behavior for a method already inherited from your superclass? That is easy, you can override particular methods. For example, you could write code that, on clicking the close button, would move the window out of view before actually closing it. Your special window class would use the same method name for closing a window as the one defined by Apple. So, when your special window receives a close message, the method executed is yours, and not Apple's. So, now the window would move out of sight, before being closed for real. Hey, closing a window for real was already programmed by Apple. From inside our own close method, we can still invoke the close method implemented by our superclass, although it requires a slightly different call to make sure our own close method is not called recursively.
//[2] // Code to move the window out of sight here. [super close]; // Use the close method of the superclass.
This stuff is way too advanced for this introductory booklet and we don't expect you to "get it" from these few lines.
08:7
The first two capitals of MAFoo stand for My Application. You can invent class names as you like. Once you start writing your own applications, we recommend that you choose a two or three letter prefix that you'll use for all your classes in order to avoid confusion with existing class names. However, don't use NS, as this may confuse you later on. NS is used for Apple's classes. It stands for NeXTStep, NeXTStep being the operating system Mac OS X was based on when Apple bought NeXT, Inc., and got Steve Jobs back as a bonus. The CocoaDev wiki contains a list of other prefixes to avoid. You should check it when choosing your own prefix: http://www.cocoadev.com/index.pl?ChooseYourOwnPrefix When creating a new class you should give it a name that conveys useful information about that class. For instance, we've already seen that in Cocoa the class used to represent windows is named NSWindow. Another example is the class that is used to represent colors, which is named NSColor. In our case, the MAFoo class we are creating is just here to illustrate the way objects communicate together in an application. This is why we gave it a generic name with no special meaning.
Next, select the Identity button in the Inspector palette (Cmd-6), then choose MAFoo from the Class pop-up menu. We have now instantiated our MAFoo class in Xcode into an object in our nib file. This will allow our code and our interface to communicate.
08:8
Become An Xcoder
Creating connections
Our next step is to create connections between the buttons (from which messages are sent) to our MAFoo object. In addition, we are going to create a connection back from the MAFoo object to the text field, because a message will be sent to the text field object. An object has no way of sending a message to another object if it doesn't have a reference to the other object. By making a connection between a button and our MAFoo object, we are providing that button with a reference to our MAFoo object. Using this reference, the button will be able to send messages to our MAFoo object. Likewise, establishing a connection from our object to the text field will allow the former to message the latter. Let us again go through what the application has to do. Each of the buttons can send, when clicked, a message corresponding to a particular action. This message contains the name of the method of the class MAFoo that has to be executed. The message is sent to the instance of the MAFoo class we've just created, the MAFoo object. (Remember: object instances themselves don't contain the code to perform the action, but the classes do.) So, this message sent to the MAFoo object triggers a method of the class MAFoo to do something: in this case, sending a message to the text field object. Like every message, this one consists of the name of a method (which the text field object will have to execute). In this case, the method of the text field object has the task of displaying a value, and that value has to be sent as part of the message (called an "argument", remember?), along with the name of the method to invoke on the text field. Our class needs two actions (methods), which will be called by the (two) button objects. Our class needs one outlet, a variable for remembering which object (i.e., the text field object) is to be sent a message. Make sure MAFoo is selected in the MainFile.nib window. On your keyboard, press Cmd-6 in order to bring on screen the Identity inspector. In the inspector window, in the Action section, click the Add (+) button to add an action (i.e., an action method) to the MAFoo class. Replace the default name provided by Interface Builder by a more meaningful name (for example, you can enter "setTo5:" because we will program this method to display the number 5 in the text field). Add another method, and give it a name (for example "reset:", because we will program it to display the number 0 in the text field). Note that our method names both end with a colon (":"). More about this later. Now, in the inspector window, select the Outlet tab, add an outlet and give it a name (for example "textField").
A Program With a GUI 08:9
Become An Xcoder
Before establishing connections between objects, we are going to give meaningful names to our two buttons. Since the first one is going to ask our MAFoo instance to display the number 5 in the text field, we name it "Set to 5" (we already learned how to change the name of a button: double click on its name on-screen, and then enter the new name). Likewise, we name the second one "Reset". Note that this step of giving this button a particular name is not required for our program to work correctly. It is just that we want our user interface to be as descriptive as possible to the enduser. Now we are ready to create the actual connections between the button "Reset" and the MAFoo instance the button "Set to 5" and the MAFoo instance the MAFoo instance and the text field. To create the connections, press the Control key on you keyboard and use the mouse to drag from the "Set to 5"" button to the MaFoo instance in the MainMenu.nib window (don't do it the other way around!). A line representing the connection will appear on screen, and a menu will pop up on the object instance icon. Choose "setTo5:" from the list.
08:10
Become An Xcoder
Now the button holds a reference to our MAFoo object, and will send it the setTo5: message whenever it is pressed. You can now connect the "Reset" button to the MAFoo object by applying the same process. To create the connection between the MAFoo object and the text field, start from the MAFoo object and control-drag to the text field object. Click "textField" in the menu to assign the connection. What was this all about? Well, as you will see in a minute, you have just in effect created some code without writing a single line.
08:11
Become An Xcoder
Generate Code
Go to the File menu in Interface Builder and choose Write Class Files. Interface Builder then asks you where you want your generated file to be put on disk. Navigate to the project folder of our application and overwrite the MAFoo class that exists there. Now, if you switch back to Xcode, you'll see the generated files in your project window, inside the Classes group. Click on the Editor toolbar button, then choose MAFoo.h.
Let's go back for a moment to Chapter 4, where we discussed functions. Do you remember the function header [11.1]? It was a kind of warning for the compiler to tell it what it could expect. One of the two files we have just created is named MAFoo.h, and it is a header file: it contains info about our class. For example, you'll recognize that there is a line [3.5] containing NSObject, which line means that our class inherits from the NSObject class.
//[3] /* MAFoo */ #import <cocoa/cocoa.h> // [3.2]
You will see that there is one outlet [3.7] to the text field object. id indicates object. "IB" stands for Interface Builder, the program you used to create this code.
08:12
Become An Xcoder
IBAction [3.9, 3.10] is equivalent to void. Nothing is returned to the object that sends the message: the buttons in our program do not get a reply from the MAFoo object in response to their message.
You can also see there are two Interface Builder Actions. These are two methods of our class. Methods are quite like functions, which we already know, but there are differences. More on that later. Earlier we have seen #import <Foundation/Foundation.h> instead of line [3.2], The former is for nonGUI apps, the latter for GUI-apps. Now let's check out the second file, MAFoo.m. Again we get a lot of code for free.
//[4] #import "MAFoo.h" @implementation MAFoo - (IBAction)reset:(id)sender { } - (IBAction)setTo5:(id)sender { } @end // [4.5]
First of all, the MAFoo.h header file is imported, so the compiler knows what to expect. Two methods can be recognized: reset: and setTo5:. These are the methods of our class. They are similar to functions in that you need to put your code between the curly braces. In our application, when a button is pressed, it sends a message to your MAFoo object, requesting the execution of one of the methods. We do not have to write any code for that. Making the connections between the buttons and the MAFoo object in Interface Builder is all what is required. However, we do have to implement the two methods, i.e. we need to write the code that performs their function. In this case, these methods do nothing but send a message each from our MAFoo object to the textField object, so we provide the statements [5.7, 5.12].
//[5] #import "MAFoo.h" @implementation MAFoo - (IBAction)reset:(id)sender { [textField setIntValue:0]; // [5.7] } - (IBAction)setTo5:(id)sender { [textField setIntValue:5]; // [5.12] } @end
As you can see, we send a message to the object referenced by the textField outlet. Since we connected this outlet to the actual text field, using Interface Builder, our message will be sent to the correct object. The message is the name of a method, setIntValue:, together with an integer value. The setIntValue: method is capable of displaying an integer value in a text field object. In the next chapter we will tell you how we found out about this method.
08:13
Become An Xcoder
Ready to rock
You are now ready to compile your application and launch it. As usual press the Build and Go button in the Xcode toolbar. It will take a few second for Xcode to build the application and to launch it. Eventually, the application will appear on screen and you'll be able to test it.
In short, you have just created a (very basic) application, for which you had to write just two lines of code yourself!
08:14
Become An Xcoder
Exercise
In Interface Builder, if you hold the pointer over an object in the palettes window, a small label pops up. If you hold your cursor above the button icon, you'll see "NSButton". If you hold it above the text field reading "System Font Text", you will see "NSTextField". Each of these names is a class name. Let's check out the NSTextField class to see what methods are available. Go to Xcode, and, in the menu, select Help Documentation. In the frame on the left, select Cocoa and then enter "NSTextField" in the search field (make sure the API-Search mode is selected; see the screen-shot below). While you are typing, the list of possible hits is reduced significantly and soon you will see NSTextField appearing on top. Click the line that reads NSTextField (of type Class) to get information on the class NSTextField displayed in the lower frame.
Finding Methods
09:1
The first thing you should notice is that this class inherits from a series of other classes. The last one in the list is topdog, king of the hill NSObject. A little lower (please scroll) there is the heading: Method Types That is where we are going to start our search. A quick glance over the subheadings will tell us that we are not going to find the method that we need to display a value in the text field object here. Because of the principle of inheritance, we need to visit the immediate superclass of the NSTextField class, which is NSControl (and if we fail, we have to scrutinize its superclass NSView, etc.). Because all the documentation is in HTML, all we have to do is click the word NSControl (as shown above in the Inherits from list). This brings us to the information on the NSControl class: NSControl Inherits from NSView : NSResponder : NSObject You can see we moved one class up. In the method list, we notice a subheading: Setting the control's value That is what we want, we want to set a value. Below this subheading we find: - setIntValue: Sounds promising, so we check out the description of this method by clicking the setIntValue: link. setIntValue: - (void)setIntValue:(int)anInt Sets the value of the receiver's cell (or selected cell) to the integer anInt. If the cell is being edited, it aborts all editing before setting the value; if the cell doesn't inherit from NSActionCell, it marks the cell's interior as needing to be redisplayed (NSActionCell performs its own updating of cells). In our application, our NSTextField object is the receiver and we have to feed it an integer. We can also see this from the signature of the method:
- (void)setIntValue:(int)anInt
In Objective-C, the minus sign marks the beginning of an instance method declaration (as opposed to a class method declaration, which we'll talk about later). void indicates that nothing is returned to the invoker of the method. That is, when we send a setIntValue: message to textField, our MAFoo object does not receive a value back from the text field object. That is ok. After the colon, (int) indicates that the variable anInt must be an integer. In our example, we send it a value of 5 or 0, which are integers, so we are fine. Sometimes it is a bit more difficult to find out which is the appropriate method to use. You will get better at this when you're more familiar with the documentation, so keep practicing.
09:2
Finding Methods
Become An Xcoder
What if you want to read the value from our text field object textField? Remember the great thing about functions being that all variables inside were shielded? The same is true for methods. Often however, objects have a pair of related methods, called "Accessors", one for reading the value, and one for setting the value. We already know the last one, this is the setIntValue: method. The first one looks like this:
//[1] - (int) intValue
As you can see, this method returns an integer. So, if we want to read the integer value associated with our textfield object we have to send it a message like this:
//[2] int resultReceived = [textField intValue];
Again, in functions (and methods) all variable names are shielded. That is great for variable names, because you do not have to fear that setting a variable in one part of your program will affect a variable with the same name in your function. However, function names must still be unique in your program. Objective-C takes shielding one step further: method names have to be unique within a class only, but different classes may have method names in common. This is a great feature for large programs, because programmers can write classes independent of each other, without having to fear conflicts in method names. But there is more. The fact that different methods in different classes can have the same name is called Polymorphism in geek (and, indeed, Greek) speak, and is one of the things that makes objectoriented programming so special. It allows you to write chunks of code without having to know in advance what are the classes of the objects you are manipulating. All that is required is that, at runtime, the actual objects understand the messages you send them. Taking advantage of polymorphism, you can write applications that are flexible and extensible by design. For instance, in the GUI application we created, if we replace the text field by an object of a different class that is able to understand the setIntValue: message, our application will still work without requiring us to modify our code, or even to recompile it. We are even able to vary the object at run-time without breaking anything. Therein lies the power of object-oriented programming.
Finding Methods
09:3
Become An Xcoder
10: awakeFromNib
Introduction
Apple has done a lot of work for you, making it easier to create your programs. In your little application, you didn't have to worry about drawing a window and buttons to a screen, amongst many other things. Most of this work is made available through two frameworks. The Foundation Kit framework that we imported in example [12] of Chapter 4, provides most services not associated with a graphical user interface. The other framework, called the Application Kit, deals with objects you see on screen and user-interaction mechanisms. Both frameworks are well documented. Let's go back to our GUI application. Suppose we want our application to display a particular value in the text field object immediately when the application is launched and the window is initially shown.
Exercise
All the information for the window is stored in a nib file (nib stands for NeXT Interface Builder). This is a good indication that the method we need may be part of the Application Kit. Let's find out how to get information about this framework. In Xcode, go to the Help menu and select Documentation. In the documentation window make sure Full-Text Search is enabled (to do that, click on the little lens in the search field and select Full-Text Search in the associated menu). Then type Application Kit in the search field and press Return. Xcode provides you with multiples results. Among them is a document named Application Kit Framework Reference. Inside it you'll find a list of services provided by this framework. Under the headings Protocols there is a link called NSNibAwaking. If you click on it, you are taken to the documentation for the NSNibAwaking class. NSNibAwaking Protocol Objective-C Reference (informal protocol) Framework /System/Library/Frameworks/AppKit.framework Declared in AppKit/NSNibLoading.h Companion document Loading Resources Protocol Description This informal protocol consists of a single method, awakeFromNib. Classes can implement this method to initialize state information after objects have been loaded from an Interface Builder archive (nib file). If we implement this method, it will be called when our object is loaded from its nib file. Thus we can use it to achieve our goal: displaying a value in the text field at launch time. By no means do I want to suggest that it is always trivial to find the correct method. Often, it will require quite a bit of browsing and the creative use of keywords for searches, to find a promising method. For that reason, it is highly important that you familiarize yourself with the documentation of both frameworks, so you will know what classes and methods are available to you. You may not need it at that time, but it will help you to figure out how to get your program to do what you want.
awakeFromNib
10:1
Become An Xcoder
Ok, now we have found our method, all we need to do is to add the method to our implementation file MAFoo.m [1.15].
//[1] #import "MAFoo.h" @implementation MAFoo - (IBAction)reset:(id)sender { [textField setIntValue:0]; } - (IBAction)setTo5:(id)sender { [textField setIntValue:5]; } - (void)awakeFromNib { [textField setIntValue:0]; } @end // [1.15]
When the window is opened, the awakeFromNib method is called automatically. As a result, the text field displays zero when you lay your eyes on the newly opened window.
10:2
awakeFromNib
Become An Xcoder
11: Pointers
Warning!
This chapter contains advanced concepts and deals with underlying C language concepts that beginners may find intimidating. If you don't understand it all now, don't worry. Thankfully, in general - although understanding how pointers work is useful - it is not essential to start programming in Objective-C.
Introduction
When you declare a variable your Mac associates this variable with some space in its memory in order to store the value of the variable. For instance, examine the following instruction:
//[1] int x = 4;
In order to execute it, your Mac finds some space in its memory that is not already in use and then notes that this space is where the value of the variable x is to be stored (of course we could and should have used a more descriptive name here for our variable). Look at the instruction [1] again. Indicating the type of the variable (here int) lets your computer know how much space in memory is needed to store the value of x. If the value were of type long long or double, more memory would need to be reserved. The assignment instruction x = 4 stores the number 4 in this reserved space. Of course, your computer remembers where the value of the variable named x is stored in its memory, or, in other words, what the address of x is. That way, each time you use x in your program, your computer can look in the right place (at the right address) and find the actual value of x. A pointer is simply a variable that contains the address of another variable.
Referencing variables
Given a variable, you can get its address by writing & before the variable. For example, to get the address of x, you write &x.
4).
When the computer evaluates the expression x it returns the value of the variable x (in our example, it will return By contrast, when the computer evaluates the expression &x, it returns the address of the variable x, not the value stored there. The address is a number that denotes a specific place in the memory of the computer (like a room number denotes a specific room in a hotel).
Using Pointers
You declare a pointer like this:
//[2] int *y;
This instruction defines a variable named y that will contain the address of a variable of type int. Again: it will not contain an int variable, but the address to such a variable. To store in variable y what the address of variable x is (in official geek speak: assign the address of x to y), you do:
//[3] y = &x;
Now y "points at" the address of x. Using y, thus, you can track down x. Here's how. Given a pointer, you can get at the variable it points to by writing an asterisk before the pointer. For instance, evaluatPointers 11:1
Become An Xcoder
will return 4. This is equivalent to evaluating the expression x. Executing the instruction:
*y = 5
Pointers are useful because sometimes you don't want to refer to the value of a variable, but to the address of that variable. For instance, you may want to program a function that adds 1 to a variable. Well, can't you just do it like this?
//[4] void increment(int x) { x = x + 1; }
Actually, no. If you call this function from a program, you won't get the results you were expecting:
//[5] int myValue = 6; increment(myValue); NSLog(@"%d:\n", myValue);
This code would display 6 on your screen. Why? Didn't you increase myValue by calling the increment function? No, you actually didn't. You see, the function in [4] just took the value of myValue (i.e. the number 6), increased it by one, and... basically, threw it away. Functions only work with the values you pass to them, not the variables that carry these values. Even if you modify the x (as you can see it in [4]), you're only modifying the value that the function received. Any such modification will be lost when the function returns. Besides, that x isn't necessarily even a variable: if you call increment(5);, what would you expect to increment? If you want to write a version of the increment function that actually works, i.e. accepts a variable as its argument and permanently increases the value of that variable, you need to pass it the address of a variable. That way, you can modify what is stored in this variable, not just use its current value. Thus, you use a pointer as argument:
//[6] void increment(int *y) { *y = *y + 1; }
11:2
Pointers
Become An Xcoder
12: Strings
Introduction
So far, we have seen several basic data types: int, long, float, double, BOOL. Plus in the last chapter we introduced pointers. While we touched on the subject of strings, we have only discussed it in relation to the NSLog() function. This function allowed us to print a string to the screen, replacing codes starting with a %-sign, such as %d, with a value.
//[1] float piValue = 3.1416; NSLog(@"Here are three examples of strings printed to the screen.\n"); NSLog(@"Pi approximates %10.4f.\n", piValue); NSLog(@"The number of eyes of a dice is %d.\n", 6);
We did not discuss strings as data types before, for good reason. Unlike ints or floats, strings are true objects, created using the class NSString or the class NSMutableString. Let's discuss these classes, beginning with NSString.
NSString
Pointers again
//[2] NSString *favoriteComputer; favoriteComputer = @"Mac!"; NSLog(favoriteComputer);
You'll probably find the second statement comprehensible, but the first one [2.1] deserves a bit of explanation. Remember that, when we declared a pointer variable, we had to tell what type of data the pointer was pointing to? Here is a statement from chapter 11 [3].
//[3] int *y;
Here we tell the compiler that the pointer variable y contains the address of a memory location where an integer can be found. In [3.1] we tell the compiler that the pointer variable favoriteComputer contains the address of a memory location where an object of type NSString can be found. We use a pointer to hold our string because in Objective-C, objects are never manipulated directly, but always through pointers to them. Don't worry too much if you don't fully understand this - it's not crucial. What is important is to always refer to an instance of NSString or NSMutableString (or indeed any object) using the * notation.
The @ symbol
Ok, why does this funny @ sign show up all the time? Well, Objective-C is an extension of the C-language, which has its own ways to deal with strings. To differentiate the new type of strings, which are fully-fledged objects, Objective-C uses an @ sign.
Strings
12:1
Become An Xcoder
Exercise
Of course, it is possible to declare and initialize the pointer variable for a string in one go [4].
//[4] NSString *favoriteActress = @"Julia";
The pointer variable favoriteActress points to a location in memory where the object representing the string "Julia" is stored. Once you have initialized the variable, i.e. favoriteComputer, you may give the variable another value, but you cannot change the string itself [5.7] because it is an instance of class NSString. More on this in a minute.
//[5] #import <foundation/foundation.h> int main (int argc, const char *argv[]) { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; NSString *favoriteComputer; favoriteComputer = @"iBook"; // [5.7] favoriteComputer = @"MacBook Pro"; NSLog(@"%@", favoriteComputer); [pool release]; return 0; }
NSMutableString
A string of class NSString is called immutable, because it cannot be modified. What good is a string you can't modify? Well, strings that can't be modified are easier for the operating system to handle, so your program can be faster. In fact when you use Objective-C to write your own programs, you'll find that most times you don't need to modify your strings. Of course, at times you will need strings that you can modify. So, there is another class, and the string objects you create with it are modifiable. The class to use is NSMutableString. We'll discuss it later in this chapter.
Exercise
First, let's make quite sure that you understand that strings are objects. As they are objects, we can send messages to them. For example, we can send the message length to a string object [6].
//[6] #import <Foundation/Foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool * pool = [[NSAutoreleasePool alloc] init]; int theLength; NSString * foo; foo = @"Julia!"; theLength = [foo length]; // [6.10] NSLog(@"The length is %d.", theLength); [pool release]; return 0; }
12:2
Strings
Become An Xcoder
Programmers often use foo and bar as variable names when explaining things. Actually, they're bad names, because they are not descriptive, just like x. We expose you to them here, so you will not be puzzled when you see them in discussions on the Internet. In line [6.10] we send the object foo, the message length. The method length is defined in the NSString class as follows: - (unsigned int)length Returns the number of Unicode characters in the receiver. You may also change the characters of the string to uppercase [7]. To that end, send the string object the appropriate message, i.e. uppercaseString, which you should be able to find in the documentation yourself (check the methods available in the NSString class). Upon reception of this message, the string object creates and returns a new string object containing the same content, with each character changed to its corresponding uppercase value.
//[7] #import <Foundation/Foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; NSString *foo, *bar; foo = @"Julia!"; bar = [foo uppercaseString]; NSLog(@"%@ is converted into %@.", foo, bar); [pool release]; return 0; }
Sometimes you might want to modify the content of an existing string instead of creating a new one. In such case you'll have to use an object of class NSMutableString to represent your string. NSMutableString provides several methods that allow you to modify the content of a string. For instance, the method appendString: appends the string passed as argument to the end of the receiver.
//[8] #import <Foundation/Foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; NSMutableString *foo; foo = [@"Julia!" mutableCopy]; [foo appendString:@" I am happy"]; NSLog(@"Here is the result: %@.", [pool release]; return 0; } foo); // [8.7] // [8.8]
Strings
12:3
Become An Xcoder
In line [8.8], the method mutableCopy (which is provided by the NSString class) creates and returns a mutable string with the same content as the receiver. That is, after the execution of the line [8.8], foo points to a mutable string object which contains the string "Julia!".
A string object
Julia!
Now suppose we assign the value of foo to the variable bar like this:
bar = foo;
The result of this operation is that both foo and bar now point to the same object:
A string object
Julia!
12:4
Strings
Become An Xcoder
In such a situation, sending a message to the object using foo as the receiver (e.g. [foo dosomething];) has the same effect as sending the message using bar (e.g. [bar dosomething];), as shown in this example:
[9] #import <Foundation/Foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; NSMutableString *foo = [@"Julia!" mutableCopy]; NSMutableString *bar = foo; NSLog(@"foo points to the string: %@.", NSLog(@"bar points to the string: %@.", NSLog(@"\n"); [foo appendString:@" I am happy"]; NSLog(@"foo points to the string: %@.", NSLog(@"bar points to the string: %@.", [pool release]; return 0; } foo); bar); foo); bar);
Being able to have references to the same object from different places at the same time is an essential feature of objectoriented languages. Actually we've already used it in previous chapters. For instance, in chapter 8, we referenced our MAFoo object from two different button objects.
Strings
12:5
Become An Xcoder
13: Arrays
Introduction
At times you will need to hold collections of data. For example, you might need to maintain a list of strings. It would be quite cumbersome to use a variable for each of those strings. Of course, there is a more convenient solution: the array. An array is an ordered list of objects (or, more exactly, a list of pointers to objects). You can add objects in an array, remove them or ask the array to let you know which object is stored at a given index (i.e., at a given position). You can also ask the array to let you know how many elements it contains. When you count items, you usually start with 1. In arrays however, the first item is at index zero, the second at index 1 and so on.
0 1 2
We will give you some example of code later in this chapter, allowing you to see the effect of counting starting with zero. Arrays are provided by two classes: NSArray and NSMutableArray. As with strings, there is an immutable and a mutable version. In this chapter, we'll consider the mutable version. These are arrays specific to Objective-C and Cocoa. There is another, simpler kind of array in the C language (which is thus also part of Objective-C), but we won't be discussing it here. This is just a reminder that you may later read about C arrays elsewhere, and be sure to understand that they won't have much to do with NSArrays or NSMutableArrays.
A class method
One way to create an array is to execute an expression like this:
[NSMutableArray array];
When evaluated, this code creates and returns an empty array. But... wait a minute... this code seems odd, doesn't it? Indeed, in this case we have used the name of the NSMutableArray class for specifying the receiver of a message. But so far we have only sent messages to instances, not to classes, right? Well, we have just learned something new: the fact that, in Objective-C, we can also send messages to classes (and the reason is that classes are also objects, instances of what we call meta-classes, but we won't explore that idea further in this introductory article).
Arrays
13:1
Become An Xcoder
It should be noted that this object is automatically autoreleased when created; that is, it is attached to an NSAutoreleasePool and set to destruct by the class method that created it. Calling the class method is equivalent to:
NSMutableArray *array = [[[NSMutableArray alloc] init] autorelease];
In the event that you want the array to persist longer than the lifespan of the autorelease pool, you must send the instance a -retain message.
In the Cocoa documentation, the methods we can call on classes are denoted by a leading "+" symbol, instead of the "-" symbol we usually see before the name of methods (for example Chapter 8 [4.5]). For instance, in the documentation we see this description for the array method: array + (id)array Creates and returns an empty array. This method is used by mutable subclasses of NSArray. See Also: + arrayWithObject:, + arrayWithObjects:
Exercise
Let's go back to coding. The following program creates an empty array, stores three strings in it, and then prints the number of elements in the array.
//[1] #import <foundation/foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; NSMutableArray *myArray = [NSMutableArray array]; [myArray addObject:@"first string"]; [myArray addObject:@"second string"]; [myArray addObject:@"third string"]; int count = [myArray count]; NSLog(@"There are %d elements in my array", count); [pool release]; return 0; }
13:2
Arrays
Become An Xcoder
The following program is the same as the previous one except that it prints the string stored at index 0 in the array. To get at this string, it uses the objectAtIndex: method [2.13].
//[2] #import <foundation/foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; NSMutableArray *myArray = [NSMutableArray array]; [myArray addObject:@"first string"]; [myArray addObject:@"second string"]; [myArray addObject:@"third string"]; NSString *element = [myArray objectAtIndex:0]; [pool release]; return 0; } // [2.13] NSLog(@"The element at index 0 in the array is: %@", element);
You'll often have to step through an array in order to do something with each element of an array. To do that, you can use a loop construct like in the following program which prints each element of the array along with its index:
//[3] #import <foundation/foundation.h> int main (int argc, const char * argv[]) { NSAutoreleasePool *pool = [[NSAutoreleasePool alloc] init]; NSMutableArray *myArray = [NSMutableArray array]; [myArray addObject:@"first string"]; [myArray addObject:@"second string"]; [myArray addObject:@"third string"]; int i; int count; for (i = 0, count = [myArray count]; i < count; i = i + 1) { NSString *element = [myArray objectAtIndex:i]; NSLog(@"The element at index %d in the array is: %@", i, element); } [pool release]; return 0; }
Note that arrays are not limited to contain strings. They can contain any object you want.
Arrays
13:3
Become An Xcoder
The NSArray and NSMutableArray classes provide many other methods, and you are encouraged to look at the documentation for these classes in order to learn more about arrays. We'll end this chapter by talking about the method that allows you to replace an object at a given index with another object. This method is named replaceObjectAtIndex:w ithObject:. Up until now we have only dealt with methods that take at most one argument. This one is different, and this is why we are looking at it here: it takes two arguments. You can tell because its name contains two colons. In objective-C methods can have any number of arguments. Here is how you can use this method:
//[4] [myArray replaceObjectAtIndex:1 withObject:@"Hello"];
After executing this method, the object at index 1 is the string @"Hello". Of course, this method should only be invoked with a valid index. That is, there must already be an object stored at the index we give to the method, in order for the method to be able to replace it in the array by the object we pass.
Conclusion
As you can see, method names in Objective-C are like sentences with holes in them (prefixed with colons). When you invoke a method you fill the holes with actual values, creating a meaningful "sentence". This way of denoting method names and method invocation comes from Smalltalk and is one of Objective-C's greatest strengths, as it makes the code very expressive. When you create your own methods, you should strive to name them in a way that they form expressive sentences when called. This helps make Objective-C code readable, which is very important in keeping your programs easily maintainable.
13:4
Arrays
Become An Xcoder
Garbage Collection
The memory management techniques used by Cocoa and introduced later in this chapter are commonly known as Reference Counting. You will find complete explanations of this Cocoa memory management system in more advanced books or articles (see Chapter 15). Mac OS X 10.5 Leopard introduces a new form of memory management to Objective-C 2.0, known as Cocoa Garbage Collection. Garbage collection manages memory automatically, removing the need to explicitly retain, release or autorelease Cocoa objects. Garbage collection's magic works on all Cocoa objects that inherit from NSObject or NSProxy, and allows a programmer to simply write less code than in earlier versions of Objective-C. There's not much more to say about it, in practice. Forget everything you learned in this chapter!
Memory Management
14:1
Become An Xcoder
Playlist 1
Playlist 1
Playlist 3
An object knows how many times it is referenced, thanks to its retain count
In order to decrease the retain count of an object, all you have to do is to send the object a release message.
[anObject release];
Autorelease
Cocoa also offers a mechanism called the "autorelease pool" which lets you send a delayed release message to an object - not immediately, but at a later time. To use it, you just have to register the object with what is called an autorelease pool, by sending your object an autorelease message.
[anObject autorelease];
The autorelease pool will take care of sending the delayed release message to your object. The statements dealing with autorelease pools that we have seen previously in our programs are instructions we give to the system in order to correctly set up the autorelease pool machinery.
14:2
Memory Management
Become An Xcoder
Sources of Information
15:1