PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION WITH C Revised Edition Alistair Moffat Department of Computing and Information Systems The University of Melbourne, Australia CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPearson Originals is an imprint of Pearson Australia, a division of Pearson Australia Group Pty Ltd. It has been established to provide academics throughout Australia and New Zealand with fast and efficient access to the printing, warchousing and distribution services of Australia’s leading educational publisher, ensuring « smooth supply channel to your campus bookseller. For more information about the Pearson Originals service, contact the Editorial Department, Pearson Australia, Unit 4, Level 3, 14 Aquatic Drive, Frenchs Forest, New South Wales, 2086. Telephone: 02 9454 2200, © 2013, 2003 by Pearson Australia (a division of Pearson Australia Group Pty Ltd) ‘The Copyright Act 1968 of Australia allows a maximum of one chapter or 10% of this book, whichever is the greater, to be copied by any educational institution for its educational purposes provided that that educational institution (or the body that administers it) has given a remuneration notice to Copyright Agency Limited (CAL) under the Act. For details of the CAL licence for educational institutions contact: Copyright Agency Limited, telephone: (02) 9394 7600, email: info@copyright.com.au Al rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording, or otherwise, without written permission of the publisher. Managing Editor: Jill Gillies Solutions Consultant: Claire Sadler Senior Project Manager: Lise D’Cruz Production Controller: Julie McAsthur Digitally printed in Australia by SOS Print + Media Group ISBN 9781486010974 ISBN 9781486010981 (Vital Source) Pearson Australia 4, Level 3,14 Aquatic Drive Frenchs Forest NSW 2086 ‘www pearson com 2u Every effort has been made to trace and acknowledge copyright. However, should any infringement have occurred, the publishers tender their apologies and invite copyright owners to contact them. Due to copyright restrictions, we may have been unable to include material from the print edition of the book in this digital edition, although every effort has been made to minimise instances of missing content. PEARSONContents Preface vii 1 Computers and Programs 1 LL Computers and computation ©... 60.0.0 vee eee 2 1.2 Programs and programming 3 13° AfirstC program... . . 6 14 The task of programming . 9 15 Becareful 10 Exercises 12 2 Numbers In, Numbers Out B 21 Woentifiers 2... ccs eccccean eee EDEL DEER 3 2.2 Constants and variables... . . . ecen ces, 23° Operators and expressions... . 3 19 24 Numbers in 3 dias G60 im Bae oe aes a1 25 Numbers out... . . 2 2.6 Assignment statements i 23 2.7 Case study: Volume of a sphere . 25 Exercises ae a's 4 26 3. Making Choices 29 3.1 Logical expressions ......... . . * 3.2 Selection ....... 32 3.3. Pitfalls to watch for . : 3 34 Case study: Calculating taxes . . 36 3.5. The switch statement 38 Exercises 2.2.2.2... 40 4 Loops 45 4.1 Controlled iteration ......... be . 45 4.2. Case study: Calculating compound interest 49 43. Program layout and style 50 44 Uncontrolled iteration... . . . 52 45° Iterating over the input data . . . 56 Exercises ......... ee 59 CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPace iv PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION 5 Getting Started with Functions 63 S.1 Abstraction. . a 5.2 Compilation with functions... 66 5.3. Library functions 10 54 Generalizing the abstraction 2 55 sean ve . 4 5.6 Case study: Calculating cube roots... . . . « anes (76 5.7 Testing functions and programs . . . : B Brorcleed ow ccc cee eee ee ee ee eee 79 6 Functions and Pointers 83 6.1 The main function. . . 83 62 Useofvoid . 84 63 Scope ......... 85 64 Global variables.......... sane ABE 65. Static variables 2 89 6.6 Pointers and pointer operations ©... eevee ee eee 90 6.7 Pointers as arguments......... SUR ae Wee Aare 408 6.8 Case study: Reading numbers 95 fiixercleec) «Gr 6 2s coekeeee a bor OSs ae: 96 7 Arrays 99 7.1 Linear collections of like objects... ee eee ee 99 7.2. Reading into an array 101 7.3. Sorting an array... . 103 7.4 Asrays and functions . 105 7.5 Two-dimensional arrays 109 7.6 Array initializers 112 7.7 Asrays and pointers . . 3 7.8 Stings... . = 116 79. Case study: Distinct words... 121 7.10 Acrays of strings 123 7.11 Program aaa - 124 Exercises ; 126 8 Structures 129 8.1 Declaring structures... ...... 129 8.2. Operations on structures = 131 8.3 Structures, pointers, and functions. 138 8.4 Structures and arrays 137 Bamciess 25.2.5. eb ou ont oe eeeeses 198 Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonPace v 9 Problem Solving Strategies 141 9.1 Generate andtest .. 2... 141 9.2 Divide and conquer . . 142 9.3 Simulation... . ae 147 9.4 Approximation techniques. - 152 9.5 Physical simulations 156 9.6 Solution by evolution .............5 aeeues (59 Erercises 202... . 0.02. : : 160 10 Dynamic Structures 163 10.1 Runtimearrays... 2.2... 163 10.2 Linked structures . . . 170 10.3 Binary search trees. . 176 10.4 Function pointers . . . pee enman nee 19D 10.5 Case study: A Polymohic tree library oe tery HES: Exercises 189 11 File Operations 193 111 Text files. . . . 193 11.2 Binary files . Eee ete 15 11.3 Case study: Merging ‘multiple aes eae . 199 fBixerleee) (2. 5 Gee eee = ee ee eee 202 12 Algorithms 203 12.1 Measuring performance... .. . « tee eee ee 203 12.2 Dictionaries and searching... . sane $5 ~ 205 12.3 Hashing... 207 12.4 Quick sort . 212 12.5 Merge sort 218 12.6 Heap sort. . 220 12.7 Other problems and algorithms - ines sews, OS Exercises 0.22 vente eee 226 13 Everything Else 229 13.1 Some more C operations ...... . ee 229 13.2 Integer representations and bit operations 230 13.3 The C preprocessor . . 235 13.4 What next? ..... eee ere ee cr seed Se ee kos ans six 290) Index 241 Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon‘Copyright © Pearson Ausbata(edivsin of Pearson Australia Group Pty Li) 2010 ~ 976406010674 -MullevProgramming, Problem Soiving, nd Abstraction wih C Revised ElionPreface to the Revised Edition ‘When I commenced this project in 2002, I was motivated by twenty years of teaching programming to first-year university students, watching their reactions to and behav- ior with a range of texts and programming languages. My observations at that time Jed to the following specification: * Length: To be palatable to undergraduate students, and accessible when re- ferred to, a programming text must be succinct. Books of 500+ pages are daunting because of their sheer size, and the underlying message tends to get lost among the trees. I set myself a length target of 250 pages, and have achieved what I wanted within that limit. For the most part I have avoided terseness; but of necessity some C features have been glossed over. I don't think that matters in a first programming subject. Value for money: Students are (quite rightly) sceptical of $100 books, and will often commence the semester without owning it. Then, if they do buy one, they sell it again at the end of the semester, in order to recoup their money. I sought, to write a book that students would not question as a purchase, nor consider for later sale. With the cooperation of the publishers, and use of the “Pearson, Original” format, this aim has also been met. © Readability: More than anything else, I wanted to write a book that students would willingly read, and with which they would engage as active leamers. ‘The prose is intended to be informative rather than turgid, and the key points in each section have been highlighted, to allow students to quickly remind themselves of important concepts. © Practicality: I didn’t want to write a reference manual, containing page upon page of function descriptions and formatting options. Students learning pro- gramming for the first time instead need to be introduced to a compact core of widely-applicable techniques, and be shown a pool of examples exploring those techniques in action. The book I have ended up with contains over 100 ‘examples, each a working C program. ‘© Context: | wanted to do more than describe the syntax of a particular language. also wanted to establish a context, by discussing the programming process itself, instead of presenting programs as static objects. [ have also not shirked ‘from expressing my personal opinions where appropriate — students should be CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPaGe vill PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION ‘encouraged to actively question the facts they are being told, to better cement their own understanding. * Excitement: Last, but certainly not least, 1 wanted a book that would enthuse students, and let them see some of the excitement in computing. Few programs ‘can match the elegance of quick sort, for example. ‘Those thoughts led to the first edition, finalized in late 2002, and published in early 2003. Now it is nearly 2013, and another decade has gone by. I have used this book every year since then in my classes, and slowly built up a list of “I wish I hadn't done ‘it that way” issues. Those “I wishes” are all addressed in this revised edition. Most of the changes are modest, and I think I have remained truc to the original goals. (One of the more interesting changes is that I have removed all mention of floppy disks!) How to use this book In terms of possible courses, this book is intended to be used in two slightly different ways. Students who are majoring in non-compating disciplines require C program- ‘ming skills as an adjunct rather than a primary focus. Chapters 1 to 8 present the core facilities available in almost all programming languages. Chapter 9 then rounds out that treatment with a discussion of problem solving techniques, including some Jarger programs, to serve as models. There are also six case studies in the first nine chapters, intended to provide a basis on which the exercises at the ends of the chap- ters can be tackled. Fora service course, use Chapters 1 to 9, and leave the more able students to read the remainder of the book on their own. Chapters 10 to 13 delve deeper into the facilities that make C the useful tool that it is, and consider dynamic structures, files, and searching and sorting algorithms They also include two more case studies. These four chapters should be included in ‘course for computer science majors, either in the initial programming subject, or, as. we do at the University of Melbourne, as a key component of their second subject. Interms of presentation, Iteach programming as a dynamic activity, and hope that you will consider doing the same. More than anything else, programmers learn by programming, in the same way that artists learn by drawing and painting. Art students also Jearn by observing an expert in action, and then mimicking the same techniques in their own work. They benefit by seeing the first lines drawn on a canvas, the way the paint is layered, and the manner in which the parts are balanced against each other to make a cohesive whole. ‘The wide availability of computers in lecture theaters has allowed the introduc- tion of similarly dynamic lessons in computing classes. By always having the com- puter available, [ have enormous flexibility to show the practical impact of whatever topic is being taught in that lecture. So my lectures consist of a mosaic of prepared slides; pre-scripted programming examples using the computer; and a healthy dose of unscripted exploratory programming, With the live demonstrations I am able to let the students see me work with the computer exactly as I am asking them to, including making mistakes, recognizing and fixing syntax errors, puzzling over logic flaws, and halting infinite loops. CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPREFACE PAGE IX ‘The idea is to show the students not just the end result of a programming exercise as an abstract object, but to also show them, with a running commentary, how that result is achieved. That the presentation includes the occasional dead end, judicious backtracking and redesign, and sometimes quite puzzling behavior, is all grist for the mill. The students engage and, for example, have at times chorused out loud at my (sometimes deliberate, sometimes inadvertent) introduction of errors. The web also helps with this style of teaching — the programs that are generated in class can be made accessible in their final form, so students are free to participate, rather than frantically copy. Running a lecture in this way requires a non-trivial amount of confidence, both to be able to get it right, and to deal with the consequences of sometimes getting it ‘wrong. More than once I have admitted that I need to go and read a manual before ‘coming back to them in the next class with an explanation of some obscure behav- ior that we have uncovered. But the benefits of taking these risks are considerable: “what will happen if...” is a recurring theme in my lectures, and whether it is a thetorical question from me, or an actual interjection from a student, we always go to the computer and find out. Supervised laboratory classes should accompany the lecture presentations. Stu- int main(int arge, char vargv{]} { printf (*Hello world! \n"); return 0; ‘The program in the box doesn’t do much, nevertheless, it is always a useful task to get this program working on any new computer system you encounter, and (0 also ‘write the equivalent program in any new language you must master. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon1.3 AFiRsTC PROGRAM PAGE 7 ‘There are several points to note. First, the text between the / and «/ pair is discarded by the C system, and is purely for the benefit of any human readers — it is a comment. Comments can be placed at almost any point of a program, and once the /+ is read, all further text is discarded until a «/ combination is encountered. Unfortunately, comments cannot appear within other comments there is no sense of nesting. Later C standards also allow a form of comment where any text that follows .// pair in a line is disregarded, but this is not legal in ANSI (C90) C programs. Programs typically commence with a comment that records the author of the program, a history of any modifications to the program and a set of associated dates, and a summary of the operation of the program. For brevity the programs shown in this book contain relatively terse commenting, and you should be more expansive in the software that you write. Second, most of the rest of the program is a kind of standard recipe that is used without discussion for the next few chapters — the #include line and int main lines are going to appear exactly the same way in every program, as are the zetusn statement and final closing brace. In fact, the only interesting part in this program is the print £ line, which says that the sequence of characters ~ or string — e110 world! is to be written to the ‘output. The next box shows a possible interaction with this program on a computer running the Unix operating system, assuming that the program’s lines have been typed into a file called hel Loworid.c using a program known as an editor. The word mac: is the prompt from the computer's operating system, and indicates that user input is expected. The commands beside each prompt (1s, which lists the files in the current directory, and gcc, which compiles the program) were typed by an imaginary user; the other lines resulted from the execution of those commands, mac? 18 helloworld.c mac: gcc -all -ansi -o hellovorld helloworld.c mac: 1s helloworld helloworld.c mac: ./helloworld Hello world! ‘The gcc command compiles the C source code in file nelloworid.c using the C compiler distributed by the Free Software Foundation’, and creates an executable file called he1 Lowor1¢ that contains machine-language instructions corresponding to the C source program. On a Windows system the executable would have a .exe filename extension, but in the Unix and MacOS contexts it is conventional to cre- atc exccutables that have no extension. All of the examples in this book presume a Unix/MacOS environment, as shown in the example. The last command executes the compiled program, and the “Hello world!” output message appears. Figure 1.2 shows a second C program, and an execution of it. This one reads a set of numbers, and calculates and prints their sum. It is presented to give you a feel for what a more substantial program looks like, and you are not expected to have *See http: //www.gnu.org. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019.Pace & PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION reads and adds a set of values. 7+ R second C program / include int main(int argc, char vargvfl} ( int sux; /+ the running oum +/ int next; /+ the next value to be added +/ sum = 0; /* get the numbers one by one «/ while (scanf("ta", énext) == 1) { sun = sumtnoxt; ? /+ and print their eum +/ print£(*The sum of the numbers is td\n", sum); return 0; mac: gcc “Wall -ansi -o addnumbers adcnumbers.c mac: ./addnunbers 345648 256984 212345, 634534 The cum of the numbers ie 2049511 Figure 1.2: A program that reads numbers, and calculates and prints their sum. The lower ‘box shows an example compilation and execution of the program. yet mastered the intricacies of how it was put together (so please don’t panic!). But with a bit of luck, by reading the code - and the comments — you can identify the basic building blocks: some variable declarations; a while loop that gets the input numbers one by one into variable next, and adds them onto a running total sum that is at first set to the value zero; and a print that prints out the value of sum. Most programs, including the one in Figure 1.2, contain these components: © Comments, which help human readers understand the program; ‘© Declarations, which tell the compiler what variables are being used in the pro- gram, and what their types are; ‘* Assignment statements, which cause arithmetic to be carried out, and the val- ues of variables to be modified; ‘© Control structures, witich direct the flow of execution, perhaps depending upon the values of the variables; and ‘Copyright © Pearson Ausbata(edivsin of Pearson Australia Group Pty Li) 2010 ~ 976406010674 -MullevProgramming, Problem Soiving, nd Abstraction wih C Revised Elion1.4 PROGRAMMING Pace 9 1/0 statements, which get values into the program (the program input), and then back out again (the program outpu:).. ‘Look at Figure 1.2, and see if you can decide which of these categories each statement falls into, The next few chapters discuss these various types of statements one by one, and give dozens more examples of their use. ‘The principal features in most programs are comments, declarations, assignment statements, control structures, and /O statements. Because different operating systems require different interactions, from now on examples usually show only the C source code for programs (or fragments of pro- grams) and their output, without giving details of the compilation process or the file names involved. 14 The task of programming How do we go about the task of writing programs? An analogy that might help is that earning programming is like learning how to ride a bike. One can study physics and know of angular momentum, but the ability to balance is quite independent of that knowledge, and only comes after (for most people) scraped knees and tears of frustra- tion. Similarly, it is possible to study the properties of a programming language in the abstract, but successfully conceiving and implementing a working program requires more than purely abstract skills. The moment of truth as they learn to “balance” is a watershed point for most students learning programming. ‘The single most important thing is to practice — exactly the same requirement as when leaming to ride a bike. You can’t develop the necessary skills if you never get around to writing small programs, on your own, or working with a friend. And you ‘can memorize this entire book, but unless you have spent time in front of a computer making mistakes, you won't have the practical skills necessary to complement your academic knowledge. ‘The basic steps in writing a program are always: ‘* Think about the problem, and decide how you, as a human “computer”, would tackle it. Write down your proposed solution as a sequence of steps using a natural lan- guage such as English. # For each step, decide how it should be solved, breaking it down into smaller sub-steps. © When the steps are straightforward enough, convert each into instructions in the programming language. Use an editor to create a file containing your program. Start with a simple skeleton, inserting as few lines as possible, and including only a small fraction of the eventual functionality. CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPace 10 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION ‘ Use the compiler to check the syntax of what you have typed, and to convert your instructions into the very simple repetitive commands that the computer actually executes. © Once it is syntactically correct, test your partial program on a range of data. # To fix errors, go back as far as necessary in the process. © Then add the next step, or unit of functionality, and start the testing process again. In general, the more time spent at the beginning of this process, the less time will need to be spent at the end. Carpenters don’t saw much wood on the first day of a house construction project ~ they spend that day making plans, knowing that in the end the time invested will be more than repaid. When they do start sawing, they know exactly what piece they are cutting, where it is going to go, and what order they are going to install the various functional components of the house ~ framework, electrical, plumbing, exterior wall cladding, interior wall linings, painting, and so on. ‘A good programmer does the same. Note also that you can test a partial program and make sure that the parts that have been implemented so far carry out the desired tasks, even when not all of the functionality has been incorporated. We don’t write programs as monolithic wholes, and then start testing when they are too big to manage ~ that is as silly as supposing that when eating pizza you should stuff three slices into your mouth before starting to chew and swallow. Little kids at birthday parties do that with lollies, but always choke in the end. Adults know to get each mouthful dealt with before embarking on the next. The same “bite, chew, swallow” routine is just as important when constructing program. The following “hacker’s motto” neatly summarizes this process: ‘A day of debugging can save an hour of planning. Another useful motto to remember is this next “KISS” one. It says that the best programs are the ones that carry out the desired transformation, from input to output, in a simple and straightforward way, without being cluttered by excessive function- ality or dubious features: Keep it simple, stupid. 15 Be careful ‘We end this first chapter with a stern warning about computers, and their propensity to fail. Yes, they are enormously reliable and accurate, often running for months or years without hardware failure. But they do eventually fail, and if the user has been naive, that failure can be heart-breaking. For a home computer, the value of the hardware may be ten times the value of the software and data. You spent perhaps $1000 on the computer, and have purchased a few hundred dollars worth of games. In this case, the actual hardware of the computer is easily replaceable when it reaches the end of its life, and there is relatively little to worry about. CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonLS BE CAREFUL, Pace 11 But if you have used that computer as an authoring or storage device ~ for exam- ple, because you are working on a programming assignment, or writing a book, or ‘managing the accounts for a small business, or storing your holiday photos — what is stored quickly becomes more valuable than the hardware it is stored on. It would not be unusual for a home computer used in this way to contain data that is worth ten times the cost of the hardware. If this is the case, you need to be careful — los- ing $10,000 worth of data is very upsetting. For a business computer, in a small or medium-sized enterprise, the value of the data could be ten times the value of the programs that manipulate it, and one hundred times the value of the hardware. That $1,000 computer may well be storing $100,000 worth of work orders, invoices, and overdue accounts. So it makes sense to be careful with your data. For you, your “data” might only be your programming work, plus some snapshots from your phone, plus some music. But even so, keeping backups should be regarded as fundamental good discipline. Somehow, the hardware always knows when you have become lazy with your backup routine, and fails at the worst possible instant. Losing a week of work because of hardware failure is stupid. Losing a month of work because of hardware failure is unforgivable. And the truth is, losing more than a couple of hours of work is completely avoidable. ‘What does this mean? Unless you have a permanent high-speed intemet con- nection, the best way to manage backups of media files (photos and videos that you have created) is to either regularly burn DVDs of files that have been added to your ‘computer, or copy them on to a removable thumb drive. Keep a careful record (in a file or even a notebook) of what files you copied and when, and use more than one thumb drive. Indeed, best practice would be to alternate between at least two thumb drives, and keep one of them somewhere other than on the desk beside your com- puter. And if you are authoring documents, such as programs and essays, every hour or two of editing time on your home computer, you should similarly make a copy of the files onto backup media; or copy them over your internet link into a cloud-based storage service. Then, the inevitable day your computer refuses to boot, and just says “Hard Disk Error’, you will be enormously glad that you have reasonably up-to-date versions of your files available. ‘As a second level of security, once a month or so you should make a complete copy of all your files (everything you have created, not just the ones you have been working on recently) onto removable disk drive, carefully label and write-protect it, and store it right away from your computer, in your locker at school, or at a friend's house. This kind of off-site backup doesn’t have to be quite as regular as the routine copying of files off the computer, but is no less important ~ a fire in your house is ‘twaumatic enough without losing a year’s worth of assignment work, even when you have carefully carried out the first level of the backup regime. Keep up this routine whenever you are working on the computer actually creating information. It is easy enough to buy a fresh copy of some game program, or a music track you can't live without, But no amount of money could, for example, resurrect the photos you took in Nepal, climbing mountains; or the effort spent writing a book such as this one. CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPage 12 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION ‘Time spent working on a computer is a resource far more precious than the hardware itself. To minimize the impact of hardware failure, you should adopt a regular backup routine. ‘If you are using a shared computer in a professional installation, then the backups should be done centrally for you, and off-site storage arranged as part of a proper dis- aster recovery plan. But you can still take some responsibility — why not periodically make a copy of your files, and load them onto your home machine? Just in case the backup regime at the central computer isn’t being done regularly? Gnashing your teeth, and calling for the resignation of the system programmer who didn’t check that the backups were working properly cannot absolve you from the responsibility of be- ing careful with your own files. If you always act as if the worst is about to happen, ‘you will be better prepared for the inescapable moment when the worst does happen. Exercises 1.1 Ask an older friend or relative about the first computer they ever bought. Try to pin down the year, the price, and some of the physical characteristics of their machine ~ disk space, memory space, and processor speed. Then look at some newspaper or magazine advertisements, and find the same values for amodem computer. Once you have some “then” and “now” figures, calculate the annual com- pound growth rate on performance that they represent. If you want to make ‘an accurate assessment, you probably need to allow for inflation too, but ig- noring it won't affect your answer by much. 1.2 Find out how to use the editor on your computer to create a file, and type in the “Hello World” program. Compile and execute it. 1.3 Study the program in Figure 1.2. Without trying to write any detailed C in- structions, discuss with a friend what would need to be added if the average (mean) of the input numbers was to be printed as well as their sum. ‘What do you think the original program does if no numbers are typed? What should the extended program do if no numbers are typed? 1.4 Suppose that you get paid $18 per hour to do some work using a computer ~ perhaps writing a book, or doing some programming. Suppose that you spend threc hours a day working on that task, five days a week. After one year of effort, what is the value of the data stored in the computer? ‘And after three years? How does that compare with the cost of a computer? CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonChapter 2 Numbers In, Numbers Out This chapter introduces the three elemental components of any C program: reading numbers from the input; doing some calculation on them; and then printing numbers back to the output. But first you need to know about the rules for forming identifiers (Section 2.1), and about constants and variables (Section 2.2). 2.1 Identifiers An identifier is a name used in a C program to represent a value. The rules for forming valid identifiers are relatively simple: © ‘They must begin with a letter or underscore (‘*.") character; They can then use any combination of letters, digits, and underscore; and * They may not contain any other characters. For example, all of date and temperature and mean and big_long-name and nitems and —funny are valid identifiers. ‘These ones are not: fast-food, because it contains a hyphen “-" character (and. ‘as the hyphen represents subtraction, the C systems interprets it as being fast take- away food); 76trombones, because it starts with a digit; and # of words, because ‘it contains a hash (““#”) character. You are also free to use identifiers like chad£t 76a£M and £5g6h7 8k9 in your programs. Using names like these does not affect the execution of your program in any way, and the C system doesn’t care what names you use. On the other hand, use of non-mnemonic strings makes it harder for human readers to browse your program ~and it is reasonably likely that the next human who has to try and understand your ‘work will be you, in a week, a month, or a year. So it pays to use appropriate variable names. Use meaningful identifiers, so that names reflect the quantities being manipulated. ‘The exception to this advice is where an identifier is used extensively over a short section of your program, in which case it is usual to have names like i and j. CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPage 14 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION ‘There is a set of words that cannot be used as identifiers, because they have special significance in the C system. These include words like int and while, both of which have already been used in Figure 1.2. Several more such reserved words will be encountered in the next few chapters. Reserved words have special meanings in C, and may not be used as identifiers. 2.2 Constants and variables Some numbers used in programs are constants — fixed values that do not (and should not) change during the course of execution. C offers a useful facility for assigning symbolic namesto these: the #def ine. To define a symbolic constant, pick a suitable ‘identifier, and make sure that the “#" is the first non-blank character in each line: define SEC_PER_MIN 60 #define Nr} fcefine EXAM_PERCENTAGS 60 fdefine PROJECT_PERCENIAGE (100 - EXAM PERCENTAGE) #detine KMPER MILE 1.609 Note that you should not put a semi-colon at the end of these lines. Note also how in this example there are several different constants that all have the same value. If a different symbolic name is used for each logically distinct quantity, even when they are numerically the same, it is much easier to correctly modify the program. For example, if the EXAM PERCENTAGE was to be changed from 60 to 70, and symbolic constants were not being used, there is a very real possibility that in hunting through ‘your program looking for occurrences of 61 to be changed you will end up with some hours containing 70 minutes, or some minutes containing 70 seconds. Note also that constants can depend upon the value of other constants: PROJECT.PERCENTAGE is definitely constant, but with a value that is not explicitly given. A change of EXAM PERCENTAGE to 70 automatically changes PROJECT PERCENTAGE too, There is no rule that says that symbolic constants should be all-uppercase letters, and any identifiers can be used. Nevertheless, use of uppercase constants is a widely followed custom, and is the style adopted in this book. It is also both permissible and sensible to use constants for any fixed character strings used in your program: #define ERROR M "values must be greater than zero Strings are dealt with in more detail in Section 7.8 on page 116. Use symbolic constants for all fixed values manipulated in your programs, to improve readability, and to facilitate subsequent modifications. ‘One extreme view is that the only permissible raw constants in your programs should be zero and one, and then only when they are being used as the additive and Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon2.2. CONSTANTS AND VARIABLES Pace 15 multiplicative identities respectively. You might not want to take it quite this far, but if you directly code any fixed value into your program, pause for a minute and consciously decide whether or not in a years time it will look like a “magic number”, and if so, give it a symbolic name. The other values manipulated in a program are variables. In mathematics, vari- ‘ables are used to represent unknown, but fixed, quantities. In a program, variables represent values that change. Consider, for example, the statement sum=sum+next in Figure 1.2 on page 8. Writing the similar s = s +n in mathematics as an as- sertion implies that n must be zero, and that s can be any value. In Figure 1.2, the statement sum~sum#next: is a directive that means: “take the current value stored in ‘the variable called sum, add to it the current value stored in the variable next, and then store the result back into sum”. At the end of that sequence, variable sum has an altered value (presuming next is non-zero); and sum changes again if the statement is repeated. That is, at any point while the program is executing, sum has a value that is determined as an exact function of the sequence of numbers entered by the program’s user. ‘Combinations of variables and constants are calculated as expressions, and their values are preserved into the same or other variables using assignment statements. ‘Here are some more assignment statements that make use of the constants that were defined earlier. Each assignment presumes that the various right-hand-side variables have been given suitable values: nitens = nitenstl; miles = kn/KY_PER MILE; tot_seconds = (hours*MIN_PER_HOUR + mins) +SEC_PER MIN + finel_mark = (exam mark+EXAM_PERCENTAGE + project_mark«PROJECT_PERCENTAG: Note the use of a semi-colon to terminate each of the calculations. This is how C knows where one statement ends and the next begins. Note also that none of the white-space layout — achieved through the judicious insertion of blank, newline, and tab characters — is significant in C. Long statements should be broken over scveral ines in a neat and tidy manner. Observe how helpful it is to use sensible identifiers — ‘well-written program text can be read almost as if it were prose. Semi-colons are used to terminate statements. Each variable used in a program must be assigned a ‘ype in a preliminary state- ‘ment called a declaration. In Figure 1.2 on page 8, sum and next are both declared tobe integer-valued variables by the initial line int sum, next. Similarly, all of km, miles, tot seconds (and so on) need to have been declared prior to the assignment statements given in the previous example. But it is probably not sensible to declare km and miles to be of type int, as they might not be integer-valued. A range of other types are described shortly, including one suitable for km and mites. Each variable in a program has a type associated with it. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonPace 16 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION ‘One common misconception is that declaring a variable also gives it an initial value. This is not the case. The declaration indicates to the compiler that the variable exists, but does not give it any particular value. That must be done by an assignment statement, which is why sum is explicitly set to zero in the program in Figure 1.2 on page 8. C does permit one small shorthand to reduce the need for duplicate han- dling, and variable declarations and an initial assignment can be combined in a single statement if desired: This statement both declares the integer variable oun, and sets it to zero. Variables must be assigned values before they can be used in expressions. If a variable is used without an initial value being assigned, the results are un- predictable. If you are unlucky, the system assigns zero or some other initial value, and your program works. If you are lucky, your program fails, and you are forced to investigate why. This second path is definitely the more desirable one - it is better to find out early that a program is erroncous, than to use it for many years, then one day transfer it to a different computer and have it mysteriously fail. ‘The type of a variable determines how its value is represented in the computer, and how operations on that variable are carried out. Integer variables store exact values, but only within a somewhat limited range. ‘On the majority of current computers the allowable range of integers is from —2°! = —2,147,483,648 to +28! — 1 = +2,147,483,647, values determined by the use of 32 bits of memory for each int variable. Computations that result in integer values outside this range result in incorrect answers, with no warning whatsoever. This unfortunate fact is demonstrated by the next program fragment: int big, bpl, bt2, bplt2; big = 2147483647; bpl = bigtt; be2 = bigh2; bplt2 = ppl+2; print£(*big=8d, bpl=td, bt2=8d, bplt2=4d\n", big, bpl, bt2, bplt2); In this program, a number of assignment statements are executed in the order they are written, and then each of the “sa” fields in the print statement is replaced by the next corresponding value in the list of variables that follows. (The print £ statement is considered in more detail shortly.) When executed on a standard 32-bit computer, the program fragment produces this output line: [ pigr2is7» multiplication 5.0*3.0 15.0 543.0 15.0 543 15 / division 19.0/4.0 4.75 19.0/4 4.75 19/4 4 % remainder after division 19.0%4.0 illegal 19.084 illegal 1984 3 + addition 7,543.0 10.5 7.543 10.5 143 10 — negation = subtraction ‘Table 2.1: Operations on numeric values. When both operands to the binary operators are of type int, the result is always of type inc. Ifeither or both of the operands are of type Float or double, the result is of type double. The modulus operator “'” can only be applied if both arguments are of type int. If both operands to a division are of type int, the result is the truncated integer quotient. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon2.4 NUMBERS IN Pace 21 print £(°17.0/0 = ¥.3£\n", 17.0/0)7 print £(*1.0/(17.0/0) = ¥-3£\n", 1.0/(17.0/0)); print£ ("sqrt (-1.0) = 8.3£\n", sqrt (-1.0)); print£(*1.0/sqrt (-1.0) ~ %.3f\n", 1.0/sqrt (-1-0))5 print£(*17/0 = ¥d\q", 17/0); When executed on one computer, the output was: 17.0/0 = int 1.0/(17.0/0) = 0.000 sqrt (-1.0) = nan 1.0/sqrt (-1.0) = nan Floating exception (core dumped) Note how, in this example, the second division 1/inf results in the calculation of zer0; while 1 /nan continues to be nan. It also remains nan when squared or multi- plied by ing. If your program is terminated, you may see the rather unhelpful error message “Floating exception (core dumped)”. The reference to “core dumped” is a link back to the early days of computers, when the internal memory in which pro- ‘grams and data were stored was composed of small magnetic donuts threaded with wires — core memory. When a program fails like this, a file called core may be written. You should delete coze files if you notice them in your directory. ‘When executed on a second computer, the final divison resulted in a bizarre num- ber being generated, rather than an execution error. Beware. 2.4 Numbers in ‘To be useful, a program must be capable of operating on a range of data, not just the fixed numbers compiled into it. It needs to be able to read numbers typed on the keyboard. In C, the easiest way to read is through the use of the standard function scané. The general form of scané requires a control string and a list of variables: scant (control string , list of variable addresses ) where each component of the control string matches one of the variables, and indi- cates how the data should be stored into that variable. Here are two examples: nt n_pass, class_size, pass_percents coubie mass, velocity, force; scanf("sdtd", un_pass, iclass_size); scanf("S1£S1£81£", 6mass, @velocity, force); Inthe first, wo int variables are to be read, and so the control string is "84%4", with each “sc” a format descriptor indicating that one integer is required. In the second, three double values are read, each controlled by a “$12” format descriptor. Had these variables been of type £1oat, format descriptors “+#” (for “float”, rather than “long float”) would be required. Similarly, “®c” is used to read one character into a variable of type chaz. For scant to operate correctly, the names of the variables must be prefixed by an ampersand character “s”. This indicates the use of the address of the variable rather ‘Copyright © Pearson Ausbata(edivsin of Pearson Australia Group Pty Li) 2010 ~ 976406010674 -MullevProgramming, Problem Soiving, nd Abstraction wih C Revised ElionPage 22 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION than its value; why addresses must be passed is discussed in Section 6.7 on page 93, and for the moment we will simply use the ampersands as part of a standard recipe. In detail, what happens is that characters in the input buffer — from the keyboard, unless Unix input redirection is used — are examined one by one. As cach character is read, it is added to a lengthening value of the type specified in the corresponding format descriptor. That format descriptor — and corresponding variable — is finished when an character that cannot be incorporated is encountered; that character is then the first one used for the next format descriptor, even if the next format descriptor is in a different scans statement. These rules mean that leading whitespace is skipped over for “4a” and “1” descriptors, but not for “8.c” descriptors. As another step in scan£ processing, any additional characters in the control string are also matched, with a blank matching any/all whitespace characters, and so on. For example, the control string “$<. 84#8c” processes the input characters “5 .67###2” (using “#” to show where the space characters are) into the int values 5 and 67, and the char value ' 2". If the control string format descriptors cannot be satisfied by the sequence of characters in the input stream, the scan halts at that point, and the remaining variables are not assigned values. For example, if “sasa%a is given the input “##1234#45. 689” (where “#” is again used to show spaces) the first integer vari- able is assigned the value 123, the second is assigned the value 45, and the third is left unchanged — the period character terminates the second integer, but cannot be used as a component of the third. These rules are reasonably complex. So an important piece of advice is this: always print out the values of variables after they have been read, so that there is at least some chance that any discrepancies (or outright errors) are detected. Another point worth noting is that scan retums as a value the number of objects that were successfully read. So it can also be used as: ant ("4d S1f %c", én, &z, Sc) There were td values success: ly read\n", num ‘There is more that you need to know about scanf before you can be a profi- cient C programmer, but the minimalist treatment in this section is enough to get you operational. Section 4.5 on page 56 returns to scan£ and explains more of its functionality. ‘The function scant is used to read numeric values from the keyboard into program variables. Values read should always be echoed back to the output, to provide validation of the input process. 2.5 Numbers out ‘The examples in this chapter have already made use of the function print £. The general structure is: print (control siring , list of expressions ) Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon2.6 ASSIGNMENT STATEMENTS Pace 23 {In simplest form, output is to the terminal screen or equivalent, a device that is char- acter and line oriented, and often with a default of 80 characters wide. As for scan#, the format descriptors must match the variables: “td” for int- valued expressions; “£” for Eloat~ and double-valued expressions, or “Se” if exponential form is required, for example, 6.67e-11; “tc” for chaz-valued ex- pressions; and “ts” for strings. Note that — confusingly “1” is used only when reading double variables, and they are written using either “S.£” or “se”. More control can be exerted by modifying the format descriptor using a field width: “3c” indicates that at least three character positions should be used, with more consumed if necessary; and “10a” indicates that at Icast ten character posi- tions should be used. In both cases, leading blank characters are inserted if the value being printed requires fewer digits than the number specified as the field width. With doubles and fioats, both the total width in characters and the number of character positions after the decimal point can be controlled: “*5£” indicates that at least five positions should be used; and “8 .2£” indicates that eight positions in total should be used, with the decimal point placed in the third-from-right position, and two characters after it If the character width value is negative, the output string is left justified within a field of the (positive) width. This is particularly useful in conjunction with the “3” format descriptor used to format character strings ~ for example, “2-208” formats a string at the left of a field of total width (at least) twenty characters. When a line is to be terminated and a new line commenced, a "\n” character is written into the format control string. The function print is used to create and structure program output. ‘The example in Figure 2.1 shows a range of formats, with hash characters used so that you can see where the blank characters are. Trace through the details, to be sure that you understand how cach of the lines was generated. ‘One of the surprising things in Figure 2.1 is that char variables can be used in arithmetic. In fact, the type char is a numeric type, and the constant ’ a” equates to the integer 97 by virtue of the character ordering used in almost all computers. Moreover, the lowercase letters form a contiguous sequence, as do the uppercase letters starting at “A” equal to 65. Exercise 4.4 on page 60 explores the properties of this “American Standard Code for Information Interchange” (ASCI) in more detail. 2.6 Assignment statements You have now seen nearly all of the parts needed to write simple programs that read numbers, do calculations, and print out answers. The only bit missing is a proper introduction of the assignment statement: variable = expression ‘There have been several examples of assignment statements in the previous parts of this chapter, and there is nothing more that needs to be said about it at this stage, except to reiterate that the value of the expression on the right -hand-side is calculated, CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPage 24 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION fdefine MARY "Mary, Mary, quite contrary" int main(int argc, char sargv{]) { int n=12345; double x=3.14159; char o~/aly printe("td %3d 86d $-6d\n", n,n, n, n)7 printe("eE %.4f 86.2f %-6.2E\N", x, x, XX; printE("tc Ad Sc\n", cy cy c#3); Print £(*¥s\n¥50s\nt-50s\n", MARY, MARY, MARY); return 0; 12345¢F1 234584 1ZSOSRFIISOSE 3141590883. 141686003.14803.1488 abto7#ed Mary, Mary, tquitetcontrary saenestesesenessereetttemary, Mary, tquitetcontrary Mary, MMary, Fquiteécontrary# #HEEHEEEAEAA FEE ARESRAEE Figure 2.1: Use of different format descriptors in print statements. The symbol used in place of blanks. and then that value is stored into the specified variable. It makes no difference if the Jeft-hand-side variable is used in the right-hand-side expression — it just contributes to the final value that is calculated, exactly the same as any other constant or variable, ‘One convenient shorthand that it worth coming to terms with is the ability to omit repeated variable names, a mechanism that allows the assignment sun=sun+next t0 be abbreviated as sum+=next. This form, with an operator prior to the “=”, can be used for any binary operator: n /=2 divides n by two, and m+=1 adds one to m. The latter of these has its own special shorthand: -++ adds one to variable m. Similarly, -= subtracts one from r. These postincrement and postdecrement forms of the assignment statement are used sparingly in this book, but are common in C programs and you need to be aware of them. There are also preincrement and predecrement ‘operators described in Table 13.1 on page 230 that are not used in any of the programs in this book. ‘The assignment statement allows the value of a variable to be changed. One slight surprise arises with assignment statements: if the expression is of a different type to the variable to which it is being assigned, an automatic cast takes place. This process is shown in the next program fragment: Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon2.7 Case sTupy Pace 25 int nj double 2; char cj 42; 2 = 42; © = 42; print£(*n-83d, 20812.8£, cmbc\n", ny 2, cdi n= 100.99 z = 100.9; ¢ = 100.95 eet 12.8f, omtc\n", n, 2, ©); exte\n", ny 2, cdi 0000000, cma Note how char, int, and double values can all be assigned to each other, without any warming message being generated by the compiler. Note also that the values printed by the “sc” format specifier interpret values as codes in the ASCII character set, hence the asterisk (code 42) and “a” (code 100) values in the final column of the output. The fundamental message to take away from this example is that you ~ the programmer — need to take care to get the types right. 2.7 Case study: Volume of a sphere Using the assignment statement, and the other C skills described in this chapter, you are now able to write a range of straightforward programs. You could also do exactly the same type of computation very easily with a calculator or spreadsheet, but be patient — you will soon sce calculations that can’t be done with a spreadsheet. To confirm that you have leamed these skills, try the following exercise before reading on. Don’t forget to include the standard boiler-plate code at the top and bottom of your program. You can copy that from Figure 1.2.0n page 8. Write a program that reads in the radius of a sphere, in meters, and calculates and outputs the volume of that sphere, in cubic meters. The volume of a sphere of radius r is given by $nr*. Figure 2.2 provides a solution to this exercise. Note the various parts: a comment at the top, followed by boiler-plate code; then declarations for two variables of type double (their values cannot be assumed to be integral); a print £ statement to print ‘2 prompt, so that the user of the program knows when input is expected; a scant statement to read in a value; an assignment statement to compute the desired output value; and a print £ to do the writing. The constant has been given a symbolic name in Figure 2.2. In fact, as will be seen in Section 5.3 on page 70, a set of commonly used numeric values, including 7, are already defined in many C installations, to greater accuracy than is shown here, ‘Section 5.3 shows how to access these predefined values. Finally, note that the program prints the input value too, as was recommended carlier. Doing so makes it more likely that you detect any errors in the scanf that might cause an incorrect value to be read and processed. Whether a scanf terminates prematurely is only partly in your hands — the person using your program also shares Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonPace 26 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION /+ Calculate the volume of a sphere given its radius. “/ include fcerine PI 3.14159 int main(int age, char ¢argv{)) ( Gouble radius, volune; print£(*Enter sphere radius: *); seant("siz", eradivs); volune = (4.0+2I+radius+radiussradius) /3.07 print£(*When radius is 4.2£ meters, ", radius); printf ("volume is %.2£ cubic meters\n", volume); reurn 0; Figure 2.2: Calculating the volume of a sphere. The lower box shows an execution of the rogram in the upper box. that responsibility, and you should not rely on them — but you can at least show what value ended up being used in your program. Programs should always be written presuming that the user will either deliberately or accidentally abuse them. ‘The program shown in Figure 2.2 is typical of programs in the simple “caleu- late the output from the inputs” category discussed in this chapter. A wide range of functional relationships can be evaluated using the same basic template. Exercises 21 Foreach of the following, decide whether or not it is a valid C identifier: —hello kn_per_hour Hel1037 speed! constant-value sensiblesnane subject #1 _num_incorrect_ A big_long_name_many_letters M Ioclock 2.2 Trace the following program fragment, and determine the final values of each of the variables. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonEXERCISES Pace 27 Pa} 24 25: 2.6 Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. a, By c Ge £, oF (atb/2#6) /3; a-bto-dte-£; Given the following declarations, double r, area, perimeter, elapsed hours; int start_hour, start_min, start_sec, finish hour, finish_min, finish_sec write assignment statements that calculate: a. The area of a square of edge length x. b. The perimeter of a square of edge length r. ©. The area of a circle of radius r. 4. The perimeter of a circle of radius =. ©. The time in elapsed hours between the start time (three variables) and the finishing time (three variables) of some event, assuming that the two times are within the same day. Write a simple program that has #include added at the top, and then print out the values of the following constants: INT.MAX, INT.MIW, FLTMIN, FLTMAX, DBL MIN, and DBL.MAK. Do the values on your system agree with the values given in this chapter? ‘Try the calculation 1.23 x 10! + 4.56789 — 1.23 x 10% ona range of com- puters and calculators (including a spreadsheet program). Which device is the most accurate? Which of the following output lines (with hash characters representing blanks in the output) could not have been generated by the print £ statement on the first line? (Note that the combination “*\"” represents a single double-quote character.) print£(*n = 43d, x = 98.42, m= \"$-15a\"\n", ny x, mF 123, #x4=#123 4567, #m#=#"hel 10, #hello##e* 1234, #x-41234.567, dm#=#"hellos sds fas HE” 123, #x#-#-1234.5670, imf#"hollo# fated EET -123, #x#=#-1236.5670, #mameMEREE EEE EH ENOL LON nd-#441, fo=-#48#3.1472, dmé=# "hello, #heLlo, #hel1o" né-#l, txt-#3.1472, ime #"HFHE FERPA EL Lo” 441, fxto8#863.1472, dmé=#"neLL0"Pace 28 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION 2.7 Study this program fragment: int num, nj double 2; char ¢; num = -1; 1 = 99; 2 = 9.99; c= "9"; nun - scanf ("ed t1f 4c", en, &2, &C); peints("num = 4a", num)? printf(", n= 4, 2 = tf, ¢ Se\n", ny 2, 0); For each of these five input lines, determine what the program generates: 12 4.6% 4.612 2 12 4.6 4.6 12 4.6 b 1234x56.6Y Ifyou are unsure, you might wish to compile and execute the program. 2.8 To convert from degrees Fahrenheit to degrees Celsius, you must first subtract 32, then multiply by 5/9. Write a program that undertakes this conversion. Confirm that 212° Fahren- heit maps to 100° Celsius, and that 82° Fahrenheit maps to a little under 28° Celsius. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019.Chapter 3 Making Choices This chapter is about making decisions ~ choosing between altemative paths. We do this all the time in our daily lives: if it is raining we take an umbrella, and if it is sunny we do not; if itis dark we turn on a lamp, and ifit is light we do not; and if we dial a wrong number, we apologize and then dial again more carefully. Inprograms we need to make the same kind of conditional choices — if the number read using a scant is negative, we might need to print an error message; ifthe scant doesn’t read enough values, we might want to repeat it; and so on. In C there are two ‘mechanisms for performing this kind of selection, and they are discussed in detail in this chapter. But first, Section 3.1 introduces a new kind of expression, one that is regarded as being either true or false. 3.1 Logical expressions ‘Table 3.1 lists the standard relational and logical operators available in C. The first group in the table is the relational operators, which are used exactly as you would imagine: to compare two numeric values. Like the arithmetic operations listed in Table 2.1 on page 20, if either of the two operands is a float or a double, the other is cast to match before the comparison is undertaken. This makes sense: the expression “1<1.1", should be true, which would not happen if the double was converted to an int rather than vice versa. Note also —and this is critical — the use of “=” in the equality test rather than “~”. Every C programmer has made the mistake of using an “=” when they really meant “==”. The difference can be disastrous. The table talks about “true” and “false” as if they are values. In fact, in C there are ‘no explicit constants true and false the way there are in some other languages, and the result type of all of the operators in Table 3.1 is int. The rule in C is that the integer value zero is always taken to mean “false” in places where a logical expression is required; and that all non-zero values (including negative ones) are taken to mean “true”. Conversely, when logical values are being generated, if a subexpression is false, the value 0 is produced from that subexpression; and if the subexpression is true, the value 1 is produced. This rather idiosyncratic interpretation means that it is possible to write statements like nun.neg+=(n<0), for example, which adds one to variable num_neg if variable n is negative. CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPace 30 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION Operator Operation True when... < _lessthan first operand is less than second <= lessthan or equal _first operand is not greater than second > greater than first operand is greater than second greater than or equal first operand is not less than second equal to first operand is equal to the second not equal to first operand is not equal to the second. and both operands are true or either operand is tue not the operand is false ‘Table 3.1: Relational and logical operators. As with the arithmetic operators in Table 2.1 on page 20, if either operand is of type Sloat or doub oe, the other is converted to the same type before the operation is applied. But here the result is always of type int. Note very carefully the use of “==" for equality testing, not “=”, which indicates assignment. Operands Operation el_e2 _elese2 eliie2 sel 0 0 0 0 1 0 NZ 0 1 1 NZ 0 0 ii 0 NZ_NZ 1 1 0 ‘Table 3.2: Details of logical operators. Both operands are of type Anz, and are cast if they are not. For the purposes of these operators non-zero (NZ) values are treated as “true”, and zero values as “false”. The three logical combining operators ~ “and”, “or”, and “not” ~ that are listed in Table 3.1 and defined in Table 3.2 may also be familiar. But there is one subtle difference to be aware of — in English, when we say something like “would you like coffee or tea", we don’t expect the answer to be “both”. But in C, a logical “or” expression is true when either of the subexpressions is true, and is also true when both of the subexpressions are true. Given these relational and logical operators, can you work out the circumstances under which each of the following integer-valued expressions is 1 (truc)? cx Sex && x comparison “ equality ae and a) or etc assignment Lowest. ‘Table 3.3: Precedence hierarchy for the arithmetic, relational, and logical operators that have ‘been introduced so far. The three unary operators not, negation, and cast, all have equal precedence. Assignment is also an operator, but unlike the other operators listed, is evaluated from right to left when precedence is equal. These rules are known as De Morgan's Laws, after the mathematician who first demonstrated their use. Relational and logical expressions generate int values that can be interpreted and used as if they correspond to true and false. Having extended the set of operations, it is time to look at precedence again. Table 3.3 lists all of the operators that have been discussed so far, in decreasing precedence order. Using the table, for example, it is possible (but not exactly easy) to work out the value of: 4243 664-5116 748 9/7 10> 21 ‘The answer here is 1. Whether or not a programmer should be rewarded - in any way at all - for writing such an awful expression is an entirely different story. There is no doubt that the best policy is to over parenthesize, especially when mixing arithmetic, relational, and logical operators in the same expression. Including redundant parentheses costs nothing. Omitting necessary parentheses can cost everything. ‘A final point to be noted from Table 3.3 is that assignment is an operator, like any other. The value of an assignment statement is the value that is assigned, and the actual assignmentis a side effect that happens while the expression is being evaluated. Assignment is also one of the few operations in C that is evaluated from right to left. Hence, expressions like ‘n=m=p=q=0" make perfect sense. That one is interpreted = (q=0)))”, and all four variables are assigned zero. More complex expressions are also possible: “n+~n++” is legal, but an absurd thing to write. If you write this kind of expression in your program, you will probably get exactly what you deserve. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonPage 32, PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION ‘The C system is free to determine the order in which the various components of, an expression are evaluated. For example, in the expression a+b+ eed, the system is permitted to evaluate a+b first and then cd, or to do it in the other order. All that the precedence rules require is that the two multiplications take place before the addition. This flexibility means that if any of the subexpressions a, 5, c, or d have side effects —as is automatically the case when any of them involve assignment operators — then the semantics are unclear. You should avoid such confusion in your programs by not trying to be too clever: remember to keep it simple, stupi Beware of the side effects caused by assignment operators inside expressions. Use them cautiously and sparingly. ‘There are two operators for which the ANSI C standard does stipulate the evalu- ation order: in the expression “a &« B” (logical and) you are guaranteed that b will be evaluated only if a is found to be non-zero; and in the expression “a | | 8” (logical or) ‘you are guaranteed that b will be evaluated only if a is found to be zero. That is, these ‘two logical operators are evaluated in a strict ordering, and the second component is evaluated only if its value determines the whole expression. Hence, it is safe to write things like “x!-0sey/x>5”: by the time “y/x>5” is evaluated, you are guaranteed that “x! =0”, so a “divide by zero” error cannot arise during computation of “y /x>5” The evaluation of the logical operators “s«” and “| |” is performed in a lazy left-to-right manner. The second operand is evaluated only if it determines the value of the expression. 3.2 Selection Now for the if statement. The most general form looks like this: 3€ ( guard ) statement] else statement ‘There is a second version that omits the e1se part and looks like this: if (guard ) statement Either or both of statement] and statement2 can be empty statements (no statement at all), single statements, or compound statements bracketed by “(" and ‘*)” pairs, ‘The examples given shortly show what is meant by this. When an i£ statement is executed, the value of the logical expression used as the ‘guard is calculated. If the expression is true (non-zero), then statement! is executed. If the expression is false, and an clse part is given, then statement2 is executed, Figure 3.1 shows this flow of control. The best way to grasp the power of this statement is to look at the examples shown in Figure 3.2. Although the two alternative branches on an i statement each CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised Editon3.3. PITFALLS Pace 33 Figure 3.1: Execution flow through a i £ statement when an ese partis given. nominally contain just a single statement, a group of several statements can also be controlled if they are enclosed in a'“(” and “}” pair to form a compound statement. If Just one statement is being controlled, as is the case in the first example in Figure 3.2, the braces are not needed. Strictly speaking, they are not needed in the last example either, because the comments are not statements, But it doesn’t hurt to have them there, and it is far easier to always put the braces in than it is to decide whether or not they are required; and potentially less confusing to a reader of the program. ‘The i £ statement allows alternative execution paths to be followed, based upon the values of variables during execution. Multiple statements can be controlled if braces are used to create a compound statement. The second example in Figure 3.2 shows how the if can be combined with scans to verify that data is being read correctly. ‘The call to function exit pass- ing the argument EXIT .FATLURE is a directive to “quit executing the program, there has been a serious error”. If execution is to be terminated without signaling an error, the call exit (2XIT-SUCCESS) should be used instead. Both EXIT_FAILURE and EXIT.SUCCESS are defined symbolically in the file std@lib.h. In many operating systems, the argument passed to the exit function is available after the termination of the program, and can be used as a diagnostic as to why the program terminated. ‘The function exit can be used at any point to terminate program execution. Return of a non-zero exit value signals an “abnormal” termination. The final example in Figure 3.2 shows a cascading 4 ¢ statement, in which more than two outcomes are available. Nothing special needs to be done to obtain this functionality, and any number of ‘else i” groups can be strung together. The last else clause catches the remaining cases not dealt with by any of the guards, and each time the overall construct is executed, exactly one of the alternatives is selected. 3.3 Pitfalls to watch for ‘There are a number of pitfalls associated with the use of i statements, usually in connection with the guard. Figure 3.3 shows the most common of these problems. CCopytght © Peersen Austata e divin of Pearson Austala Group Ply Ll) 2018 976486010674 -MofletPrograrvning, Problem Galng, and Abstraction wih © Revised EditonPage 34 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION if (n <0) if (scanf(sdtdtd", en, om, 6x) ! print£("scanf failed to read ‘exit (EXIT_FAILURE) ; 3 three items\n"); if (yeart4==0 &6 (yearti00!=0 11 yeart40\ 7+ need to allow for leap years */ length_of_year = 3667 length_of_feb = 29; ) else { J= not a leap year «/ Jength_of_year = 365; Jength_of_feb ~ 28; = length_of_feb; I month==6 I 9 || month==11) { Js thirty daye hath september, april, june, and november */ Jengeh_of_month = 30; ) else { J= all the rest have 31, except february... */ Jength_of_month = 31; Figure 3.2: Examples of the use of the if statement. Statements between a “{” and “}” pair are a compound statement, and treated as a single statement. ‘At face value, when the number zero is entered the program should print a message that more students can be accepted. But when the program is executed (the lower box in Figure 3.3), the other message is printed, and itis clear that the {£ statement took the first of the two alternative paths. Why? Well, you were warned about this earlier — look carefully at the guard in the i, and count the number of “=” characters. ‘Two are required for an equality test, and when there is only one, it is an assignment statement. So that guard says “assign the value MAX.CLASS.SIZ# to the variable clacs.size, and then, if the value that was assigned is non-zero, execute the first problem. The gcc compiler used while preparing this book does not naturally give such warnings (sce the first compilation linc in the lower box of Figure 3.3), but can be instructed to look for potential problems via the use of the “a1” option to the “1” compiler wamnings-level flag (the second compilation line), which asks for all ‘warming messages to be listed. The ~ansi flag similarly asks that the gcc compiler Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon3.3. PITFALLS Pace 35 define MAX_CLASS_sIZE 50 int class_size; printf ("Enter class size: "); scant ("sa", &class_size); Af (class_size = VAX_CLASS SIZE) ( print£("Class is now full\n"); } else { printf("Wore students can be accepted\n"); ) gee -0 equalinif equalinif.c : .fequalinit Enter class size: 0 Clase ie now full mac: gcc Mall -ansi -o equalinif equalinif.c equalinif.c: In function ‘main’: equalinif.c:11: warning: suggest parentheses around assignzent used as truth value Figure 3.3: A flawed i statement. The program always prints “Class is now full”, even if 2 class.sizo of 0 is entered. The reason why is revealed in the second compilation line shown in the lower box, when all warning messages are requested using “-Wa11”. Now the compiler draws attention to the assignment statement inside the guard of the { statement, and suggests that it might be problematic, despite that fact that the program is technically correct as it stands. int x=3, v4, 2-6 if (02) SE (y>6) 2am else z= 87 printf(*after the if statement 2=4d\n", 2)7 Figure 3.4: Another flawed i statement. not accept any constructs that are not part of the ANSI standard definition of C. ‘There is no reason at all why you shouldn’t always use these warning and diagnostic facilities if they are available. What about the program fragment in Figure 3.4? What is the final value of 2’? Should be 6, right? The first guard is uue, but the second guard is false, right? So z remains unchanged, right? Wrong! The rule is that an ice part attaches to the most recent unmatched if that is available to it, irrespective of program layout. So the e1se in this fragment belongs to the second i, and when the second guard is false, the statement 2~8 is executed. Like the previous pitfall, the program in Figure 3.4 is correct according to the rules of C, and when compiled using the default gcc options, no warning message is ‘Copyright © Pearson Ausbata(edivsin of Pearson Australia Group Pty Li) 2010 ~ 976406010674 -MullevProgramming, Problem Soiving, nd Abstraction wih C Revised ElionPace 36 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION generated. But when -via11 is used: mac: gee Wall ~ansi -o danglingelse danglingelse.< denglingelse.c: In function ‘main’: danglingelee.ci10: warning: euggest explicit braces to avoid ambiguous ‘else’ This issue is known as the dangling else question, and the C rule is the one usually adopted in programming languages in which it arises. If you always put the braces in, as was recommended a few paragraphs ago, you will never encounter a dangling else in your programs. If your C compiler provides for additional warning and diagnostic facilities, you should gratefully accept the assistance. ‘One last example draws out another mistake that it is easy to make: what is the final value of z in this program fragment? int x3, yah, e6) af (Cy <2) z= atl; if (2 > y> x) 2 = 242; printf(*after the two if statements z=sd\n", 2); The obvious answer is 9, but by now you are probably (and rightly) suspicious of the obvious answer, In fact, the program prints 7. To understand why, insert the implicit precedence-derived parentheses, which make the second guard (2>y) >x. ‘Then remember that (>y) will be either zero or one. When the first version of this book was written, using -Wa11 didn’t help find this type of problem — the compiler was silent even when you asked to be warned of possibly erroneous constructs. But the gcc compiler is under constant development, and it now recognizes this potential problem and issues a warning message that “comparisons like x < y < 2do not have their mathematical meaning”, Take the hint, and always use ~ma11. 3.4 Case study: Calculating taxes You now have enough C knowledge to tackle a non-trivial program. See if you can implement a solution to this calculation: The 2012/13 Australian tax rates stipulate that annual income between $18,200 and $37,000 is taxed at 19.0 cents in the dollar; income be- tween $37,000 and $80,000 at 32.5 cents in the dollar; and income between $80,000 and $180,000 at 37.0 cents in the dollar. Any further income is taxed at 45.0 cents in the dollar. In addition, a Medicare levy of 1.5 cents in the dollar is applied to all income. For a given gross income, what net income is received? Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised Editon3.4 Case srupy Pace 37 7s Calculate Australian tax and Medicare levy, 2012 scales. “/ include include fdefine RATED 0.000 fdefine RATE 0.190 fdefine RATE2 0.325 fdefine RATE3 0.370 fdefine RATEG 0.450 fcefine RATEM 0.015 fcefine THRESH! 1820.00 define THRESH2 3700.00 fdefine THRESE3 80000.00 fcefine THRESHs 180000.00 define BASED 0.00 fdefine BASEL (BASEO + RATEQ*THRESH1) ficefine BASE2 (BASE1 + RATE1s (THRESH2-THRESH1)) fcefine BASE3 (BASE2 + RATE2« (THRESH3-THRESH2)) fdefine BASE4 (BASE3 + RATE3«(THRESHI-THRESH3)) int, main(int arge, char sargv{]) { Gouble gross, tax, medicare, net; print£(*snter gross salary: $*); 4£ (scanf ("$1£", goross) != 1) { print£("No value was entered\n"); exit (EXIT_FATLORE) ; ) 4£ (gross <= THRESHL) ( tax = BASEO + gross+RATEO; } else if (gross <= THRESH2) { tax = BASEL + (gros~THRESH1) «RATE1; ) else sf (gross <= THRESH3) ( tax = BASE2 + (gross-THRESH2) #RATE2; } else if (grose <- THRESH) [ tax = BASE3 + (gross-THRESH3) «RATES; ) else { tax = BASES + (gross-THRESH4) «RATES; ) medicare = RATENsgross; net = gross ~ tax ~ medicare; print£(*grose incone $89.2£\n", gross} printf ("less tax $89.2f\n", -tax); print£(*less medicare 989.2£\n", -medicare); print£("net income $89.2£\n", net); return 0} ‘Figure 3.5: A program to calculate net income es a function of gross income. ‘Copyright © Pearson Ausbata(edivsin of Pearson Australia Group Pty Li) 2010 ~ 976406010674 -MullevProgramming, Problem Soiving, nd Abstraction wih C Revised ElionPace 38 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION Figure 3.5 shows how the task can be tackled using a cascading if statement. Note how the use of symbolic constants avoids much of the complexity inherent to this computation — in a sense, the program is controlled by two tables, one of threshold amounts, and another of corresponding tax rates. Changes to the thresholds or tax rates can thus be made relatively easily, and even if the number of steps in the tax scale changes, it is clear how the program should be altered. On the other hand, the program is also repetitive in its computation, and there is scope for exrors to creep in because of that. Chapter 7 introduces techniques that allow a more elegant implementation of this kind of task. Until then, the program in Figure 3.5 is pretty much the best we can do for this problem. Here is what happens when it is executed: mac: ./taxation Enter gross salary: $98765.43 gross incone § 98765.43 less tax $-24490.21 less medicare $ 1481.48 net. income $ 72793.73 Neat and tidy output is important if the eventual users of a program are to place their confidence in a program, and you should never forget that the only part of your program seen by most users is the interface though which input values are read and output values are reported. If the interface to a program is untidy or awkward to use, users will ‘assume that the program itself is of poor quality. 3.5 The switch statement ‘There is another mechanism provided in the C language to achieve selection ~ the switch statement. It doesn’t really offer any extra functionality compared to the it statement, nor is it any more succinct. Indeed, the one way in which it differs from the if statement is capable of causing more confusion than it is worth, and you are advised to avoid any use of the switch. Nevertheless, you may encounter the switch in programs that you are asked to ‘maintain. The general form of itis switch ( integer expression ) ( case ( integer! ): statement] break; case ( integer? ): statement2 break; case (integer n): statement n break; Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019.3.5. THE SWITCH STATEMENT Pace 39 default: statement ) When executed, the value of the integer expression is calculated. That value is then compared in turn against the various case expressions, until either a match is found, or the default (if it exists) is reached. The corresponding statement or list of statements is then executed, continuing through until cither the closing brace of the switch is encountered, or until a break statement is executed. Note that the general form shown here presumes that there is a break for each case. Unfortunately, it is not mandatory. If there is a break after each group of statements, the behavior of the switch is the same as that of a corresponding cascading if. The following box shows a switch that is well structured in this regard: ‘switch (month) { case 2: length_of_month = 28 + (year’4==0 4& (yeart100!=0 || yeart40 break; case 6: ease 11: length_of_ronth break; default: Length_of_nonth break; Note how in this case the accumulation of cases is plausible ~ a bit like several differ- cent tests connected with “or” operations in an 4 ¢-statement guard. In this fragment there are three mutually exclusive sets of statements controlled by the switch, and three matching break statements. If the programmer does not follow the discipline of having one reak for every ‘group of statements, cases flow together in a chaotic and confusing manner. It is for this reason that the switch statement is to be distrusted. Program execution starts at the first case with a matching value, and moves through all subsequent statements ‘until a break is encountered. Programmers trying to be “clever” then order the cases to exploit this flexibility. To see an example of this “cleverness”, trace the switch statement in Figure 3.6 for each possible starting value of x between 1 and 10, and determine the final value of x that corresponds to each starting value. Confusing, isn't it! Should a programmer who writes things like this be rewarded? The switch statement offers a general form of the cascading i¢. Use it with caution, and maintain the discipline of having a brea for each non-empty group of statements, Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonPace 40 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION witch (x) { case 1 case 4 case 7: break; case 3) case 9: xi 3; default: Figure 3.6: A dangerous switch statement, Exercises 3.1 Number each of the operators in this expression with the order in which it is applied. Then work out the value of the expression. 7 7 10> a<5 a6 6 & Tee-2s 3.2 Trace the action of these statements, and determine the values printed out by each of the print £ statements. Assume that all variables have been declared to be of type int. Be careful — some of them illustrate common errors. a T £ (i fdefine NAX_YEARS 7 /+ number of years to compute for */ fdefine MONTHLY 100 /+ monthly savings amount «/ fcefine MIN.RATE 2 /+ minimum annual percentage rate +/ fdefine NAXRATE 7 /+ maximum interest rate */ int main(int arge, char vargv(}) ( int month, rate, year; double balance, nonthly_rate; /s print the table heading lines +/ print£(*Monthly savings of $%d", MONTHLY); print£(", with monthly compounded interest\n"); printf(*Annual Rate 1"); for (rate-MIN RATE; rate<-MAX RATS; ratet+) [ printe("e4d¥s ", rate); ) print£(*"\n")5 /+ and now the rows of the table, one by one +/ for (year=1; year<-MAK_YEARS; yeart+) ( print£("After tld years | ", year); for (ratenMIN_RATE; rate<-MAX RATE; rate++) { /* compute one value */ balance = 0. monthly_rate = 1.00 + ((rate/100.09)/12); for (month=1; monthe=12syear; month++) ( balance «= monthly_rate; balance += MONTHLY; ) print£("$5.0£ *, balance); } J* end of one row +/ print£(*\n"); ) return 0; Figure 4.6: Using nested loops to generate a table. An example of the output generated by this program appears in Figure 4.5. braces should be prominently displayed. Yet another style tucks the closing brace of each compound statement up at the end of the last statement. What is important is that you adopt a single style, and then stick with it. Consistency is more impor- tant that the particular style used. Indeed, to obtain consistency, organizations often ‘Copyright © Pearson Ausbata(edivsin of Pearson Australia Group Pty Li) 2010 ~ 976406010674 -MullevProgramming, Problem Soiving, nd Abstraction wih C Revised ElionPage 52 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION printi("Enter date in dd/mm/yyyy format: ");if (scant ( "ad/ad/%d", edd, om, eyyyy) !=3) (printf ("Invalid input \n") jexit (EXIT_FAILURE) ; }length_of_feb~28+ (yyyyt4==066 (yyy §100!-0] | yyyy#400--0) ) ;daynum-dd; for (nonth~1 jmonth1;) is identical, and would have served equally well, but the use of the word white in some sense captures the open-ended nature of the computation ~ we are not actually in a position to demonstrate that this loop will terminate, since there is no guarantec that every path through the loop results in the guard moving closer to becoming false. Note also that in this program a single nested loop has been used, despite the fact that the output of the program again appears as a table. In this case the tabular layout is purely a convenience, and the underlying computation is essentially one- dimensional. So a single loop is used, and the two-dimensional output is obtained through the use of an if statement within the loop that periodically inserts newline characters into the output character stream. <_ gues > Non-orol statement Figure 4.8: Execution flow through a whi Le statement. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonCopyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. Page 54 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION 7+ @ value for n has been read +/ cycles = 0; while (n>1) ( printf (*"’5d ", n); ) else { n= Sent; ) Sf (npmax) { max = nj 1 eyeles 4 1; Sf (cyclestPERLINE print£("\n"); ) ) print£(*\n$d cycles consumed, maximum was d\n", cycles, max); mac: ./threen Enter starting value for n: 8 ay” wh (2 3 cycles consumed, maximum was 8 mac: ./threen Enter starting value for ni 9 9 2 14 7 22 WW 34 17 52 26 «413 «400 «20 «10tiSC a 19 cycles consumed, maximum was 52 Figure 4.9: A program to explore some properties of the 37 + 1 problem. Both for statements and wh:.1e statements share a common attribute — they test the guard prior to each and every iteration, and continue with that execution only if, the guard is true, This means that they cannot terminate until the guard is false. More to the point, it means that when such loops do terminate, the guard is always false. When a for or while loop terminates, the guard is false, This observation means that one way of creating a loop guard is to decide what is supposed to be true at the end of the loop, and then make the guard the negation of, that. For example, if n==1 is required at the end of a loop, the guard should be n !=1 Every iteration of the loop should then move a step closer towards changing the guard from true to false. If you think about your loops as needing to make progress towards falsifying the guard, you will be less likely to get bogged down debugging infinite loops. If a loop is to terminate, every iteration must move one step closer to ‘making the loop guard false. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonCopyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 4.4 UNCONTROLLED ITERATION Pace 55 7+ @ value for n has been read +/ isprine = 1; for (divisor=2; divisorsdivisormax) { print£("td values read, maximum is $d\n", n, max); mac: ./readloopl Enter values, control-D to end: 12326527154 ‘D 9 values read, maximum is 27 Figure 4.12: Use of scan¢ in a loop to process a set of input items of arbitrary size. The Jower box shows an execution of the program fragment in the upper box. Copyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonCopyright © Pearson Ausbata (edison of Pearson Australia Group Pty Lt) 2019. Pace 58 PROGRAMMING, PROBLEM SOLVING, AND ABSTRACTION int ch, incomment=0, newlinelast=1; while ((ch=getchar()) != EOF) { if (ch == 'C’ && newlinelast) { incomment ~ 17 3 if (!incomment) { putchar (ch); ) if (ch == "\n") { incomment = Rewlinelast ) else { newlinelest = 0; ) mac: more data.txt Ca comment in the first line ut no Comtent here or here © but there ie one here mac: ./fortconm < data.txt but no Content here or here Figure 4.13: Use of get char and putchar to process the input stream character by char- ‘acter. The program fragment removes lines that start with an uppercase “C”, The lower box shows an execution of the program fragment in the upper box. and output — the functions get char and putchar. To show the use of these two functions, suppose that we are required to write a program that deletes all of the lines in a file that start with an uppercase “C”. (This is the convention for indicating comments in Fortran programs.) Figure 4.13 shows a program fragment that achieves: this goal. The first character of each line is the character after the previous newline character; and a newline always ends a comment. ‘Note the loop guard, and the way in which the guard both assigns to variable ch the integer ASCII value of the next input character, and then also tests to see if the assignment resulted in a valid value. This is one situation in which assignment side effects can be both tolerated and exploited, to make a recipe for reading that all C programmers know and use. ‘The predefined constant EOF takes on a value outside the ASCII set, and in most, implementations has the value —1. The need to represent this non-character is why cch must be declared to be an int rather than a char. Functions getcha® and put char allow processing of the input as a stream of characters. The variable used must be declared to be an int. in order for end of file value EOF to be manipulated properly. And now for a challenge. Earlier in this book you saw a C program that read a set of numbers and printed out their sum, Without looking back at page 8, can you 76486010074 - MollProganning, Problems Golvng, and Abstraction wih © Revised EditonEXERCISES Pace 59 write an equivalent program using the C programming techniques described so far’? If you can, you are doing well; if you can’t, you should review Chapters 1, 2, and 3, before proceeding to the next chapter. Now might also be a good time to implement solution to Exercise 1.3 on page 12. Exercises 4.1 Trace the action of these loops, and determine the values printed out by each of the print £é statements. Assume that all variables have been declared to be of type int. Be careful with the last two — they illustrate two common errors. a for (1=0; 1<20; 1-143) { print£("82d\n", i); } b. for (i=1; i1<2000000; i=24i) { print£("87d\n", 1); ) G (ist; isto; itt) ¢ peint£("S(#2d) = #2d\n", i, sum); ) a. 7 i<8; i+) 1 for (Sith; 3<8) 5 printf ("i=td, } ) e for (1-0; i<8; i++) { for (Jeitt; 3<8; 3483) ( if (its=57) ( break; > print£("ietd, jmtd\n", i, 4); 1 » f. 7-5 for (1-0; ict; At { printf ("i-td, j-td\n", 3, 3); d B a= 5 For (in0; i

You might also like

• The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  

  From Everand

  The Subtle Art of Not Giving a F*ck: A Counterintuitive Approach to Living a Good Life

  Mark Manson

  4/5 (6125)
• Principles: Life and Work

  

  From Everand

  Principles: Life and Work

  Ray Dalio

  4/5 (627)
• The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  

  From Everand

  The Gifts of Imperfection: Let Go of Who You Think You're Supposed to Be and Embrace Who You Are

  Brené Brown

  4/5 (1148)
• Never Split the Difference: Negotiating As If Your Life Depended On It

  

  From Everand

  Never Split the Difference: Negotiating As If Your Life Depended On It

  Chris Voss

  4.5/5 (932)
• The Glass Castle: A Memoir

  

  From Everand

  The Glass Castle: A Memoir

  Jeannette Walls

  4/5 (8214)
• Grit: The Power of Passion and Perseverance

  

  From Everand

  Grit: The Power of Passion and Perseverance

  Angela Duckworth

  4/5 (631)
• Sing, Unburied, Sing: A Novel

  

  From Everand

  Sing, Unburied, Sing: A Novel

  Jesmyn Ward

  4/5 (1253)
• The Perks of Being a Wallflower

  

  From Everand

  The Perks of Being a Wallflower

  Stephen Chbosky

  4/5 (8365)
• Shoe Dog: A Memoir by the Creator of Nike

  

  From Everand

  Shoe Dog: A Memoir by the Creator of Nike

  Phil Knight

  4.5/5 (860)
• Her Body and Other Parties: Stories

  

  From Everand

  Her Body and Other Parties: Stories

  Carmen Maria Machado

  4/5 (877)
• Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  

  From Everand

  Hidden Figures: The American Dream and the Untold Story of the Black Women Mathematicians Who Helped Win the Space Race

  Margot Lee Shetterly

  4/5 (954)
• The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  

  From Everand

  The Hard Thing About Hard Things: Building a Business When There Are No Easy Answers

  Ben Horowitz

  4.5/5 (361)
• Steve Jobs

  

  From Everand

  Steve Jobs

  Walter Isaacson

  4/5 (2922)
• Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  

  From Everand

  Elon Musk: Tesla, SpaceX, and the Quest for a Fantastic Future

  Ashlee Vance

  4.5/5 (484)
• The Emperor of All Maladies: A Biography of Cancer

  

  From Everand

  The Emperor of All Maladies: A Biography of Cancer

  Siddhartha Mukherjee

  4.5/5 (277)
• Brooklyn: A Novel

  

  From Everand

  Brooklyn: A Novel

  Colm Toibin

  3.5/5 (2061)
• A Man Called Ove: A Novel

  

  From Everand

  A Man Called Ove: A Novel

  Fredrik Backman

  4.5/5 (4972)
• Angela's Ashes: A Memoir

  

  From Everand

  Angela's Ashes: A Memoir

  Frank McCourt

  4.5/5 (444)
• The Art of Racing in the Rain: A Novel

  

  From Everand

  The Art of Racing in the Rain: A Novel

  Garth Stein

  4/5 (4281)
• The Yellow House: A Memoir (2019 National Book Award Winner)

  

  From Everand

  The Yellow House: A Memoir (2019 National Book Award Winner)

  Sarah M. Broom

  4/5 (100)
• The Little Book of Hygge: Danish Secrets to Happy Living

  

  From Everand

  The Little Book of Hygge: Danish Secrets to Happy Living

  Meik Wiking

  3.5/5 (447)
• The World Is Flat 3.0: A Brief History of the Twenty-first Century

  

  From Everand

  The World Is Flat 3.0: A Brief History of the Twenty-first Century

  Thomas L. Friedman

  3.5/5 (2283)
• Bad Feminist: Essays

  

  From Everand

  Bad Feminist: Essays

  Roxane Gay

  4/5 (1068)
• Yes Please

  

  From Everand

  Yes Please

  Amy Poehler

  4/5 (1987)
• Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  

  From Everand

  Devil in the Grove: Thurgood Marshall, the Groveland Boys, and the Dawn of a New America

  Gilbert King

  4.5/5 (278)
• The Outsider: A Novel

  

  From Everand

  The Outsider: A Novel

  Stephen King

  4/5 (1993)
• The Woman in Cabin 10

  

  From Everand

  The Woman in Cabin 10

  Ruth Ware

  3.5/5 (2619)
• A Tree Grows in Brooklyn

  

  From Everand

  A Tree Grows in Brooklyn

  Betty Smith

  4.5/5 (1936)
• The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  

  From Everand

  The Sympathizer: A Novel (Pulitzer Prize for Fiction)

  Viet Thanh Nguyen

  4.5/5 (125)
• A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  

  From Everand

  A Heartbreaking Work Of Staggering Genius: A Memoir Based on a True Story

  Dave Eggers

  3.5/5 (692)
• Team of Rivals: The Political Genius of Abraham Lincoln

  

  From Everand

  Team of Rivals: The Political Genius of Abraham Lincoln

  Doris Kearns Goodwin

  4.5/5 (1912)
• Wolf Hall: A Novel

  

  From Everand

  Wolf Hall: A Novel

  Hilary Mantel

  4/5 (4074)
• On Fire: The (Burning) Case for a Green New Deal

  

  From Everand

  On Fire: The (Burning) Case for a Green New Deal

  Naomi Klein

  4/5 (75)
• Fear: Trump in the White House

  

  From Everand

  Fear: Trump in the White House

  Bob Woodward

  3.5/5 (830)
• Rise of ISIS: A Threat We Can't Ignore

  

  From Everand

  Rise of ISIS: A Threat We Can't Ignore

  Jay Sekulow

  3.5/5 (143)
• Manhattan Beach: A Novel

  

  From Everand

  Manhattan Beach: A Novel

  Jennifer Egan

  3.5/5 (901)
• John Adams

  

  From Everand

  John Adams

  David McCullough

  4.5/5 (2530)
• The Light Between Oceans: A Novel

  

  From Everand

  The Light Between Oceans: A Novel

  M L Stedman

  4.5/5 (790)
• The Unwinding: An Inner History of the New America

  

  From Everand

  The Unwinding: An Inner History of the New America

  George Packer

  4/5 (45)
• Little Women

  

  From Everand

  Little Women

  Louisa May Alcott

  4/5 (105)
• The Constant Gardener: A Novel

  

  From Everand

  The Constant Gardener: A Novel

  John le Carré

  3.5/5 (109)