\input texinfo @c -*-texinfo-*- @c ********************************************************************* @c @c This is a TEXINFO file. It generates both TEX documentation and @c the "on line" documentation "info" files. @c @c The file is structured like a programming language. Each chapter @c starts with a chapter comment. @c @c Menus list the subsections so that an online inforeader can parse @c the file hierachically. @c @c Each time you add or change a node, you need to update the @c menus and forward/backward references. You can do this from @c emacs automatically by writing: @c @c ESC-x texinfo-every-node-update @c ESC-x texinfo-all-menus-update @c ESC-x texinfo-master-menu @c @c In practice you can just type the first few letters and hit TAB @c to write these long names. @c @c Last change: MB 31/08/95 @c @c *********************************************************************** @c %** start of header @setfilename unix.info @settitle The unix programming environment @setchapternewpage odd @c %** end of header @include version.texi @titlepage @title The unix programming environment @subtitle Edition @value{EDITION}, @value{UPDATED} @author Mark Burgess @author Centre of Science and Technology @author Faculty of Engineering, Oslo College @c @smallbook @page @vskip 0pt plus 1filll @ifinfo Copyright @copyright{} 1996/7 Mark Burgess Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided also that the section entitled "GNU General Public License" is included exactly as in the original, and provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that the section entitled "GNU General Public License" may be included in a translation approved by the author instead of in the original English. @end ifinfo Copyright @copyright{} 1996/7 Mark Burgess Permission is granted to make and distribute verbatim copies of this manual provided the copyright notice and this permission notice are preserved on all copies. Permission is granted to copy and distribute modified versions of this manual under the conditions for verbatim copying, provided also that the section entitled "GNU General Public License" is included exactly as in the original, and provided that the entire resulting derived work is distributed under the terms of a permission notice identical to this one. Permission is granted to copy and distribute translations of this manual into another language, under the above conditions for modified versions, except that the section entitled "GNU General Public License" may be included in a translation approved by the author instead of in the original English. @end titlepage @ifinfo @node Top, Comment, (dir), (dir) @top UNIX @end ifinfo @c ********************************************************************** @c INTRO @c ********************************************************************** @unnumbered Foreword This is a revised version of the UNIX compendium which is available in printed form and online via the WWW and info hypertext readers. It forms the basis for a one or two semester course in UNIX. The most up-to-date version of this manual can be found at @smallexample http://www.iu.hioslo.no/~mark/unix/unix.html. @end smallexample It is a reference guide which contains enough to help you to find what you need from other sources. It is not (and probably can never be) a complete and self-contained work. Certain topics are covered in more detail than others. Some topics are included for future reference and are not intended to be part of an introductory course, but will probably be useful later. The chapter on X11 programming has been deleted for the time being. Comments to Mark.Burgess@@iu.hioslo.no Oslo, June 1996 @c ********************************************************************** @c CHAPTER @c ********************************************************************** @menu * Comment:: * Overview:: * Getting started:: * The login environment:: * Files and access:: * C shell:: * Bourne shell:: * Perl:: * WWW and CGI programming:: * C programming:: * Network programming:: * Programming idioms:: * Variable Index:: * Concept Index:: @end menu @node Comment, Overview, Top, Top @unnumbered Welcome If you are coming to unix for the first time, from a Windows or MacIntosh environment, be prepared for a rather different culture than the one you are used to. Unix is not about `products' and off-the-shelf software, it is about open standards, free software and the ability to change just about everything. @itemize @bullet @item What you personally might perceive as user friendliness in other systems, others might perceive as annoying time wasting. Unix offers you just about every level of friendliness and unfriendliness, if you choose your programs right. In this book, we take the programmer's point of view. @item Unix is about functionality, not about simplicity. Be prepared for powerful, not necessarily `simple' solutions. @end itemize @noindent You should approach Unix the way you should approach any new system: with an open mind. The journey begins... @node Overview, Getting started, Comment, Top @chapter Overview @emph{In this manual the word "host" is used to refer to a single computer system -- i.e. a single machine which has a name termed its "hostname".} @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * What is unix?:: * Flavours of unix:: * How to use this reference guide:: * NEVER-DOs in UNIX:: * What you should know before starting:: * The superuser and nobody:: * The file hierarchy :: * Symbolic links:: * Hard links:: @end menu @node What is unix?, Flavours of unix, Overview, Overview @section What is unix? @cindex UNIX @cindex UNIX history Unix is one of the most important operating system in use today, perhaps even @emph{the} most important. Since its invention around the beginning of the 1970s it has been an object of continual research and development. UNIX is not popular because it is the best operating system one could imagine, but because it is an extremely flexible system which is easy to extend and modify. It is an ideal platform for developing new ideas. Much of the success of UNIX may be attributed to the rapid pace of its development (a development to which all of its users have been able to contribute) its efficiency at running programs and the many powerful tools which have been written for it over the years, such as the C programming language, @code{make}, shell, @code{lex} and @code{yacc} and many others. UNIX was written by programmers for programmers. It is popular in situations where a lot of computing power is required and for database applications, where timesharing is critical. In contrast to some operating systems, UNIX performs equally well on large scale computers (with many processors) and small computers which fit in your suitcase! All of the basic mechanisms required of a multi-user operating system are present in UNIX. During the last few years it has become ever more popular and has formed the basis of newer, though less mature, systems like NT. One reason for this that now computers have now become powerful enough to run UNIX effectively. UNIX places burdens on the resources of a computer, since it expects to be able to run potentially many programs simultaneously. @cindex DOS If you are coming to UNIX from DOS you may well be used to using applications software or helpful interactive utilities to solve every problem. UNIX is not usually like this: the operating system has much greater functionality and provides the possibilities for making your own, so it is less common to find applications software which implements the same things. UNIX has long been in the hands of academics who are used to making their own applications or writing their own programs, whereas as the DOS world has been driven by businesses who are willing to spend money on software. For that reason commerical UNIX software is often very expensive and therefore not available at this college. On the other hand, the flexibility of UNIX means that it is easy to write programs and it is possible to fetch gigabytes of @emph{free software} from the internet to suit your needs. It may not look like what you are used to on your PC, but then you have to remember that UNIX users are a different kind of animal altogether Like all operating systems, UNIX has many faults. The biggest problem for any operating system is that it evolves without being redesigned. Operating systems evolve as more and more patches and hacks are applied to solve day-to-day problems. The result is either a mess which works somehow (like UNIX) or a blank refusal to change (like DOS or MacIntosh). From a practical perspective, Unix is important and successful because it is a multi-process system which @itemize @bullet @item has an enormous functionality built in, and the capacity to adapt itself to changing technologies, @item is relatively portable, @item is good at sharing resources (but not so good at security), @item has tools which are each developed to do @emph{one thing well}, @item allows these tools to be combined in every imaginable way, using pipes and channeling of data streams, @item incorporates networking almost trivially, because all the right mechanisms are already there for providing services and sharing, building client-server pairs etc,. @item it is very adaptable and is often used to develop new ideas because of the rich variety of tools it possesses. @end itemize Unix has some problems: it is old, it contains a lot of rubbish which no one ever bothered to throw away. Although it develops quickly (at light speed compared to either DOS or MacIntosh) the user interface has been the slowest thing to change. Unix is not user-friendly for beginners, it is user-friendly for advanced users: it is made for users who @emph{know} about computing. It sometimes makes simple things difficult, but above all it makes things possible! The aim of this introduction is to @itemize @bullet @item introduce the unix system basics and user interface, @item develop the unix philosophy of using and combining tools, @item learn how to make new tools and write software, @item learn how to understand existing software. @end itemize @noindent To accomplish this task, we must first learn something about the shell (the way in which UNIX starts programs). Later we shall learn how to solve more complex problems using Perl and C. Each of these is a language which can be used to put UNIX to work. We must also learn when to use which tool, so that we do not waste time and effort. Typical uses for these different interfaces are @table @emph @item shell Command line interaction, making scripts which perform simple jobs such as running programs in batch, installing new software, simple system configuration and administration. @cindex shell @item perl Text interpretation, text formatting, output filters, mail robots, WWW cgi (common gateway interface) scripts in forms, password testing, simple database manipulation, simple client-server applications. @cindex perl @item C Nearly all of UNIX is written in C. Any program which cannot be solved quickly using shell or perl can be written in C. One advantage is that C is a compiled language and many simple errors can be caught at compile time. @cindex C @end table Much of UNIX's recent popularity has been a result of its networking abilities: unix is the backbone of the internet. No other widely available system could keep the internet alive today. Once you have mastered the unix interface and philosophy you will find that i) the PC and MacIntosh window environments seem to be easy to use, but simplistic and primitive by comparison; ii) UNIX is far from being the perfect operating system---it has a whole different set of problems and flaws. @cindex Windows on PC @cindex MacIntosh @cindex PC windows The operating system of the future will not be UNIX as we see it today, nor will is be DOS or MacIntosh, but one thing is for certain: it will owe a lot to the UNIX operating system and will contain many of the tools and mechanisms we shall describe below. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Flavours of unix, How to use this reference guide, What is unix?, Overview @section Flavours of unix @cindex AT&T @cindex BSD @cindex System V @cindex System 5 Unix is not a single operating system. It has branched out in many different directions since it was introduced by AT&T. The most important @samp{fork()} in its history happened early on when the university of Berkeley, California created the BSD (Berkeley Software Distribution), adding network support and the C-shell. Here are some of the most common implementations of unix. @table @emph @item BSD: Berkeley, BSD @item SunOS: Sun Microsystems, BSD/sys 5 @item Solaris: Sun Microsystems, Sys 5/BSD @item Ultrix: Digital Equipment Corperation, BSD @item OSF 1: Digital Equipment Corperation, BSD/sys 5 @item HPUX: Hewlett-Packard, Sys 5 @item AIX: IBM, Sys 5 / BSD @item IRIX: Silicon Graphics, Sys 5 @item GNU/Linux: GNU, BSD/Posix @end table @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node How to use this reference guide, NEVER-DOs in UNIX, Flavours of unix, Overview @section How to use this reference guide This programming guide is something between a user manual and a tutorial. The information contained here should be sufficient to get you started with the unix system, but it is far from complete. To use this programming guide, you will need to work through the basics from each chapter. You will find that there is much more information here than you need straight away, so try not to be overwhelmed by the amount of material. Use the contents and the indices at the back to find the information you need. If you are following a one-semester UNIX course, you should probably concentrate on the following: @itemize @bullet @item The remainder of this introduction @item The detailed knowledge of the C shell @item An appreciation of the Bourne shell @item A detailed knowledge of Perl, guided by chapter 6. This chapter provides pointers on how to get started in perl. It is not a substitute for the perl book. @item Everything in chapter 7 about C programming. This chapter is written in note form, since it is assumed that you know a lot about C programming already. @item A sound appreciation of chapter 8 on network programming. @end itemize The @emph{only} way to learn UNIX is to sit down and try it. As with any new thing, it is a pain to get started, but once you are started, you will probably come to agree that UNIX contains a wealth of possibilities, perhaps more than you had ever though was possible or useful! One of the advantages of the UNIX system is that the entire UNIX manual is available on-line. You should get used to looking for information in the online manual pages. For instance, suppose you do not remember how to create a new directory, you could do the following: @smallexample nexus% man -k dir dir ls (1) - list contents of directories dirname dirname (1) - strip non-directory suffix from file name dirs bash (1) - bash built-in commands, see bash(1) find find (1) - search for files in a directory hierarchy ls ls (1) - list contents of directories mkdir mkdir (1) - make directories pwd pwd (1) - print name of current/working directory rmdir rmdir (1) - remove empty directories @end smallexample @cindex Making directories @vindex mkdir @cindex mkdir @cindex apropos @vindex apropos @vindex man -k @cindex Finding commands @noindent The @samp{man -k} command looks for a keyword in the manual and lists all the references it finds. The command @samp{apropos} is completely equivalent to @samp{man -k}. Having discovered that the command to create a directory is @samp{mkdir} you can now look up the specific manaul page on @samp{mkdir} to find out how to use it: @smallexample man mkdir @end smallexample @noindent Some but no all of the UNIX commands also have a help option which is activated with the @samp{-h} or @samp{--help} command-line option. @cindex Help function for commands @vindex -h @vindex --help @smallexample dax% mkdir --help Usage: mkdir [OPTION] DIRECTORY... -p, --parents no error if existing, make parent directories as needed -m, --mode=MODE set permission mode (as in chmod), not 0777 - umask --help display this help and exit --version output version information and exit dax% @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node NEVER-DOs in UNIX, What you should know before starting, How to use this reference guide, Overview @section NEVER-DO's in UNIX @cindex Never do in unix There are some things that you should never do in UNIX. Some of these will cause you more serious problems than others. You can make your own list as you discover more. @itemize @bullet @item You should NEVER EVER switch off the power on a Unix computer unless you know what you are doing. A Unix machine is not like a PC running DOS. Even when you are not doing anything, the system is working in the background. If you switch off the power, you could interrupt the system while it is writing to the disk drive and destroy your disk. You must also remember that several users might be using the system even though you cannot see them: they do not have to be sitting at the machine, they could be logged in over the network. If you switch off the power, you might ruin their valuable work. @item Once you have deleted a UNIX file using @code{rm} it is impossible to recover it! Don't use wildcards with @code{rm} without thinking quite carefully about what you are doing! It has happened to very many users throughout the history of UNIX that one tries to type @smallexample rm *~ @end smallexample but instead, by a slip of the hand, one writes @smallexample rm * ~ @end smallexample Unix then takes these wildcards in turn, so that the first command is @code{rm *} which deletes all of your files! BE CAREFUL! @cindex undelete @cindex delete @item Don't ever call a program or an important file @file{core}. Many scripts go around deleting files called @file{core} because the, when a program crashes, UNIX dumps the entire kernel image to a file called @file{core} and these files use up a lot of disk space. If you call a file @file{core} it might get deleted! @cindex core @item Don't call test programs @code{test}. There is a UNIX command which is already called test and chances are that when you try to run your program you will start the UNIX command instead. This can cause a lot of confusion because the UNIX command doesn't seem to do very much at all! @end itemize @cindex test programs @cindex test, don't call your program this @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node What you should know before starting, The superuser and nobody, NEVER-DOs in UNIX, Overview @section What you should know before starting @c ..................................................... @c SUBSECTION @c ..................................................... @menu * One library:: * Unix commands are files:: * Kernel and shell:: * The role of C:: * Stdin:: @end menu @node One library, Unix commands are files, What you should know before starting, What you should know before starting @subsection One library: several interfaces The core of unix is the library of functions (written in C) which access the system. Everything you do on a unix system goes through this set of functions. However, you can choose your own interface to these library functions. Unix has very many different interfaces to its libraries in the form of languages and command interpreters. You can use the functions directly in C, or you can use command programs like @samp{ls}, @samp{cd} etc. These functions just provide a simple user interface to the C calls. You can also use a variety of `script' languages: C-shell, Bourne shell, Perl, Tcl, scheme. You choose the interface which solves your problem most easily. @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Unix commands are files:: * Kernel and shell:: * The role of C:: * Stdin:: @end menu @node Unix commands are files, Kernel and shell, One library, What you should know before starting @subsection Unix commands are files @cindex Commands as files @cindex Built in commands @cindex shell With the exception of a few simple commands which are built into the command interpreter (shell), all unix commands and programs consist of executable files. In other words, there is a separate executable file for each command. This makes it extremely simple to add new commands to the system. One simply makes a program with the desired name and places it in the appropriate directory. @cindex Commands path @cindex @code{PATH} @cindex @code{path} Unix commands live in special directories (usually called @code{bin} for binary files). The location of these directories is recorded in a variable called @code{path} or @code{PATH} which is used by the system to search for binaries. We shall return to this in more detail in later chapters. @c ..................................................... @c SUBSECTION @c ..................................................... @node Kernel and shell, The role of C, Unix commands are files, What you should know before starting @subsection Kernel and Shell @cindex kernel @cindex shell Since users cannot command the kernel directly, UNIX has a command language known as the @emph{shell}. The word shell implies a layer around the kernel. A shell is a user interface, or command interpreter. There are two main versions of the shell, plus a number of enhancements. @table @emph @item /bin/sh The Bourne Shell. The shell is most often used for writing system scripts. It is part of the original unix system. @cindex Bourne shell @vindex @file{/bin/sh} @cindex @file{sh} @item /bin/csh The C-shell. This was added to unix by the Berkeley workers. The commands and syntax resemble C code. C-shell is better suited for interactive work than the Bourne shell. @cindex C shell @cindex @file{csh} @vindex @file{/bin/csh} @end table @cindex tcsh @cindex bash @cindex ksh The program @code{tcsh} is a public-domain enhancement of the csh and is in common use. Two improved versions of the Bourne shell also exist: @code{ksh}, the Korn shell and @code{bash}, the Bourne-again shell. Although the shells are mainly tools for typing in commands (which are excutable files to be loaded and run), they contain features such as aliases, a command history, wildcard-expansions and job control functions which provide a comfortable user environment. @c ..................................................... @c SUBSECTION @c ..................................................... @node The role of C, Stdin, Kernel and shell, What you should know before starting @subsection The role of C @cindex Role of C in unix @cindex C, role in unix Most of the unix kernel and daemons are written in the C programming language @footnote{In particular they are written in Kernighan and Ritchie C, not the newer ANSI standard C.}. Calls to the kernel and to services are made through functions in the standard C library. The commands like @code{chmod}, @code{mkdir} and @code{cd} are all C functions. The binary files of the same name @code{/bin/chmod}, @code{/bin/mkdir} etc. are just trivial "wrapper" programs for these C functions. @cindex Wrapper functions @cindex C library calls and shell commands @cindex Shell commands and C library calls Until @emph{Solaris 2}, the C compiler was a standard part of the UNIX operating system, thus C is the most natural language to program in in a UNIX environment. Some tools are provided for C programmers: @table @emph @item dbx A symbolic debugger. Also @emph{gdb}, @emph{xxgdb} @emph{ddd}. @cindex Debugger for C @item make A development tool for compiling large programs. @cindex make @item lex A `lexer'. A program which generates C code to recognize words of text. @cindex lex @item yacc A `parser'. This is a tool which generates C code for checking the syntax of groups of textual words. @cindex yacc @item rpcgen A protocol compiler which generates C code from a higher level langauge, for programming RPC applications. @cindex rpcgen @end table @c ..................................................... @c SUBSECTION @c ..................................................... @node Stdin, , The role of C, What you should know before starting @subsection Stdin, stdout, stderr Unix has three logical @emph{streams} or @emph{files} which are always open and are available to any program. @cindex I/O streams @cindex Standard input @cindex Standard output @cindex Standard error @vindex stdin @vindex stdout @vindex stderr @table @emph @item stdin The standard input - file descriptor 0. @item stdout The standard output - file descriptor 1. @item stderr The standard error - file descriptor 2. @end table The names are a part of the C language and are defined as pointers of type @code{FILE}. @smallexample #include /* FILE *stdin, *stdout, *stderr; */ fprintf(stderr,"This is an error message!\n"); @end smallexample @noindent The names are `logical' in the sense that they do not refer to a particular device, or a particular place for information to come from or go. Their role is analogous to the `.' and `..' directories in the filesystem. Programs can write to these files without worrying about where the information comes from or goes to. The user can personally define these places by @emph{redirecting standard I/O}. This is discussed in the next chapter. @vindex . @vindex .. @cindex . directory @cindex .. directory @cindex Error messages A separate stream is kept for error messages so that error output does not get mixed up with a program's intended output. @c ------------------------------------------------------------------------------- @c SECTION @c ------------------------------------------------------------------------------- @node The superuser and nobody, The file hierarchy , What you should know before starting, Overview @section The superuser (root) and @emph{nobody} When logged onto a UNIX system directly, the user whose name is @code{root} has unlimited access to the files on the system. @code{root} can also become any other user without having to give a password. @code{root} is reserved for the system administrator or @emph{trusted users}. @cindex root user @cindex superuser Certain commands are forbidden to normal users. For example, a regular user should not be able to halt the system, or change the ownership of files (see next paragraph). These things are reserved for the @code{root} or @emph{superuser}. In a networked environment, @code{root} has no automatic authority on remote machines. This is to prevent the system administrator of one machine in Canada from being able to edit files on another in China. He or she must log in directly and supply a password in order to gain access privileges. On a network where files are often accessible in principle to anyone, the username @code{root} gets mapped to the user @code{nobody}, who has no rights at all. @cindex nobody @c ------------------------------------------------------------------------------- @c SECTION @c ------------------------------------------------------------------------------- @node The file hierarchy , Symbolic links, The superuser and nobody, Overview @section The file hierarchy @cindex File hierarchy @cindex Hierarchy, file Unix has a hierachical filesystem, which makes use of directories and sub-directories to form a tree. The root of the tree is called the root filesystem or `/'. Although the details of where every file is located differ for different versions of unix, some basic features are the same. The main sub-directories of the root directory together with the most important file are shown in the figure. Their contents are as follows. @table @file @item /bin Executable (binary) programs. On most systems this is a separate directory to /usr/bin. In SunOS, this is a pointer (link) to /usr/bin. @vindex @file{/bin} @item /etc Miscellaneous programs and configuration files. This directory has become very messy over the history of UNIX and has become a dumping ground for almost anything. Recent versions of unix have begun to tidy up this directory by creating subdirectories @file{/etc/mail}, @file{/etc/services} etc! @vindex @file{/etc} @item /usr This contains the main meat of UNIX. This is where application software lives, together with all of the basic libraries used by the OS. @vindex @file{/usr} @item /usr/bin More executables from the OS. @vindex @file{/usr/bin} @item /usr/local This is where users' custom software is normally added. @vindex @file{/usr/local} @item /sbin A special area for statically linked system binaries. They are placed here to distinguish commands used solely by the system administrator from user commands and so that they lie on the system root partition where they are guaranteed to be accessible during booting. @vindex @file{/sbin} @item /sys This holds the configuration data which go to build the system kernel. (See below.) @vindex @file{/sys} @item /export Network servers only use this. This contains the disk space set aside for client machines which do not have their own disks. It is like a `virtual disk' for diskless clients. @vindex @file{/export} @item /dev, /devices A place where all the `logical devices' are collected. These are called `device nodes' in unix and are created by @code{mknod}. Logical devices are UNIX's official entry points for writing to devices. For instance, @code{/dev/console} is a route to the system console, while @code{/dev/kmem} is a route for reading kernel memory. Device nodes enable devices to be treated as though they were files. @vindex @file{/dev} @vindex @file{/devices} @item /home (Called @emph{/users} on some systems.) Each user has a separate login directory where files can be kept. These are normally stored under @code{/home} by some convention decided by the system administrator. @vindex @file{/home} @vindex @file{/users} @item /var System 5 and mixed systems have a separate directory for spooling. Under old BSD systems, @code{/usr/spool} contains spool queues and system data. @code{/var/spool} and @code{/var/adm} etc are used for holding queues and system log files. @vindex @file{/var} @vindex @file{/vr/spool} @vindex @file{/var/adm} @item /vmunix This is the program code for the unix @emph{kernel} (see below). On HPUX systems with file is called @file{hp-ux}. On linux it is called @file{linux}. @vindex @file{vmunix} @cindex Kernel @item /kernel On newer systems the kernel is built up from a number of modules which are placed in this directory. @end table @noindent Every unix directory contains two `virtual' directories marked by a single dot and two dots. @smallexample ls -a . .. @end smallexample @cindex . directory @cindex .. directory @noindent The single dot represents the directory one is already in (the current directory). The double dots mean the directory one level up the tree from the current location. Thus, if one writes @smallexample cd /usr/local cd .. @end smallexample @noindent the final directory is @code{/usr}. The single dot is very useful in C programming if one wishes to read `the current directory'. Since this is always called `.' there is no need to keep track of what the current directory really is. `.' and `..' are `hard links' to the true directories. @cindex Hard links @c ------------------------------------------------------------------------------- @c SECTION @c ------------------------------------------------------------------------------- @node Symbolic links, Hard links, The file hierarchy , Overview @section Symbolic links @cindex Soft links @cindex Symbolic links @vindex @code{ln -s} @cindex @code{ln -s} A symbolic link is a pointer or an alias to another file. The command @smallexample ln -s fromfile /other/directory/tolink @end smallexample @noindent makes the file @code{fromfile} appear to exist at @code{/other/directory/tolink} simultaneously. The file is not copied, it merely appears to be a part of the file tree in two places. Symbolic links can be made to both files and directories. A symbolic link is just a small file which contains the name of the real file one is interested in. It cannot be opened like an ordinary file, but may be read with the C call @code{readlink()} @xref{lstat and readlink}. If we remove the file a symbolic link points to, the link remains -- it just points nowhere. @c ------------------------------------------------------------------------------- @c SECTION @c ------------------------------------------------------------------------------- @node Hard links, , Symbolic links, Overview @section Hard links @cindex Hard links @vindex @code{ln} A @emph{hard link} is a duplicate @emph{inode} in the filesystem which is in every way equivalent to the original file inode. If a file is pointed to by a hard link, it cannot be removed until the link is removed. If a file has @math{n} hard links -- all of them must be removed before the file can be removed. The number of hard links to a file is stored in the filesystem @emph{index node} for the file. @cindex inodes @cindex Index nodes @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node Getting started, The login environment, Overview, Top @chapter Getting started If you have never met unix, or another multiuser system before, then you might find the idea daunting. There are several things you should know. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Logging in:: * Mouse buttons:: * Reading mail:: * Simple commands:: * Text editing and word processing:: @end menu @node Logging in, Mouse buttons, Getting started, Getting started @section Logging in Each time you use unix you must log on to the system by typing a username and a password. Your login name is sometimes called an `account' because some unix systems implement strict quotas for computer resources which have to be paid for with real money@footnote{This is seldom true these days.}. @cartouche @example login: mark password: @end example @end cartouche Once you have typed in your password, you are `logged on'. What happens then depends on what kind of system you are logged onto and how. If you have a colour monitor and keyboard in front of you, with a graphical user interface, you will see a number of windows appear, perhaps a menu bar. You then use a mouse and keyboard just like any other system. This is not the only way to log onto unix. You can also log in remotely, from another machine, using the @code{telnet} or @code{rlogin} @cindex telnet @cindex Logging on programs. If you use these programs, you will normally only get a text or command line interface (though graphics can still be arranged). Once you have logged in, a short message will be printed (called Message of the Day or motd) and you will see the C-shell prompt: the name of the host you are logged onto followed by a percent sign, e.g. @cartouche @example SunOS Release 5.5 Version Generic [UNIX(R) System V Release 4.0] Copyright (c) 1983-1995, Sun Microsystems, Inc. Please report problems to sysadm@@iu.hioslo.no dax% @end example @end cartouche @noindent Remember that every unix machine is a separate entity: it is not like logging onto a PC system where you log onto the `network' i.e. the PC file server. Every unix machine is a server. The network, in unix-land, has lots of players. The first thing you should do once you have logged on is to set a reliable password. A poor password might be okay on a PC which is not attached to a large network, but once you are attached to the internet, you have to remember that the whole world will be trying to crack your password. Don't think that no one will bother: some people really have nothing better to do. A password should not contain any word that could be in a list of words (in any language), or be a simple concatenation of a word and a number (e.g. mark123). It takes seconds to crack such a password. Choose instead something which is easy to remember. Feel free to use the PIN number from your bankers card in your password! This will leave you with fewer things to remember. e.g. Ma9876rk). Passwords can be up to eight characters long. Some sites allow you to change your password anywhere. Other sites require you to log onto a special machine to change your password: @cartouche @example dax% dax% passwd Change your password on host nexus You cannot change it here dax% rlogin nexus password: ****** nexus% passwd Changing password for mark Enter login password: ******** Enter new password: ******** Reenter new passwd: ******** @end example @end cartouche @noindent You will be prompted for your old password and your new password twice. If your network is large, it might take the system up to an hour or two to register the change in your password, so don't forget the old one right away! @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Mouse buttons:: @end menu @node Mouse buttons, Reading mail, Logging in, Getting started @section Mouse buttons Unix has three mouse buttons. On some PC's running GNU/Linux or some other PC unix, there are only two, but the middle mouse button can be simulated by pressing both mouse buttons simultaneously. The mouse buttons have the following general functions. They may also have additional functions in special software. @table @emph @item index finger This is used to select and click on objects. It is also used to mark out areas and copy by dragging. This is the button you normally use. @item middle finger Used to pull down menus. It is also used to paste a marked area somewhere at the mouse position. @item outer finger Pulls down menus. @end table @noindent On a left-handed system right and left are reversed. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Reading mail, Simple commands, Mouse buttons, Getting started @section E-mail Reading electronic mail on unix is just like any other system, but there are many programs to choose from. There are very old programs from the seventies such as @example mail @end example @noindent and there are fully graphical mail programs such as @example tkrat mailtool @end example Choose the program you like best. Not all of the programs support modern multimedia extensions because of their age. Some programs like @code{tkrat} have immediate mail notification alerts. To start a mail program you just type its name. If you have an icon-bar, you can click on the mail-icon. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Simple commands, Text editing and word processing, Reading mail, Getting started @section Simple commands Inexperienced computer users often prefer to use file-manager programs to avoid typing anything. With a mouse you can click your way through directories and files without having to type anything (e.g. the @code{fmgr} or @code{tkdesk} programs). More experienced users generally find this to be slow and tedious after a while and prefer to use written commands. Unix has many short cuts and keyboard features which make typed commands extremely fast and much more powerful than use of the mouse. If you come from a DOS environment, the unix commands can be a little strange. Because they stem from an era when keyboards had to be hit with hammer force, and machines were very slow, the command names are generally as short as possible, so they seem pretty cryptic. Some familar ones which DOS borrowed from unix include, @example cd mkdir @end example @noindent which change to a new directory and make a new directory respectively. To list the files in the current directory you use, @example ls @end example @noindent To rename a file, you `move' it: @example mv @var{old-name} @var{new-name} @end example @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Text editing and word processing, , Simple commands, Getting started @section Text editing and word processing Text editing is one of the things which people spend most time doing on any computer. It is important to distingiush text editing from word processing. On a PC or MacIntosh, you are perhaps used to Word or WordPerfect for writing documents. Unix has a Word-like program called @code{lyx}, but for the most part Unix users do not use word processors. It is more common in the unix community to write all documents, regardless of whether they are letters, books or computer programs, using a non-formatting text editor. (Unix word processors like @code{Framemaker} do exist, but they are very expensive. A version of MS-Word also exists for some unices.) Once you have written a document in a normal text editor, you call up a text formatter to make it pretty. You might think this strange, but the truth of the matter is that this two-stage process gives you the most power and flexibilty---and that is what most unix folks like. For writing programs, or anything else, you edit a file by typing: @example emacs myfile @end example @noindent @code{emacs} is one of dozens of text-editors. It is not the simplest or most intuitive, but it is the most powerful and if you are going to spend time learning an editor, it wouldn't do any harm to make it this one. You could also click on emacs' icon if you are relying on a window system. Emacs is almost certainly the most powerful text editor that exists on any system. It is not a word-processor, it not for formatting printed documents, but it can be linked to almost any other program in order to format and print text. It contains a powerful programming language and has many intelligent features. We shall not go into the details of document formatting in this book, but only mention that programs like @code{troff} and @code{Tex} or @code{Latex} are used for this purpose to obtain typset-quality printing. Text formatting is an area where Unix folks do things differently to PC folks. @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node The login environment, Files and access, Getting started, Top @chapter The login environment @cindex Environment, unix user @cindex User environment @cindex Login environment Unix began as a timesharing mainframe system in the seventies, when the only terminals available were text based @emph{teletype} terminals or @emph{tty}-s. Later, the Massachusetts Institute of Technology (MIT) developed the X-windows interface which is now a standard across UNIX platforms. Because of this history, the X-window system works as a front end to the standard UNIX shell and interface, so to understand the user environment we must first understand the shell. @cindex X-windows @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Shells:: * UNIX command overview:: * Terminals:: * The X window system:: * Multiple screens:: @end menu @node Shells, UNIX command overview, The login environment, The login environment @section Shells @cindex Command interpreter A shell is a command interpreter. In the early days of unix, a shell was the only way of issuing commands to the system. Nowadays many window-based application programs provide menus and buttons to perform simple commands, but the UNIX shell remains the most powerful and flexible way of interacting with the system. @cindex @emph{init} After logging in and entering a password, the unix process @emph{init} starts a shell for the user logging in. Unix has several different kinds of shell to choose from, so that each user can pick his/her favourite command interface. The type of shell which the system starts at login is determined by the user's entry in the @emph{passwd} database. On most systems, the standard login shell is a variant of the C-shell. Shells provide facilities and commands which @itemize @bullet @item Start and stop processes (programs) @item Allow two processes to communicate through a @emph{pipe} @item Allow the user to redirect the flow of input or output @item Allow simple command line editing and command history @item Define aliases to frequently used commands @item Define global "environment" variables which are used to configure the default behaviour of a variety of programs. These lie in an "associated array" for each process and may be seen with the @samp{env} command. Environment variables are inherited by all processes which are started from a shell. @cindex Environment variables @item Provide wildcard expansion (joker notation) of filenames using @samp{*,?,[]} @cindex Wildcards @item Provide a simple script language, with tests and loops, so that users can combine system programs to create new programs of their own. @item Change and remember the location of the current working directory, or location within the file hierarchy. @end itemize The shell does not contain any more specific functions---all other commands, such as programs which list files or create directories etc., are executable programs which are independent of the shell. When you type @samp{ls}, the shell looks for the executable file called @samp{ls} in a special list of directories called @emph{the command path} and attempts to start this program. This allows such programs to be developed and replaced independently of the actual command interpreter. Each shell which is started can be customized and configured by editing a setup file. For the C-shell and its variants this file is called @file{.cshrc}, and for the Bourne shell and its variants it is called @file{.profile}. (Note that files which begin with leading dots are not normally visible with the @samp{ls} command. Use @samp{ls -a} to view these.) Any commands which are placed in these files are interpreted by the shell before the first command prompt is issued. These files are typically used to define a command search path and terminal characteristics. @emph{On each new command line you can use the cursor keys to edit the line. The up-arrow browses back through earlier commands. @kbd{CTRL-a} takes you to the start of the line. @kbd{CTRL-e} takes you to the end of the line. The @key{TAB} can be used to save typing with the `completion' facility @xref{Command/filename completion}.} @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Shell commands generally:: * Environment and shell variables:: * Wildcards:: * Regular expressions:: * Nested shells and reverse apostrophes:: @end menu @node Shell commands generally, Environment and shell variables, Shells, Shells @subsection Shell commands generally Shell commands are commands like @code{cp}, @code{mv}, @code{passwd}, @code{cat}, @code{more}, @code{less}, @code{cc}, @code{grep}, @code{ps} etc.. Very few commands are actually built into the shell command line interpreter, in the way that they are in DOS -- commands are mostly programs which exist as files. When we type a command, the shell searches for a program with the same name and tries to execute it. The file must be executable, or a @code{Command not found} error will result. To see what actually happens when you type a command like @code{gcc}, try typing in the following C-shell commands @emph{directly into a C-shell}. (We shall discuss these commands soon.) @smallexample foreach dir ( $path ) # for every directory in the list path if ( -x $dir/gcc ) then # if the file is executable echo Found $dir/gcc # Print message found! break # break out of loop else echo Searching $dir/gcc endif end @end smallexample @noindent The output of this command is something like @smallexample Searching /usr/lang/gcc Searching /usr/openwin/bin/gcc Searching /usr/openwin/bin/xview/gcc Searching /physics/lib/framemaker/bin/gcc Searching /physics/motif/bin/gcc Searching /physics/mutils/bin/gcc Searching /physics/common/scripts/gcc Found /physics/bin/gcc @end smallexample If you type @smallexample echo $path @end smallexample you will see the entire list of directories which are searched by the shell. If we had left out the `break' command, we might have discovered that UNIX often has several programs with the same name, in different directories! For example, @smallexample /bin/mail /usr/ucb/mail /bin/Mail /bin/make /usr/local/bin/make. @end smallexample Also, different versions of unix have different conventions for placing the commands in directories, so the path list needs to be different for different types of unix machine. In the C-shell a few basic commands like @code{cd} and @code{kill} are built into the shell (as in DOS). You can find out which directory a command is stored in using the @smallexample which @end smallexample command. For example @smallexample nexus% which cd cd: shell built-in command. nexus% which cp /bin/cp nexus% @end smallexample @code{which} only searches the directories in @code{$path} and quits after the first match, so if there are several commands with the same name, you will only see the first of them using which. @cindex path @cindex Command path Finally, in the C-shell, the @code{which} command is built in. In the Bourne shell it is a program: @smallexample nexus% which which which: shell built-in command. nexus% sh $ which which /bin/which $ exit nexus% @end smallexample Take a look at the script @code{/usr/ucb/which}. It is a script written in the C-shell. @cindex which command @vindex which @c ..................................................... @c SUBSECTION @c ..................................................... @node Environment and shell variables, Wildcards, Shell commands generally, Shells @subsection Environment and shell variables @cindex Environment variables Environment variables are variables which the shell keeps. They are normally used to configure the behaviour of utility programs like @code{lpr} (which sends a file to the printer) and @code{mail} (which reads and sends mail) so that special options do not have to be typed in every time you run these programs. Any program can read these variables to find out how you have configured your working environment. We shall meet these variables frequently. Here are some important variables @vindex PATH @vindex TERM @vindex DISPLAY @vindex LD_LIBRARY_PATH @vindex HOST @vindex PRINTER @vindex HOME @smallexample PATH # The search path for shell commands (sh) TERM # The terminal type (sh and csh) DISPLAY # X11 - the name of your display LD_LIBRARY_PATH # Path to search for object and shared libraries HOST # Name of this unix host PRINTER # Default printer (lpr) HOME # The path to your home directory (sh) path # The search path for shell commands (csh) term # The terminal type (csh) noclobber # See below under redirection prompt # The default prompt for csh home # The path to your home directory (csh) @end smallexample These variables fall into two groups. Traditionally the first group always have names in uppercase letters and are called @emph{environment variables}, whereas variables in the second group have names with lowercase letters and are called @emph{shell variables}-- but this is only a convention. The uppercase variables are @emph{global variables}, whereas the lower case variables are @emph{local variables}. Local variables are not defined for programs or sub-shells started by the current shell, while global variables are inherited by all sub-shells. @cindex Global variables @cindex Variables, global @cindex Local variables @cindex Variables, local The Bourne-shell and the C-shell use these conventions differently and not always consistently. You will see how to define these below. For now you just have to know that you can use the following commands from the C-shell to list these variables. The command @code{env} can be used in either C-shell or Bourne shell to see all of the defined environment variables. @cindex @code{env} command @vindex env @c ..................................................... @c SUBSECTION @c ..................................................... @node Wildcards, Regular expressions, Environment and shell variables, Shells @subsection Wildcards @cindex Wildcards @cindex Joker notation @vindex ? @vindex * @vindex [] @cindex Matching filenames Sometimes you want to be able to refer to several files in one go. For instance, you might want to copy all files ending in `.c' to a new directory. To do this one uses @emph{wildcards}. Wildcards are characters like @code{* ?} which stand for any character or group of characters. In card games the joker is a `wild card' which can be substituted for any other card. Use of wildcards is also called @emph{filename substitution} in the unix manuals, in the sections on @code{sh} and @code{csh}. The wildcard symbols are, @table @samp @item ? Match single character. e.g. @code{ls /etc/rc.????} @item * Match any number of characters. e.g. @code{ls /etc/rc.*} @item [...] Match any character in a list enclosed by these brackets. e.g. @code{ls [abc].C} @end table @noindent Here are some examples and explanations. @table @samp @item /etc/rc.???? Match all files in /etc whose first three characters are @code{rc.} and are 7 characters long. @item *.c Match all files ending in `.c' i.e. all C programs. @item *.[Cc] List all files ending on `.c' or `.C' i.e. all C and C++ programs. @item *.[a-z] Match any file ending in .a, .b, .c, ... up to .z etc. @end table @noindent @cindex Matching filenames It is important to understand that @emph{the shell expands wildcards}. When you type a command, the program is not invoked with an argument that contains @code{*} or @code{?}. The shell expands the special characters first and invokes commands with the entire list of files which match the patterns. The programs never see the wildcard characters, only the list of files they stand for. To see this in action, you can type @smallexample echo /etc/rc.* @end smallexample which gives @smallexample /etc/rc /etc/rc.boot /etc/rc.ip /etc/rc.local /etc/rc.local% /etc/rc.local~ /etc/rc.single /etc/rc~ @end smallexample All shell commands are invoked with a command line of this form. This has an important corollary. It means that multiple renaming @emph{cannot work}! Unix files are renamed using the @code{mv} command. In many microcomputer operating systems one can write @smallexample rename *.x *.y @end smallexample @noindent which changes the file extension of all files ending in `.x' to the same name with a `.y' extension. This cannot work in unix, because the shell tries expands everything before passing the arguments to the command line. @emph{The local shell variable @code{noglob} switches off wildcard expansion in the C shell, but you still cannot rename multiple files using @code{mv}. Some free-software programs make this possible.} @c ..................................................... @c SUBSECTION @c ..................................................... @node Regular expressions, Nested shells and reverse apostrophes, Wildcards, Shells @subsection Regular expressions @cindex Matching strings @cindex Regular expressions @cindex Expressions, regular The wildcards belong to the shell. They are used for matching filenames. UNIX has a more general and widely used mechanism for matching @emph{strings}, this is through @emph{regular expressions}. Regular expressions are used by the @code{egrep} utility, text editors like @code{ed}, @code{vi} and @code{emacs} and @code{sed} and @code{awk}. They are also used in the C programming language for matching input as well as in the Perl programming language and @code{lex} tokenizer. Here are some examples using the @code{egrep} command which print lines from the file @code{/etc/rc} which match certain conditions. The @code{} contruction is part of @code{egrep}. Everything in between these symbols is a regular expression. Notice that special shell symbols @code{! * &} have to be preceded with a backslash @code{\} in order to prevent the shell from expanding them! @cindex @code{egrep} command @smallexample # Print all lines beginning with a comment # egrep '(^#)' /etc/rc # Print all lines which DON'T begin with # egrep '(^[^#])' /etc/rc # Print all lines beginning with e, f or g. egrep '(^[efg])' /etc/rc # Print all lines beginning with uppercase egrep '(^[A-Z])' /etc/rc # Print all lines NOT beginning with uppercase egrep '(^[^A-Z])' /etc/rc # Print all lines containing ! * & egrep '([\!\*\&])' /etc/rc # All lines containing ! * & but not starting # egrep '([^#][\!\*\&])' /etc/rc @end smallexample @noindent Regular expressions are made up of the following `atoms'. @noindent These examples assume that the file @file{/etc/rc} exists. If it doesn't exist on the machine you are using, try to find the equivalent by, for instance, replacing @code{/etc/rc} with @code{/etc/rc*} which will try to find a match beginning with the rc. @vindex . in regular expressions @vindex ^ in regular expressions @vindex $ in regular expressions @vindex [] in regular expressions @vindex * in regular expressions @vindex + in regular expressions @vindex ? in regular expressions @cindex . in regular expressions @cindex ^ in regular expressions @cindex $ in regular expressions @cindex [] in regular expressions @cindex * in regular expressions @cindex + in regular expressions @cindex ? in regular expressions @table @samp @item . Match any single character except the end of line. @item ^ Match the beginning of a line as the first character. @item $ Match end of line as last character. @item [..] Match any character in the list between the square brackets.(see below). @item * Match zero or more occurrances of the preceding expression. @item + Match one or more occurrences of the preceding expression. @item ? Match zero or one occurrance of the preceding expression. @end table You can find a complete list in the unix manual pages. The square brackets above are used to define a @emph{class} of characters to be matched. Here are some examples, @itemize @bullet @item If the square brackets contain a list of characters, $[a-z156]$ then a single occurrance of any character in the list will match the regular expression: in this case any lowercase letter or the numbers 1, 5 and 6. @item If the first character in the brackets is the caret symbol @samp{^} then any character @emph{except} those in the list will be matched. @item Normally a dash or minus sign @samp{-} means a range of characters. If it is the first character after the @samp{[} or after @samp{[^} then it is treated literally. @end itemize @c ..................................................... @c SUBSECTION @c ..................................................... @node Nested shells and reverse apostrophes, , Regular expressions, Shells @subsection Nested shell commands and `` @vindex `` shell construction @vindex `...` The backwards apostrophes `...` can be used in all shells and also in the programming language Perl. When these are encountered in a string the shell tries to execute the command inside the quotes and replace the quoted expression by the result of that command. For example: @smallexample unix% echo "This system's kernel type is `/bin/file /vmunix`" This system's kernel type is /vmunix: sparc executable not stripped unix% foreach file ( `ls /etc/rc*` ) ? echo I found a config file $file ? echo Its type is `/bin/file $file` ? end I found a config file /etc/rc Its type is /etc/rc: executable shell script I found a config file /etc/rc.boot Its type is /etc/rc.boot: executable shell script I found a config file /etc/rc.ip Its type is /etc/rc.ip: executable shell script I found a config file /etc/rc.local Its type is /etc/rc.local: ascii text I found a config file /etc/rc.local~ Its type is /etc/rc.local~: ascii text I found a config file /etc/rc.single Its type is /etc/rc.single: executable shell script I found a config file /etc/rc~ Its type is /etc/rc~: executable shell script @end smallexample This is how we insert the result of a shell command into a text string or variable. @cindex Inserting a command into a string @cindex Result of a command into a string @cindex Getting command output into a string @c ------------------------------------------------------------------------------- @c SECTION @c ------------------------------------------------------------------------------- @node UNIX command overview, Terminals, Shells, The login environment @section UNIX command overview @menu * Keys:: * Alternative shells:: * terminal emulators:: * Remote login:: * Text editors:: * File handling:: * File browsers:: * Access control:: * Extracting from files:: * Locating files:: * Disk usage:: * Other users:: * Contacting users:: * Mail:: * File transfer:: * Compilers:: * Interpreted languages:: * System statistics:: * System identity:: * Internet resources:: * Text/postscript:: * Pictures/graphics:: * Miscellaneous:: @end menu @node Keys, Alternative shells, UNIX command overview, UNIX command overview @subsection Important keys @table @key @item CTRL-A Jump to start of line. If @samp{screen} is active, this prefixes all control key commands for @samp{screen} and then the normal @kbd{CTRL-A} is replaced by @kbd{CTRL-a a}. @cindex @key{CTRL-A} @vindex @key{CTRL-A} @item CTRL-C Interrupt or break key. Sends signal 15 to a process. @cindex @key{CTRL-C} @vindex @key{CTRL-C} @item CTRL-D Signifies @samp{EOF} (end of file) or shows expansion matches in command/filename completion @xref{Command/filename completion}. @cindex @key{CTRL-D} @vindex @key{CTRL-D} @item CTRL-E Jump to end of line. @cindex @key{CTRL-E} @vindex @key{CTRL-E} @item CTRL-L Clear screen in newer shells and in emacs. Same as @samp{clear} in the shell. @cindex @key{CTRL-L} @vindex @key{CTRL-L} @item CTRL-Z Suspend the present process, but do not destroy it. This sends signal 18 to the process. @cindex @key{CTRL-Z} @vindex @key{CTRL-Z} @end table @c ---------------------------------------------- @node Alternative shells, terminal emulators, Keys, UNIX command overview @subsection Alternative shells @cindex @key{Shells, various} @table @code @item bash The Bourne Again shell, an improved sh. @cindex bash @item csh The standard C-shell. @cindex csh @item jsh The same as sh, with C-shell style job control. @cindex jsh @item ksh The Korn shell, an improved sh. @cindex ksh @item sh The original Bourne shell. @cindex sh @item sh5 On ULTRIX systems the standard Bourne shell is quite stupid. sh5 corresponds to the normal Bourne shell on these systems. @cindex sh5 @item tcsh An improved C-shell. @cindex tcsh @item zsh An improved sh. @cindex zsh @end table @c ---------------------------------------------- @node terminal emulators, Remote login, Alternative shells, UNIX command overview @subsection Window based terminal emulators @cindex Terminals @cindex Command window @table @code @item xterm The standard X11 terminal window. @cindex @samp{xterm} @vindex @samp{xterm} @item shelltool, cmdtool Openwindows terminals from Sun Microsystems. These are not completely X11 compatible during copy/paste operations. @cindex @samp{shelltool} @vindex @samp{shelltool} @cindex @samp{cmdtool} @vindex @samp{cmdtool} @item screen This is not a window in itself, but allows you to emulate having several windows inside a single (say) xterm window. The user can switch between different windows and open new ones, but can only see one window at a time @xref{Multiple screens}. @cindex @samp{screen} @vindex @samp{screen} @end table @c ---------------------------------------------- @node Remote login, Text editors, terminal emulators, UNIX command overview @subsection Remote shells and logins @table @code @item rlogin Login onto a remote unix system. @cindex @samp{rlogin} @vindex @samp{rlogin} @item rsh Open a shell on a remote system (require access rights). @cindex @samp{rsh} @vindex @samp{rsh} @item telnet Open a connection to a remove system using the telnet protocol. @cindex @samp{telnet} @vindex @samp{telnet} @end table @c ---------------------------------------------- @node Text editors, File handling, Remote login, UNIX command overview @subsection Text editors @table @code @item ed An ancient line-editor. @cindex @samp{ed} @vindex @samp{ed} @item vi Visual interface to @code{ed}. This is the only "standard" unix text editor supplied by vendors. @cindex @samp{vi} @vindex @samp{vi} @item emacs The most powerful UNIX editor. A fully configurable, user programmable editor which works under X11 and on tty-terminals. @cindex @samp{emacs} @vindex @samp{emacs} @item xemacs A pretty version of emacs for X11 windows. @cindex @samp{xemacs} @vindex @samp{xemacs} @item pico A tty-terminal only editor, comes as part of the PINE mail package. @cindex @samp{pico} @vindex @samp{pico} @item xedit A test X11-only editor supplied with X-windows. @cindex @samp{xedit} @vindex @samp{xedit} @item textedit A simple X11-only editor supplied by Sun Microsystems. @cindex @samp{textedit} @vindex @samp{textedit} @end table @c ---------------------------------------------- @node File handling, File browsers, Text editors, UNIX command overview @subsection File handling commands @table @code @item ls List files in specified directory (like @code{dir} on other systems). @cindex @samp{ls command} @vindex @samp{ls} @item cp Copy files. @cindex @samp{cp} command @vindex @samp{cp} @item mv Move or rename files. @cindex @samp{mv} command @cindex Renaming files @cindex Moving files @vindex @samp{mv} @item touch Creates an empty new file if none exists, or updates date and time stamps on existing files. @cindex @samp{touch} command @vindex @samp{touch} @cindex Creating files @cindex Time stamp, updating @cindex Date stamp, updating @cindex Updating file time stamp @item rm, unlink Remove a file or link (delete). @cindex @samp{rm} command @vindex @samp{unlink} @cindex @samp{unlink} command @vindex @samp{unlink} @item mkdir, rmdir Make or remove a directory. A directory must be empty in order to be able to remove it. @cindex @samp{mkdir} command @vindex @samp{mkdir} @cindex Creating directories @cindex Directories, creating @cindex Directories, deleting @cindex @samp{rmdir} command @vindex @samp{rmdir} @item cat Concatenate or join together a number of files. The output is written to the standard output by default. Can also be used to simply print a file on screen. @cindex @samp{cat} command @vindex @samp{cat} @item lp, lpr Line printer. Send a file to the default printer, or the printer defined in the @samp{PRINTER} evironment variable. @cindex @samp{PRINTER} variable @vindex @samp{PRINTER} @cindex @samp{lp} command @vindex @samp{lp} @cindex @samp{lpr} command @vindex @samp{lp} @cindex Printing a file @item lpq, lpstat Show the status of the print queue. @cindex @samp{lpq} @vindex @samp{lpq} @cindex Printer queue @cindex Printer status @cindex @samp{lpstat} @vindex @samp{lpstat} @end table @c ---------------------------------------------- @node File browsers, Access control, File handling, UNIX command overview @subsection File browsing @table @code @item more Shows one screen full at a time. Possibility to search for a string and edit the file. This is like @samp{type @var{file} | more} in DOS. @cindex @samp{more} command @vindex @samp{more} @cindex Viewing a file @cindex Browsing through a file @cindex @samp{type} in DOS @item less An enhanced version of more. @cindex @samp{less} command @vindex @samp{less} @cindex nc @cindex mc @item mc Midnight commander, a free version of the `Norton Commander' PC utility for unix. (Only for non-serious UNIX users...) @item fmgr A window based file manager with icons and all that nonsense. @end table @c ---------------------------------------------- @node Access control, Extracting from files, File browsers, UNIX command overview @subsection Ownership and granting access permission @cindex Access control @cindex Granting permission @table @code @item chmod Change file access mode. @cindex @samp{chmod} command @vindex @samp{chmod} @cindex Changing file mode @cindex File mode, changing @item chown, chgrp Change owner and group of a file. The GNU version of @code{chown} allows both these operations to be performed together using the syntax @code{chown @var{owner}.@var{group} @var{file}}. @cindex @samp{chown} command @vindex @samp{chown} @cindex @samp{chgrp} command @vindex @samp{chgrp} @item acl On newer Unices, Access control lists allow access to be granted on a per-user basis rather than by groups. @cindex Access control lists @cindex ACLs @cindex @samp{fmgr} file manager @vindex @samp{fmgr} @end table @c ---------------------------------------------- @node Extracting from files, Locating files, Access control, UNIX command overview @subsection Extracting from and rebuilding files @table @code @item cut Extract a column in a table @cindex @samp{cut} command @vindex @samp{cut} @item paste Merge several files so that each file becomes a column in a table. @cindex @samp{paste} command @vindex @samp{paste} @item sed A batch text-editor for searching, replacing and selecting text without human intervention. @cindex @samp{sed} editor @vindex @samp{sed} @item awk A prerunner to the Perl language, for extracting and modifying textfiles. @cindex @samp{awk} @vindex @samp{awk} @item rmcr Strip carriage return (ASCII 13) characters from a file. Useful for converting DOS files to unix. @end table @c ---------------------------------------------- @node Locating files, Disk usage, Extracting from files, UNIX command overview @subsection Locating files @table @code @item find Search for files from a specified directory using various criteria. @cindex @samp{find} command @vindex @samp{find} @item locate Fast search in a global file database for files containing a search-string. @cindex @samp{locate} command @vindex @samp{locate} @item whereis Look for a command and its documentation on the system. @cindex @samp{whereis} command @vindex @samp{whereis} @end table @c ---------------------------------------------- @node Disk usage, Other users, Locating files, UNIX command overview @subsection Disk usage. @cindex Disk usage. @table @code @item du Show number of blocks used by a file or files. @cindex @samp{du} command @vindex @samp{du} @item df Show the state of usage for one or more disk partitions. @cindex @samp{df} command @vindex @samp{df} @end table @c ---------------------------------------------- @node Other users, Contacting users, Disk usage, UNIX command overview @subsection Show other users logged on @table @code @item users Simple list of other users. @cindex @samp{users} command @vindex @samp{users} @item finger Show who is logged onto this and other systems. @cindex @samp{finger} service @vindex @samp{finger} @item who List of users logged into this system. @cindex @samp{who} command @vindex @samp{who} @item w Long list of who is logged onto this system and what they are doing. @cindex @samp{w} command @vindex @samp{w} @end table @node Contacting users, Mail, Other users, UNIX command overview @subsection Contacting other users @table @code @item write Send a simple message to the named user, end with @key{CTRL-D}. The command @samp{mesg n} switches off messages receipt. @cindex Sending messages @cindex Messages @cindex @samp{write} command @vindex @samp{write} @cindex @samp{mesg} @vindex @samp{mesg} @item talk Interactive two-way conversation with named user. @cindex @samp{talk} service @vindex @samp{talk} @item irc Internet relay chat. A conferencing system for realtime multi-user conversations, for addicts and losers. @cindex @samp{IRC} @cindex Internet relay chat @vindex @samp{irc} @end table @c ---------------------------------------------- @node Mail, File transfer, Contacting users, UNIX command overview @subsection Mail senders/readers @cindex Mail clients @table @code @item mail The standard (old) mail interface. @item Mail Another mail interface. @item elm Electronic Mail program. Lots of functionality but poor support for multimedia. @cindex @samp{elm} mailer @vindex @samp{elm} @item pine Rumours (untrue) are that pine stands for Pine is Not Elm; it actually stands for nothing at all. Improved support for multimedia but very slow and rather stupid at times. Some of the best features of elm have been removed! @cindex @samp{pine} mailer @vindex @samp{pine} @item mailtool Sun's openwindows client program. @item rmail A mail interface built into the emacs editor. @cindex @samp{rmail} in emacs @vindex @samp{rmail} @item netscape mail A mail interface built into the netscape navigator. @item zmail A commerical mail package. @cindex @samp{zmail} client @vindex @samp{zmail} @item tkrat A graphical mail reader which supports most MIME types, written in tcl/tk. This program has a nice feel and allows you to create a searchable database of old mail messages, but has a hopeless locking mechanism. @end table @c ---------------------------------------------- @node File transfer, Compilers, Mail, UNIX command overview @subsection File transfer @cindex File transfer @table @code @item ftp The File Transfer program - copies files to/from a remote host. @cindex @samp{ftp} program @vindex @samp{ftp} @item ncftp An enhanced ftp for anonymous login. @cindex @samp{ncftp} program @vindex @samp{ncftp} @end table @c ---------------------------------------------- @node Compilers, Interpreted languages, File transfer, UNIX command overview @subsection Compilers @cindex Compilers @table @code @item cc The C compiler. @cindex @samp{cc} @vindex @samp{cc} @item CC The C++ compiler. @cindex @samp{CC} @vindex @samp{CC} @item gcc The GNU C compiler. @cindex @samp{gcc} @vindex @samp{gcc} @item g++ The GNU C++ compiler. @cindex @samp{g++} @vindex @samp{g++} @item ld The system linker/loader. @cindex @samp{ld} loader/linker @vindex @samp{ld} @item ar Archive library builder. @cindex @samp{ar} archiver @vindex @samp{ar} @item dbx A symbolic debugger. @cindex @samp{dbx} debugger @cindex Debugger @vindex @samp{dbx} @item gdb The GNU symbolic debugger. @cindex @samp{gdb} debugger @vindex @samp{gdb} @item xxgdb The GNU debugger with a windown driven front-end. @item ddd A motif based front-end to the gdb debugger. @vindex ddd @end table @cindex @samp{xxgdb} @cindex Debugger GUI @vindex @samp{xxgdb} @c ---------------------------------------------- @node Interpreted languages, System statistics, Compilers, UNIX command overview @subsection Other interpreted languages @table @code @item perl Practical extraction an report language. @item tcl A perl-like language with special support for building user interfaces and command shells. @cindex @samp{TCL} @vindex @samp{tcl} @item scheme A lisp-like extensible scripting language from GNU. @cindex scheme @item mercury A prolog-like language for artificial intelligence. @cindex Mercury @end table @node System statistics, System identity, Interpreted languages, UNIX command overview @subsection Processes and system statistics @table @code @item ps List system process table. @cindex @samp{ps} command @vindex @samp{ps} @item vmstat List kernel virtual-memory statistics. @cindex @samp{vmstat} virtual memory stats @vindex @samp{vmstat} @item netstat List network connections and statistics. @cindex @samp{netstat} network statistics @vindex @samp{netstat} @item rpcinfo Show rpc information. @cindex @samp{rpcinfo} @vindex @samp{rcpinfo} @item showmount Show clients mounting local filesystems. @cindex @samp{showmount} @cindex Mounted file systems @vindex @samp{showmount} @end table @node System identity, Internet resources, System statistics, UNIX command overview @subsection System identity @table @code @item uname Display system name and operating system release. @cindex @samp{System details} @vindex @samp{uname} @cindex Operating system name @cindex System name @item hostname Show the name of this host. @cindex @samp{hostname} command @vindex @samp{hostname} @item domainname Show the name of the local NIS domain. Normally this is chosen to be the same as the BIND/DNS domain, but it need not be. @cindex @samp{domainname} command @vindex @samp{domainname} @cindex Domainname @item nslookup Interrogate the DNS/BIND name service (hostname to IP address conversion). @cindex @samp{nslookup} command @vindex @samp{nslookup} @end table @c ---------------------------------------------- @node Internet resources, Text/postscript, System identity, UNIX command overview @subsection Internet resources @cindex Internet resources @table @code @item archie, xarchie Search the internet ftp database for files. @cindex @samp{archie} program @vindex @samp{archie} @cindex FTP resources, finding @cindex Finding FTP files @cindex @samp{xarchie} client @vindex @samp{xarchie} @item xrn, fnews Read news (browser). @cindex @samp{xrn} news reader @vindex @samp{xrn} @cindex @samp{fnews} news reader @vindex @samp{fnews} @item netscape, xmosaic Read world wide web (WWW) (browser). @cindex @samp{mosaic} @vindex @samp{xmosaic} @end table @c ---------------------------------------------- @node Text/postscript, Pictures/graphics, Internet resources, UNIX command overview @subsection Text formatting and postscript @cindex Postscript viewers @cindex Text formatting @table @code @item tex, latex Donald Knuth's text formatting language, pronounced "tek" (the x is really a greek "chi"). Used widely for technical publications. Compiles to dvi (device independent) file format. @cindex @samp{tex} @vindex @samp{tex} @cindex @samp{latex} @vindex @samp{latex} @item texinfo A hypertext documentation system using tex and "info" format. This is the GNU documentation system. This UNIX guide is written in texinfo!!! @cindex Hypertext @cindex @samp{texinfo} system @vindex @samp{texinfo} @item xdvi View a tex dvi file on screen. @vindex @samp{xdvi} @item dvips Convert dvi format into postscript. @cindex dvi to postscript @vindex @samp{dvips} @item ghostview, ghostscript View a postscript file on screen. @cindex @samp{ghostview} postscript previewer @vindex @samp{ghostview} @cindex @samp{ghostscript} "GNU postscript" interpreter @vindex @samp{ghostscript} @end table @c ---------------------------------------------- @node Pictures/graphics, Miscellaneous, Text/postscript, UNIX command overview @subsection Picture editors and processors @cindex Picture processing @table @code @item xv Handles, edits and processes pictures in a variety of standard graphics formats (gif, jpg, tiff etc). Use @code{xv -quit} to place a picture on your root window. @cindex @samp{xv} picture processor @vindex @samp{xv} @cindex gif @cindex jpg @cindex tiff @item xpaint A simple paint program. @cindex @samp{xpaint} program @vindex @samp{xpaint} @cindex Painting program @item xfig A line drawing figure editor. Produces postscript, tex, and a variety of other output formats. @cindex @samp{xfig} drawing program @vindex @samp{xfig} @cindex Drawing program @item xsetroot Load an X-bitmap image into the screen (root window) background. Small images are tiled. @cindex @samp{Background picture} @vindex @samp{setroot} @end table @c ---------------------------------------------- @node Miscellaneous, , Pictures/graphics, UNIX command overview @subsection Miscellaneous @table @code @item date Print the date and time. @cindex @samp{date} command @vindex @samp{date} @cindex Time and date @item ispell Spelling checker. @cindex Spelling checker @vindex @samp{ispell} @item xcalc A graphical calculator. @cindex Calculator, X windows @vindex @samp{xcalc} @item dc,bc Text-based calculators. @cindex Calculator, shell @vindex @samp{dc} @item xclock A clock! @item ping Send a "sonar" ping to see if another unix host is alive. @cindex @samp{Sonar} ping @vindex @samp{ping} @cindex Testing reponse from other hosts. @end table @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Terminals, The X window system, UNIX command overview, The login environment @section Terminals @cindex tty @cindex Teletype terminal @cindex xterm program In order to communicate with a user, a shell needs to have access to a terminal. Unix was designed to work with many different kinds of terminals. Input/output commands in Unix read and write to a virtual terminal. In @emph{reality} a terminal might be a text-based Teletype terminal (called a @emph{tty} for short) or a graphics based terminal; it might be 80-characters wide or it might be wider or narrower. Unix take into account these possibility by defining a number of instances of terminals in a more or less object oriented way. Each user's terminal has to be configured before cursor based input/output will work correctly. Normally this is done by choosing one of a number of standard terminal types a list which is supplied by the system. In practice the user defines the value of the environment variable @samp{TERM} to an appropriate name. Typical examples are @samp{vt100} and @samp{xterm}. If no standard setup is found, the terminal can always be configured manually using UNIX's most cryptic and opaque of commands: @samp{stty}. The job of configuring terminals is much easier now that hardware is more standard. Users' terminals are usually configured centrally by the system administrator and it is seldom indeed that one ever has to choose anything other than @samp{vt100} or @samp{xterm}. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node The X window system, Multiple screens, Terminals, The login environment @section The X window system Because UNIX originated before windowing technology was available, the user-interface was not designed with windowing in mind. The X window system attempts to be like a virtual machine park, running a different program in each window. Although the programs appear on one screen, they may in fact be running on unix systems anywhere in the world, with only the output being local to the user's display. The standard shell interface is available by running an X client application called @samp{xterm} which is a graphical front-end to the standard UNIX textual interface. @cindex rlogin program The @samp{xterm} program provides a virtual terminal using the X windows graphical user interface. It works in exactly the same way as a @emph{tty} terminal, except that standard graphical facilities like copy and paste are available. Moreover, the user has the convenience of being able to run a different shell in every window. For example, using the @samp{rlogin} command, it is possible to work on the local system in one window, and on another remote system in another window. The X-window environment allows one to cut and paste between windows, regardless of which host the shell runs on. @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Components of X:: * How to set up X windows:: * X displays and authority:: @end menu @node Components of X, How to set up X windows, The X window system, The X window system @subsection The components of the X-window system The X11 system is based on the client-server model. You might wonder why a window system would be based on a model which was introduced for interprocess communication, or network communication. The answer is straightforward. @cindex X window system The designers of the X window system realized that network communication was to be the paradigm of the next generation of computer systems. They wanted to design a system of windows which would enable a user to sit at a terminal in Massachusetts and work on a machine in Tokyo -- and still be able to get high quality windows displayed on their terminal. The aim of X windows from the beginning is to create a @emph{distributed} window environment. When I log onto my friend's Hewlett Packard workstation to use the text editor (because I don't like the one on my EUNUCHS workstation) I want it to work correctly on my screen, with my keyboard -- even though my workstation is manufactured by a different company. I also want the colours to be right despite the fact that the HP machine uses a completely different video hardware to my machine. When I press the curly brace key @code{@{}, I want to see a curly brace, and not some hieroglyphic because the HP station uses a different keyboard. These are the problems which X tries to address. In a network environment we need a @emph{common window system} which will work on any kind of hardware, and hide the differences between different machines as far as possible. But it has to be flexible enough to allow us to change all of the things we don't like -- to choose our own colours, and the kind of window borders we want etc. Other windowing systems (like Microsoft windows) ignore these problems and thereby lock the user to a single vendors products and a single operating system. (That, of course, is no accident.) The way X solves this problem is to use the client server model. Each program which wants to open a window on somebody's compute screen is a client of the @emph{X window service}. To get something drawn on a user's screen, the client asks a server on the host of interest to draw windows for it. No client ever draws anything itself -- it asks the server to do it on its behalf. There are several reasons for this: @itemize @bullet @item The clients can all talk a common `window language' or @emph{protocol}. We can hide the difference between different kinds of hardware by making the @emph{machine-specific} part of drawing graphics entirely a problem of implementing the server on the particular hardware. When a new type of hardware comes along, we just need to modify the server -- none of the clients need to be modified. @cindex X protocol @item We can contact different servers and send our output to different hardware -- thus even though a program is running on a CPU in Tokyo, it can ask the server in Massachusetts to display its window for it. @item When more than one window is on a user's display, it eventually becomes necessary to move the windows around and then figure out which windows are on top of which other windows etc. If all of the drawing information is kept in a server, it is straightforward to work out this information. If every client drew where it wanted to, it would be impossible to know which window was supposed to be on top of another. @cindex X display @cindex Display, X @end itemize In X, the window manager is a different program to the server which does the drawing of graphics -- but the client-server idea still applies, it just has one more piece to its puzzle. @c ..................................................... @c SUBSECTION @c ..................................................... @node How to set up X windows, X displays and authority, Components of X, The X window system @subsection How to set up X windows The X windows system is large and complex and not particularly user friendly. When you log in to the system, X reads two files in your home directory which decide which applications will be started what they will look like. The files are called @vindex .xsession @cindex .xsession file @cindex Setting up the x environment @table @emph @item .Xsession This file is a shell script which starts up a number of applications as background processes and exits by calling a window manager. Here is a simple example file @smallexample #!/bin/csh # # .xsession file # # setenv PATH /usr/bin:/bin:/local/gnu/bin:/usr/X11R6/bin # # List applications here, with & at the end # so they run in the background # xterm -T NewTitle -sl 1000 -geometry 90x45+16+150 -sb & xclock & xbiff -geometry 80x80+510+0 & # Start a window manager. Exec replaces this script with # the fvwm process, so that it doesn't exist as a separate # (useless) process. exec /local/bin/fvwm @end smallexample @item .Xdefaults This file specifies all of the resources which X programs use. It can be used to change the colours used by applications, or font types etc. The subject of X-rescources is a large one and we don't have time for it here. Here is a simple example, which shows how you can make your over-bright xterm and emacs windows less bright grey shade. @smallexample xterm*background: LightGrey Emacs*background: grey92 Xemacs*background: grey92 @end smallexample @end table @c ..................................................... @c SUBSECTION @c ..................................................... @node X displays and authority, , How to set up X windows, The X window system @subsection X displays and authority In the terminology used by X11, every client program has to contact a @emph{display} in order to open a window. A display is a virtual screen which is created by the X server on a particular host. X can create several separate displays on a given host, though most machines only have one. @cindex DISPLAY variable @cindex X display When an X client program wants to open a window, it looks in the UNIX environment variable @samp{DISPLAY} for the IP address of a host which has an X server it can contact. For example, if we wrote @example setenv DISPLAY myhost:0 @end example the client would try to contact the X server on `myhost' and ask for a window on display number zero (the usual display). If we wrote @example setenv DISPLAY 198.112.208.35:0 @end example @noindent the client would try to open display zero on the X server at the host with the IP address @samp{198.112.208.35}. Clearly there must be some kind of security mechanism to prevent just anybody from opening windows on someone's display. X has two such mechanisms: @cindex X windows authentification @cindex X windows access @table @emph @cindex xhost mechanism @item xhost This mechanism is now obsolete. The @samp{xhost} command is used to define a list of hosts which are allowed to open windows on the user's display. It cannot destinguish between individual users. i.e. the command @code{xhost yourhost} would allow @emph{anyone} using yourhost to access the local display. This mechanism is only present for backward compatibility with early versions of X windows. Normally one should use the command @code{xhost -} to exclude all others from accessing the display. @item Xauthority @cindex Xauthority mechanism @cindex X access control The Xauthority mechanism has replaced the xhost scheme. It provides a security mechanism which can distinguish individual users, not just hosts. In order for a user to open a window on a display, he/she must have a ticket---called a "magic cookie". This is a binary file called @file{.Xauthority} which is created in the user's home directory when he/she first starts the X-windows system. Anyone who does not have a recent copy of this file cannot open windows or read the display of the user's terminal. This mechanism is based on the idea that the user's home directory is available via NFS on all hosts he/she will log onto, and thus the owner of the display will always have access to the magic cookie, and will therefore always be able to open windows on the display. Other users must obtain a copy of the file in order to open windows there. The command @code{xauth} is an interactive utility used for controlling the contents of the @file{.Xauthority} file. See the @samp{xauth} manual page for more information. @end table @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Multiple screens, , The X window system, The login environment @section Multiple screens The window paradigm has been very successful in many ways, but anyone who has used a window system knows that the screen is simply not big enough for all the windows one would like! Unix has several solutions to this problem. @cindex Screens @cindex Multiple screens One solution is to attach several physical screens to a terminal. The X window system can support any number of physical screens of different types. A graphical designer might want a high resolution colour screen for drawing and a black and white screen for writing text, for instance. The disadvantage with this method is the cost of the hardware. A cheaper solution is to use a window manager such as @samp{fwvm} which creates a virtual screen of unlimited size on a single monitor. As the mouse pointer reaches the edge of the true screen, the window manager replaces the display with a new "blank screen" in which to place windows. A miniaturized image of the windows on a control panel acts as a map which makes it possible to find the applications on the virtual screen. Yet another possibility is to create virtual displays inside a single window. In other words, one can collapse several shell windows into a single @samp{xterm} window by running the program @samp{screen}. The screen command allows you to start several shells in a single window (using @key{CTRL-a CTRL-c}) and to switch between them (by typing @key{CTRL-a CTRL-n}). It is only possible to see one shell window at a time, but it is still possible to cut and paste between windows and one has a considerable saving of space. The @samp{screen} command also allows you to suspend a shell session, log out, log in again later and resume the session precisely where you left off. Here is a summary of some useful screen commands: @table @kbd @item screen Start the screen server. @item screen -r Resume a previously suspended screen session if possible. @item CTRL-a CTRL-c Start a new shell on top of the others (a fresh `screen') in the current window. @item CTRL-a CTRL-n Switch to the next `screen'. @item CTRL-a CTRL-a Switch to the last screen used. @item CTRL-a a When screen is running, @kbd{CTRL-a} is used for screen commands and cannot therefore be used in its usual shell meaning of `jump to start of line'. @kbd{CTRL-a a} replaces this. @item CTRL-a CTRL-d Detach the screen session from the current window so that it can be resumed later. It can be resumed with the @samp{screen -r} command. @item CTRL-a ? Help screen. @end table @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node Files and access, C shell, The login environment, Top @chapter Files and access To prevent all users from being able to access all files on the system, unix records information about @emph{who} creates files and also who is allowed to access them later. @cindex Access to files @cindex Permissions on files @cindex Access rights @cindex File access permission Each user has a unique @emph{username} or @emph{loginname} together with a unique @emph{user id} or @emph{uid}. The user id is a number, whereas the login name is a text string -- otherwise the two express the same information. A file belongs to user A if it is @emph{owned} by user A. User A then decides whether or not other users can read, write or execute the file by setting the @emph{protection bits} or the @emph{permission} of the file using the command @code{chmod}. @cindex groups @cindex @file{/etc/group} In addition to user identities, there are groups of users. The idea of a group is that several named users might want to be able to read and work on a file, without other users being able to access it. Every user is a member of at least one group, called the @emph{login group} and each group has both a textual name and a number (@emph{group id}). The @emph{uid} and @emph{gid} of each user is recorded in the file @code{/etc/passwd} (See chapter 6). Membership of other groups is recorded in the file @code{/etc/group} or on some systems @code{/etc/logingroup}. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Protection bits:: * chmod:: * Umask:: * s-bit and t-bit:: @end menu @node Protection bits, chmod, Files and access, Files and access @section Protection bits @cindex File protection bits @cindex Protection bits @cindex Access bits @cindex Long file listing @cindex @code{ls -l} The following output is from the command @code{ls -lag} executed on a SunOS type machine. @smallexample lrwxrwxrwx 1 root wheel 7 Jun 1 1993 bin -> usr/bin -r--r--r-- 1 root bin 103512 Jun 1 1993 boot drwxr-sr-x 2 bin staff 11264 May 11 17:00 dev drwxr-sr-x 10 bin staff 2560 Jul 8 02:06 etc drwxr-sr-x 8 root wheel 512 Jun 1 1993 export drwx------ 2 root daemon 512 Sep 26 1993 home -rwxr-xr-x 1 root wheel 249079 Jun 1 1993 kadb lrwxrwxrwx 1 root wheel 7 Jun 1 1993 lib -> usr/lib drwxr-xr-x 2 root wheel 8192 Jun 1 1993 lost+found drwxr-sr-x 2 bin staff 512 Jul 23 1992 mnt dr-xr-xr-x 1 root wheel 512 May 11 17:00 net drwxr-sr-x 2 root wheel 512 Jun 1 1993 pcfs drwxr-sr-x 2 bin staff 512 Jun 1 1993 sbin lrwxrwxrwx 1 root wheel 13 Jun 1 1993 sys->kvm/sys drwxrwxrwx 6 root wheel 732 Jul 8 19:23 tmp drwxr-xr-x 27 root wheel 1024 Jun 14 1993 usr drwxr-sr-x 10 bin staff 512 Jul 23 1992 var -rwxr-xr-x 1 root daemon 2182656 Jun 4 1993 vmunix @end smallexample The first column is a textual representation of the protection bits for each file. Column two is the number of hard links to the file (See exercises below). The third and fourth columns are the user name and group name and the remainder show the file size in bytes and the creation date. Notice that the directories @code{/bin} and @code{/sys} are symbolic links to other directories. There are sixteen protection bits for a UNIX file, but only twelve of them can be changed by users. These twelve are split into four groups of three. Each three-bit number corresponds to one @emph{octal} number. The leading four invisible bits gives information about the type of file: is the file a @emph{plain file}, a @emph{directory} or a @emph{link}. In the output from @code{ls} this is represented by a single character: @code{-}, @code{d} or @code{l}. The next three bits set the so-called @emph{s-bits} and @emph{t-bit} which are explained below. The remaining three groups of three bits set flags which indicate whether a file can be read @samp{r}, written to @samp{w} or executed @samp{x} by (i) the user who created them, (ii) the other users who are in the group the file is marked with, and (iii) any user at all. @cindex Text form of access bits @cindex Access bits, text form For example, the permission @example Type Owner Group Anyone d rwx r-x --- @end example @noindent tells us that the file is a directory, which can be read and written to by the owner, can be read by others in its group, but not by anyone else. @emph{Note about directories. It is impossible to @code{cd} to a directory unless the @code{x} bit is set. That is, directories must be `executable' in order to be accessible.} Here are some examples of the relationship between binary, octal and the textual representation of file modes. @cindex Access bits, octal form @example Binary Octal Text 001 1 x 010 2 w 100 4 r 110 6 rw- 101 5 r-x - 644 rw-r--r-- @end example @noindent It is well worth becoming familiar with the octal number representation of these permissions. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node chmod, Umask, Protection bits, Files and access @section chmod @cindex @code{chmod} command The @code{chmod} command changes the permission or @emph{mode} of a file. Only the owner of the file or the superuser can change the permission. Here are some examples of its use. Try them. @smallexample # make read/write-able for everyone chmod a+w myfile # add the 'execute' flag for directory chmod u+x mydir/ # open all files for everyone chmod 755 * # set the s-bit on my-dir's group chmod g+s mydir/ # descend recursively into directory opening all files chmod -R a+r dir @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Umask, s-bit and t-bit, chmod, Files and access @section Umask @cindex @code{umask} variable @cindex s-bit @cindex setuid bit @cindex setgid bit When a new file gets created, the operating system must decide what default protection bits to set on that file. The variable @code{umask} decides this. @code{umask} is normally set by each user in his or her @code{.cshrc} file (see next chapter). For example @smallexample umask 077 # safe umask 022 # liberal @end smallexample @noindent According the UNIX documentation, the value of @code{umask} is @samp{XOR}ed (exclusive @samp{OR}) with a value of @code{666 & umask} for plain files or @code{777 & umask} for directories in order to find out the standard protection. Actually this is not quite true: @samp{umask} only removes bits, it never sets bits which were not already set in @code{666}. For instance @cindex @code{umask} variable @example umask Permission 077 600 (plain) 077 700 (dir) 022 644 (plain) 022 755 (dir) @end example The correct rule for computing permissions is not XOR but @samp{NOT AND}. @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Making programs executable:: * chown and chgrp:: * Making a group:: @end menu @node Making programs executable, chown and chgrp, Umask, Umask @subsection Making programs executable @cindex Masking programs executable @cindex Executable, making programs A unix program is normally executed by typing its pathname. If the @code{x} execute bit is not set on the file, this will generate a `Permission denied' error. This protects the system from interpreting nonsense files as programs. To make a program executable for someone, you must therefore ensure that they can execute the file, using a command like @example chmod u+x @var{filename} @end example @noindent This command would set execute permissions for the owner of the file; @example chmod ug+x @var{filename} @end example @noindent would set execute permissions for the owner and for any users in the same group as the file. Note that script programs must also be readable in order to be executable, since the shell has the interpret them by reading. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node chown and chgrp, Making a group, Making programs executable, Umask @subsection chown and chgrp @cindex @code{chown} command @cindex @code{chgrp} command These two commands change the ownership and the group ownership of a file. Only the superuser can change the ownership of a file on most systems. This is to prevent users from being able to defeat quota mechanisms. (On some systems, which do not implement quotas, ordinary users can give a file away to another user but not get it back again.) The same applies to group ownership. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Making a group, , chown and chgrp, Umask @subsection Making a group Normally users other than root cannot define their own groups. This is a weakness in Unix from older times which no one seems to be in a hurry to change. At Oslo College, Computer Science, we use a local solution whereby users can edit a file to create their own groups. This file is called @file{/iu/nexus/local/iu/etc/iu-group}. The format of the group file is: @smallexample @var{group-name}::@var{group-number}:@var{comma-separated-list-of-users} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node s-bit and t-bit, , Umask, Files and access @section s-bit and t-bit (sticky bit) @cindex Sticky bit @cindex s-bit @cindex t-bit The @code{s} and @code{t} bits have special uses. They are described as follows. @example Octal Text Name 4000 chmod u+s Setuid bit 2000 chmod g+s Setgid bit 1000 chmod +t Sticky bit @end example @noindent The effect of these bits differs for plain files and directories and differ between different versions of UNIX. You should check the manual page @code{man sticky} to find out about your system! The following is common behaviour. For executable files, the setuid bit tells UNIX that @emph{regardless of who runs the program} it should be executed with the permissions and rights of owner of the file. This is often used to allow normal users limited access to @code{root} privileges. A @emph{setuid-root} program is executed as @code{root} for any user. The setgid bit sets the group execution rights of the program in a similar way. In BSD unix, if the setgid bit is set on a directory then any new files created in that directory assume the group ownership of the parent directory and not the logingroup of the user who created the file. This is standard policy under system 5. A directory for which the sticky bit is set restrict the deletion of files within it. A file or directory inside a directory with the t-bit set can only be deleted or renamed by its owner or the superuser. This is useful for directories like the mail spool area and @code{/tmp} which must be writable to everyone, but should not allow a user to delete another user's files. (Ultrix) If an executable file is marked with a sticky bit, it is held in the memory or system swap area. It does not have to be fetched from disk each time it is executed. This saves time for frequently used programs like @code{ls}. (Solaris 1) If a non-executable file is marked with the sticky bit, it will @emph{not} be held in the disk page cache -- that is, it is never copied from the disk and held in RAM but is written to directly. This is used to prevent certain files from using up valuable memory. On some systems (e.g. ULTRIX), only the superuser can set the sticky bit. On others (e.g. SunOS) any user can create a sticky directory. @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node C shell, Bourne shell, Files and access, Top @chapter C shell The C shell is the command interpreter which you use to run programs and utilities. It contains a simple programming language for writing tailor-made commands, and allows you to join together unix commands with pipes. It is a configurable environment, and once you know it well, it is the most efficient way of working with unix. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * .cshrc and .login:: * Defining variables with set and setenv:: * Arrays:: * Pipes and redirection:: * tee and script:: * Command history:: * Command/filename completion:: * Single and double quotes:: * Job control:: * Scripts with argumemts:: * subshells:: * Tests and conditions:: * Loops in csh:: * Input from the user:: * Extracting parts of a pathname:: * Arithmetic:: * Examples:: @end menu @node .cshrc and .login, Defining variables with set and setenv, C shell, C shell @section .cshrc and .login files @cindex @file{.cshrc} file @cindex @file{.login} file @cindex Login evironment @cindex C shell setup files @cindex Setting up the C shell Most users run the C-shell @file{/bin/csh} as their login environment, or these days, preferably the @file{tcsh} which is an improved version of csh. When a user logs in to a UNIX system the C-shell starts by reading some files which configure the environment by defining variables like @code{path}. @itemize @bullet @item The file @file{.cshrc} is searched for in your home directory. i.e. @file{~/.cshrc}. If it is found, its contents are interpreted by the C-shell as C-shell instructions, before giving you the command prompt@footnote{Under HPUX, two other files are also read by the C-shell. These are called @file{/etc/csh.login} and @file{/etc/src.csh}, enabling some standard set-up to be configured globally. GNU/Linux has a similar system. On solaris systems @file{/etc/.login} is read.}. @item If and only if this is the @emph{login shell} (not a sub-shell that you have started after login) then the file @file{~/.login} is searched for and executed. @end itemize @noindent With the advent of the X11 windowing system, this has changed slightly. Since the window system takes over the entire login procedure, users never get to run `login shells', since the login shell is used up by the X11 system. On an X-terminal or host running X the @file{.login} file normally has no effect. With some thought, the @file{.login} file can be eliminated entirely, and we can put everything into the @code{.cshrc} file. Here is a very simple example @file{.cshrc} file. @smallexample # # .cshrc - read in by every csh that starts. # # Set the default file creation mask umask 077 # Set the path set path=( /usr/local/bin /usr/bin/X11 /usr/ucb /bin /usr/bin . ) # Exit here if the shell is not interactive if ( $?prompt == 0 ) exit # Set some variables set noclobber notify filec nobeep set history=100 @cindex Setting the prompt @cindex Prompt, redefining set prompt="`hostname`%" set prompt2 = "%m %h>" # tcsh, prompt for foreach and while setenv PRINTER myprinter setenv LD_LIBRARY_PATH /usr/lib:/usr/local/lib:/usr/openwin/lib # Aliases are shortcuts to unix commands alias passwd yppasswd alias dir 'ls -lg \!* | more' alias sys 'ps aux | more' alias h history @end smallexample It is possible to make a much more complicated .cshrc file than this. The advent of distributed computing and NFS (Network file system) means that you might log into many different machines running different versions of unix. The command path would have to be set differently for each type of machine. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Defining variables with set and setenv, Arrays, .cshrc and .login, C shell @section Defining variables with set, setenv We have already seen in the examples above how to define variables in C-shell. Let's formalize this. To define a local variable -- that is, one which will not get passed on to programs and sub-shells running under the current shell, we write @smallexample set local = "some string" set myname = "`whoami`" @end smallexample @noindent These variables are then referred to by using the dollar @samp{$} symbol. i.e. The value of the variable @samp{local} is @samp{$local}. @smallexample echo $local $myname @end smallexample @noindent Global variables, that is variables which all sub-shells inherit from the current shell are defined using @samp{setenv} @smallexample setenv GLOBAL "Some other string" setenv MYNAME "`who am i`" @end smallexample @noindent Their values are also referred to using the @samp{$} symbol. Notice that @code{set} uses an @samp{=} sign while @samp{setenv} does not. @cindex @code{set} command @cindex Local variables in csh @cindex @code{setenv} command @cindex Global variables in csh Variables can be also created without a value. The shell uses this method to switch on and off certain features, using variables like @samp{noclobber} and @samp{noglob}. For instance @smallexample nexus% set flag nexus% if ($?flag) echo 'Flag is set!' Flag is set! nexus% unset flag nexus% if ( $?flag ) echo 'Flag is set!' nexus% @end smallexample @noindent The operator @samp{$?@var{variable}} is @samp{true} if @var{variable} exists and @samp{false} if it does not. It does not matter whether the variable holds any information. @vindex set @vindex unset @cindex @code{unset} command The commands @samp{unset} and @samp{unsetenv} can be used to undefine or delete variables when you don't want them anymore. @cindex Undefining variables @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Arrays, Pipes and redirection, Defining variables with set and setenv, C shell @section Arrays @cindex Arrays in csh @cindex () operators to make array in csh @vindex () in csh A useful facility in the C-shell is the ability to make arrays out of strings and other variables. The round parentheses @samp{(..)} do this. For example, look at the following commands. @smallexample nexus% set array = ( a b c d ) nexus% echo $array[1] a nexus% echo $array[2] b nexus% echo $array[$#array] d nexus% set noarray = ( "a b c d" ) nexus% echo $noarray[1] a b c d nexus% echo $noarray[$#noarray] a b c d @end smallexample @noindent The first command defines an array containing the elements @samp{a b c d}. The elements of the array are referred to using square brackets @samp{[..]} and the first element is @samp{$array[1]}. The last element is @samp{$array[4]}. @emph{NOTE: this is not the same as in C or C++ where the first element of the array is the zeroth element!} The special operator @samp{$#} returns the number of elements in an array. This gives us a simple way of finding the end of the array. For example @smallexample nexus% echo $#path 23 nexus% echo "The last element in path is $path[$#path]" The last element in path is . @end smallexample @noindent To find the next last element we need to be able to do arithmetic. We'll come back to this later. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Pipes and redirection, tee and script, Arrays, C shell @section Pipes and redirection in csh @cindex Pipes @cindex @samp{|} symbol @vindex | @vindex > @vindex < @vindex >> @vindex << @vindex & @cindex Redirection of stdio @cindex Standard I/O, redirection The symbols @smallexample < > >> << | & @end smallexample @noindent have a special meaning in the shell. By default, most commands take their input from the file @samp{stdin} (the keyboard) and write their output to the file @samp{stdout} and their error messages to the file @samp{stderr} (normally, both of these output files are defined to be the current terminal device @samp{/dev/tty}, or @samp{/dev/console}). @samp{stdin}, @samp{stdout} and @samp{stderr}, known collectively as @samp{stdio}, can be redefined or @emph{redirected} so that information is taken from or sent to a different file. The output direction can be changed with the symbol @samp{>}. For example, @cindex Output, sending to a file @smallexample echo testing > myfile @end smallexample @noindent produces a file called @file{myfile} which contains the string `testing'. The single @samp{>} (greater than) sign always creates a new file, whereas the double @samp{>>} appends to the end of a file, if it already exists. So the first of the commands @smallexample echo blah blah >> myfile echo Newfile > myfile @end smallexample @noindent adds a second line to @samp{myfile} after `testing', whereas the second command writes over @file{myfile} and ends up with just one line `Newfile'. Now suppose we mistype a command @smallexample ehco test > myfile @end smallexample @noindent The command @samp{ehco} does not exist and so the error message @samp{ehco: Command not found} appears on the terminal. This error message was sent to @emph{stderr} -- so even though we redirected output to a file, the error message appeared on the screen to tell us that an error occurred. Even this can be changed. @samp{stderr} can also be redirected by adding an ampersand @samp{&} character to the @samp{>} symbol. The command @smallexample ehco test >& myfile @end smallexample @noindent results in the file @file{myfile} being created, containing the error message `ehco: Command not found'. The input direction can be changed using the @samp{<} symbol for example @smallexample /bin/mail mark < message @end smallexample @noindent would send the file @samp{message} to the user @samp{mark} by electronic mail. The mail program takes its input from the file instead of waiting for keyboard input. @cindex @code{noclobber} variable @cindex Protecting files from overwrite with @samp{>} @cindex @code{noclobber} overwrite protection There are some refinements to the redirection symbols. First of all, let us introduce the C-shell variable @samp{noclobber}. If this variable is set with a command like @smallexample set noclobber @end smallexample @noindent then files will not be overwritten by the @samp{>} command. If one tries to redirect output to an existing file, the following happens. @smallexample unix% set noclobber unix% touch blah # create an empty file blah unix% echo test > blah blah: File exists. @end smallexample If you are nervous about overwriting files, then you can set @samp{noclobber} in your @file{.cshrc} file. @samp{noclobber} can be overridden using the pling @samp{!} symbol. So @smallexample unix% set noclobber unix% touch blah # create an empty file blah unix% echo test >! blah @end smallexample @noindent writes over the file @samp{blah} even though @samp{noclobber} is set. Here are some other combinations of redirection symbols @cindex Appending to a file with @samp{>>} @table @samp @item >> Append, including @samp{stderr} @item >>! Append, ignoring @samp{noclobber} @item >>&! Append @samp{stdout}, @samp{stderr}, ignore @samp{noclobber} @item << See below. @end table @noindent The last of these commands reads from the standard input until it finds a line which contains a word. It then feeds all of this input into the program concerned. For example, @smallexample nexus% mail mark < Hello mark nexus 2> Nothing much to say... nexus 2> so bye nexus 2> nexus 2> quit Sending mail... Mail sent! @end smallexample @noindent The mail message contains all the lines up to, but not including `marker'. This method can also be used to print text verbatim from a file without using multiple echo commands. Inside a script one may write: @cindex Formatting text in a file @cindex Printing multiple lines @smallexample cat << "marker"; MENU 1) choice 1 2) choice 2 ... marker @end smallexample The @code{cat} command writes directly to @code{stdout} and the input is redirected and taken directly from the script file. @vindex << @cindex Input over many lines @cindex Pipe @cindex @samp{|} symbol A very useful construction is the `pipe' facility. Using the @samp{|} symbol one can feed the @samp{stdout} of one program straight into the @samp{stdin} of another program. Similarly with @samp{|&} both @samp{stdout} and @samp{stderr} can be piped into the input of another program. This is very convenient. For instance, look up the following commands in the manual and try them. @cindex Piping to more to prevent scrolling @smallexample ps aux | more echo 'Keep on sharpenin them there knives!' | mail henry vmstat 1 | head ls -l /etc | tail @end smallexample @noindent Note that when piping both standard input and standard error to another program, the two files @emph{do not mix synchronously}. Often @samp{stderr} appears first. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node tee and script, Command history, Pipes and redirection, C shell @section @samp{tee} and @samp{script} @cindex @samp{tee} command @cindex Splitting output to several files @cindex Copy of output to file @cindex Output to file Occasionally you might want to have a copy of what you see on your terminal sent to a file. @samp{tee} and @samp{script} do this. For instance, @smallexample find / -type l -print | tee myfile @end smallexample @noindent sends a copy of the output of @samp{find} to the file `myfile'. @samp{tee} can split the output into as many files as you want: @smallexample @var{command} | tee @var{file1} @var{file2} .... @end smallexample @noindent You can also choose to record the output an entire shell session using the @samp{script} command. @smallexample nexus% script mysession Script started, file is mysession nexus% echo Big brother is scripting you Big brother is scripting you nexus% exit exit Script done, file is mysession @end smallexample @noindent The file `mysession' is a text file which contains a transcript of the session. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Command history, Command/filename completion, tee and script, C shell @section Command history @cindex Command history @cindex Up arrow The history feature in C-shell means that you do not have to type commands over and over again. In the @samp{tcsh} version of the C shell, and the @samp{bash} version of the Bourne shell, you can use the @key{UP ARROW} key to browse back through the list of commands you have typed previously. In the normal C-shell (@samp{csh}) there are three main commands. @table @samp @item !! Execute the last command again. @item !-3 Execute the third last command again. @item !4 Execute command number 4. @end table The first of these simply repeats the last command. The second counts backwards from the last command to three commands-ago. The final command gives an absolute number. The absolute command number can be seen by typing @samp{history}. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Command/filename completion, Single and double quotes, Command history, C shell @section Command/filename completion @cindex Filename completion @cindex Command completion @cindex @key{TAB} completion key In the @samp{tcsh} extension of the C-shell, you can save hours worth of typing errors by using the completion mechanism. This feature is based on the @key{TAB} key. The idea is that if you type half a filename and press @key{TAB}, the shell will try to guess the remainder of the filename. It does this by looking at the files which match what you have already typed and trying to fill in the rest. If there are several files which match, the shell sounds the "bell" or beeps. You can then type @key{CTRL-D} to obtain a list of the possible alternatives. Here is an example: suppose you have just a single file in the current directory called @file{very_long_filename}, typing @smallexample more @key{TAB} @end smallexample @noindent results in the following appearing on the command line @smallexample more very_long_filename @end smallexample @noindent The shell was able to identify a unique file. Now suppose that you have two files called @file{very_long_filename} and @file{very_big_filename}, typing @smallexample more @key{TAB} @end smallexample @noindent results in the following appearing on the command line @smallexample more very_ @end smallexample @noindent and the shell beeps, indicating that the choice was not unique and a decision is required. Next, you type @key{CTRL-D} to see which files you ahve to choose from and the shell lists them and returns you to the command line, exactly where you were. You now choose @file{very_long_filename} by typing @samp{l}. This is enough to uniquely identify the file. Pressing the @key{TAB} key again results in @smallexample more very_long_filename @end smallexample @noindent on the screen. As long as you have written enough to select a file uniquely, the shell will be able to complete the name for you. Completion also works on shell commands, but it is a little slower since the shell must serach through all the directories in the command path to complete commands. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Single and double quotes, Job control, Command/filename completion, C shell @section Single and double quotes @cindex Single and double quotes @cindex @key{'} and @key{"} @vindex ' @vindex " @vindex ` @cindex Backwards quotes @cindex @key{`} symbol and embedded shells @cindex Embedded shell Two kinds of quotes can be used in shell apart from the backward quotes we mentioned above. The essential difference between them is that certain shell commands work inside @emph{double} quotes but not inside single quotes. For example @smallexample nexus% echo /etc/rc.* /etc/rc.boot /etc/rc.ip /etc/rc.local nexus% echo "/etc/rc.*" /etc/rc.* nexus% echo "`who am i` -- my name is $user ???" nexus!mark ttyp7 Jul 13 10:16 -- my name is mark ??? nexus% echo '`who am i` -- my name is $user ???' `who am i` -- my name is $user ??? @end smallexample We see that the single quotes prevent @emph{variable substitution} and @emph{sub-shells}. Wildcards do not work inside either single or double quotes. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Job control, Scripts with argumemts, Single and double quotes, C shell @section Job control, break key, @samp{fg}, @samp{bg} @cindex Job control @cindex Foreground process @cindex Background process So far we haven't mentioned UNIX's ability to multitask. In the Bourne shell (@samp{sh}) there are no facilities for controlling several user processes @footnote{Newer versions of the Bourne shell, like the Korn shell and Bourne-again-shell do have these facilities.}. C-shell provides some commands for starting and stopping processes. These originate from the days before windows and X11, so some of them may seem a little old-fashioned. They are still very useful nonetheless. Let's begin by looking at the commands which are true for any shell. Most programs are run in the @emph{foreground} or @emph{interactively}. That means that they are connected to the standard input and send their output to the standard output. A program can be made to run in the background, if it does not need to use the standard I/O. For example, a program which generates output and sends it to a file could run in the background. In a window environment, programs which create their own windows can also be started as background processes, leaving standard I/O in the shell free. @emph{Background processes run independently of what you are doing in the foreground.} @c .............................. @c SUBSECTION @c .............................. @menu * Processes and signals:: * Child processes and zombies:: * Builtins:: @end menu @node Processes and signals, Child processes and zombies, Job control, Job control @subsection Unix Processes and BSD signals @cindex @samp{find} command @cindex Processes @cindex Background process A background process is started using the special charcter @samp{&} at the end of the command line. @smallexample find / -name '*lib*' -print >& output & @end smallexample @noindent The final @samp{&} on the end of this line means that the job will be run in the background. Note that this is not confused with the redirection operator @samp{>&} since it must be the last character on the line. The command above looks for any files in the system containing the string `lib' and writes the list of files to a file called `output'. This might be a useful way of searching for missing libraries which you want to include in your environment variable @samp{LD_LIBRARY_PATH}. Searching the enire disk from the root directory @samp{/} could take a long time, so it pays to run this in the background. If we want to see what processes are running, we can use the @samp{ps} command. @samp{ps} without any arguments lists all of your processes, i.e. all processes owned by the user name you logged in with. @samp{ps} takes many options, for instance @samp{ps auxg} will list all processes in gruesome detail. (The "g" is for group, not gruesome!) @samp{ps} reads the kernel's process tables directly. Processes can be stopped and started, or killed one and for all. The @samp{kill} command does this. There are, in fact, two versions of the @samp{kill} command. One of them is built into the C-shell and the other is not. If you use the C-shell then you will never care about the difference. We shall nonetheless mention the special features of the C-shell built-ins below. The kill command takes a number called a @emph{signal} as an argument and another number called the @emph{process identifier} or @emph{PID} for short. Kill send signals to processes. Some of these are fatal and some are for information only. The two commands @smallexample kill -15 127 kill 127 @end smallexample @noindent are identical. They both send signal 15 to PID 127. This is the normal @emph{termination} signal and it is often enough to stop any process from running. Programs can choose to ignore certain signals by trapping signals with a special handler. One signal they cannot ignore is signal 9. @smallexample kill -9 127 @end smallexample @noindent is a sure way of killing PID 127. Even though the process dies, it may not be removed from the kernel's process table if it has a parent (see next section). Here is the complete list of unix signals which the kernel send to processes in different circumstances. @smallexample 1 "SIGHUP", /* hangup */ 2 "SIGINT", /* interrupt */ 3 "SIGQUIT", /* quit */ 4 "SIGILL", /* illegal instruction (not reset when caught) */ 5 "SIGTRAP", /* trace trap (not reset when caught) */ 6 "SIGIOT/SIGABRT", /* IOT instruction */ 7 "SIGEMT", /* EMT instruction */ 8 "SIGFPE", /* floating point exception */ 9 "SIGKILL", /* kill (cannot be caught or ignored) */ 10 "SIGBUS", /* bus error */ 11 "SIGSEGV", /* segmentation violation */ 12 "SIGSYS", /* bad argument to system call */ 13 "SIGPIPE", /* write on a pipe with no one to read it */ 14 "SIGALRM", /* alarm clock */ 15 "SIGTERM", /* software termination signal from kill */ 16 "SIGURG", /* urgent condition on IO channel */ 17 "SIGSTOP", /* sendable stop signal not from tty */ 18 "SIGTSTP", /* stop signal from tty */ 19 "SIGCONT", /* continue a stopped process */ 20 "SIGCHLD/SIGCLD", /* to parent on child stop or exit */ 21 "SIGTTIN", /* to readers pgrp upon background tty read */ 22 "SIGTTOU", /* like TTIN for output if (tp->t_local<OSTOP) */ 23 "SIGIO/SIGPOLL", /* input/output possible signal */ 24 "SIGXCPU", /* exceeded CPU time limit */ 25 "SIGXFSZ", /* exceeded file size limit */ 26 "SIGVTALRM", /* virtual time alarm */ 27 "SIGPROF", /* profiling time alarm */ 28 "SIGWINCH", /* window changed */ 29 "SIGLOST", /* resource lost (eg, record-lock lost) */ 30 "SIGUSR1", /* user defined signal 1 */ 31 "SIGUSR2" @end smallexample We have already mentioned 15 and 9 which are the main signals for users. Signal 1, or @samp{HUP} can be sent to certain programs by the superuser. For instance @smallexample kill -1 kill -HUP @end smallexample @noindent which forces @samp{inetd} to reread its configuration file. Sometimes it is useful to @emph{suspend} a process temporarily and then @emph{restart} it later. @smallexample kill -18 # suspend process kill -19 # resume process @end smallexample @c .............................. @c SUBSECTION @c .............................. @node Child processes and zombies, Builtins, Processes and signals, Job control @subsection Child Processes and zombies When you start a process from a shell, regardless of whether it is a background process or a foreground process, the new process becomes a @emph{child} of the original shell. Remember that the shell is just a unix process itself. Moreover, if one of the children starts a new process then it will be a child of the child (a grandchild?)! Processes therefore form @emph{hierachies}. Several children can have a common @emph{parent}. If we kill a parent, then (unless the child has detached itself from the parent) all of its children die too. If a child dies, the parent is not affected. Sometimes when a child is killed, it does not die but becomes "defunct" or a zombie process. This means that the child has a parent which is @emph{waiting} for it to finish. If the parent has not yet been informed that the child has died, for example because it has been suspended itself, then the dead child is not removed from the kernel's process table. When the parent wakes up and receives the message that the child has terminated, the process entry for the dead child can be removed. @c .............................. @c SUBSECTION @c .............................. @node Builtins, , Child processes and zombies, Job control @subsection C-shell builtins: @samp{jobs}, @samp{kill}, @samp{fg},@samp{bg}, break key @cindex Starting shell jobs @cindex @samp{fg} command @cindex @samp{bg} command @cindex Job numbers in csh Now let's look at some commands which are built into the C-shell for starting and stopping processes. C-shell refers to user programs as `jobs' rather than processes -- but there is no real difference. The added bonus of the C-shell is that each shell has a @emph{job number} in addition to its PID. The job numbers are simpler and are private for the shell, whereas the PIDs are assigned by the kernel and are often very large numbers which are difficult to to remember. When a command is executed in the shell, it is assigned a job number. If you never run any background jobs then there is only ever one job number: 1, since every job exits before the next one starts. However, if you run background tasks, then you can have several jobs "active" at any time. Moreover, by @emph{suspending} jobs, C-shell allows you to have several interactive programs running on the same terminal -- the @samp{fg} and @samp{bg} commands allow you to move commands from the background to the foreground and vice-versa. Take a look at the following shell session. @smallexample nexus% emacs myfile & [1] 4990 nexus% ( other commands ... , edit myfile and close emacs ) [1] Exit 70 emacs myfile @end smallexample @noindent When a background job is done, the shell prints a message at a suitable moment between prompts. @smallexample [1] Done emacs myfile @end smallexample @noindent This tells you that job number 1 finished normally. If the job exits abnormally then the word `Done' may be replaced by some other message. For instance, if you kill the job, it will say @cindex @samp{kill} command @smallexample unix% kill %12 [12] Terminated textedit file @end smallexample @noindent You can list the jobs you have running using the @samp{jobs} command. The output looks something like @smallexample [1] + Running textedit c.tex [3] Running textedit glossary.tex [4] Running textedit net.tex [5] Running textedit overview.tex [6] Running textedit perl.tex [7] Running textedit shell.tex [8] Running textedit sysadm.tex [9] Running textedit unix.tex [10] Running textedit x11.tex [11] - Running shelltool [15] Suspended emacs myfile @end smallexample @noindent @cindex Suspending a job @cindex Moving a job to the background @cindex Job, moving to background @cindex Process. moving to background To suspend a program which you are running in the foreground you can type @key{CTRL-z} (this is like sending a @samp{kill -18} signal from the keyboard). @footnote{This does not seem to work in solaris?!} You can suspend any number of programs and then restart them one at a time using @samp{fg} and @samp{bg}. If you want job 5 to be restarted in the foreground, you would type @smallexample fg %5 @end smallexample @noindent When you have had enough of job 5, you can type CTRL-z to suspend it and then type @noindent @smallexample fg %6 @end smallexample @noindent to activate job 6. Provided a job does not want to send output to @samp{stdout}, you can restart any job in the background, using a command like. @smallexample bg %4 @end smallexample @emph{This method of working was useful before windows were available.} @emph{Using @samp{fg} and @samp{bg}, you can edit several files} @emph{or work on several programs without have to quit to} @emph{move from one to another.} See also some related commands for batch processing @samp{at}, @samp{batch} and @samp{atq}, @samp{cron}. @cindex Break key @cindex Panic button NOTE: @key{CTRL-c} sends a @samp{kill -2} signal, which send a standard interrupt message to a program. This is always a safe way to interrupt a shell command. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Scripts with argumemts, subshells, Job control, C shell @section Scripts with arguments @cindex Script, making @cindex Making a script @cindex Writing a script One of the useful features of the shell is that you can use the normal unix commands to make programs called @emph{scripts}. To make a script, you just create a file containing shell commands you want to execute and make sure that the first line of the file looks like the following example. @smallexample #!/bin/csh -f # # A simple script: check for user's mail # # set path = ( /bin /usr/ucb ) # Set the local path cd /var/spool/mail # Change dir foreach uid ( * ) echo "$uid has mail in the intray! " # space prevents an error! end @end smallexample @cindex @samp{#!program} sequence The sequence @samp{#!/bin/csh} means that the following commands are to be fed into @file{/bin/csh}. The two symbols @samp{#!} must be the very first two characters in the file. The @samp{-f} option means that your @samp{.cshrc} file is not read by the shell when it starts up. The file containing this script must be executable (see @samp{chmod}) and must be in the current path, like all other programs. @cindex Command line arguments @cindex Argument vector in csh @cindex @samp{argv} @cindex Arguments, command line Like C programs, C-shell scripts can accept command line arguments. Suppose you want to make a program to say hello to some other users who are logged onto the system. @smallexample say-hello mark sarah mel @end smallexample @noindent To do this you need to know the names that were typed on the command line. These names are copied into an array in the C-shell called the @emph{argument vector}, or @samp{argv}. To read these arguments, you just treat @samp{argv} as an array. @cindex @code{foreach} example @cindex @code{write} example @smallexample #!/bin/csh -f # # Say hello # foreach name ( $argv ) echo Saying hello to $name echo "Hello from $user! " | write $name end @end smallexample The elements of the array can be referred to as @samp{argv[1]}..@samp{argv[$#argv]} as usual. They can also be referred to as @samp{$1}..@samp{$3} upto the last acceptable number. This makes C-shell compatible with the Bourne shell as far as arguments are concerned. One extra flourish in this method is that you can also refer to the name of the program itself as @samp{$0}. For example, @smallexample #!/bin/csh -f echo This is program $0 running for $user @end smallexample @noindent @samp{$argv} represents all the arguments. You can also use @samp{$*} from the Bourne shell. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node subshells, Tests and conditions, Scripts with argumemts, C shell @section Sub-shells () @cindex Subshells and () @cindex () and subshells The C-shell does not allow you to define subroutines or functions, but you can create a local shell, with its own private variables by enclosing commands in parentheses. @smallexample #!/bin/csh cd /etc ( cd /usr/bin; ls * ) > myfile pwd @end smallexample This program changes the working directory to /etc and then executes a subshell which @emph{inside the brackets} changes directory to /usr/bin and lists the files there. The output of this private shell are sent to a file `myfile'. At the end we print out the current working directory just to show that the @samp{cd} command in brackets had no effect on the main program. Normally both parentheses must be on the same line. If a subshell command line gets too long, so that the brackets are not on the same line, you have to use backslash characters to continue the lines, @smallexample ( command \ command \ command \ ) @cindex Continuing long lines @cindex Long lines, continuing @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Tests and conditions, Loops in csh, subshells, C shell @section Tests and conditions @cindex Tests and conditions in csh @cindex @code{if..then..else} in csh No programming language would be complete without tests and loops. C-shell has two kinds of decision structure: the @samp{if..then..else} and the @samp{switch} structure. These are closely related to their C counterparts. The syntax of these is @smallexample if (@var{condition}) @var{command} if (@var{condition}) then command command.. else command command.. endif @cindex @code{switch..case} in csh switch (@var{string}) case @var{one}: commands breaksw case @var{two}: commands breaksw ... endsw @end smallexample In the latter case, no commands should appear on the same line as a `case' statement, or they will be ignored. Also, if the @samp{breaksw} commands are omitted, then control flows through all the commands for case 2, case 3 etc, exactly as it does in the C programming language. We shall consider some examples of these statements in a moment, but first it is worth listing some important tests which can be used in @samp{if} questions to find out information about files. @cindex Testing files @vindex -r @var{file} @vindex -w @var{file} @vindex -e @var{file} @vindex -z @var{file} @vindex -f @var{file} @vindex -d @var{file} @table @samp @item -r @var{file} True if the file exists and is readable @item -w @var{file} True if the file exists and is writable @item -x @var{file} True if the file exists and is executable @item -e @var{file} True if the file simply exists @item -z @var{file} True if the file exists and is empty @item -f @var{file} True if the file is a plain file @item -d @var{file} True if the file is a directory @end table We shall also have need of the following comparision operators. @cindex Comparison operators in csh @cindex Operators in csh @vindex @samp{==} @vindex @samp{!=} @vindex @samp{>} @vindex @samp{<} @vindex @samp{>=} @vindex @samp{<=} @vindex @samp{=~} @vindex @samp{!~} @table @samp @item == is equal to (string comparison) @item != is not equal to @item > is greater than @item < is less than @item >= is greater than or equal to @item <= is less than or equal to @item =~ matches a wildcard @item !~ does not match a wildcard @end table The simplest way to learn about these statements is to use them, so we shall now look at some examples. @smallexample #!/bin/csh -f # # Safe copy from to # # if ($#argv != 2) then echo "Syntax: copy " exit 0 endif if ( -f $argv[2] ) then echo "File exists. Copy anyway?" switch ( $< ) # Get a line from user case y: breaksw default: echo "Doing nothing!" exit 0 endsw endif echo -n "Copying $argv[1] to $argv[2]..." cp $argv[1] $argv[2] echo done endif @end smallexample This script tries to copy a file from one location to another. If the user does not type exactly two arguments, the script quits with a message about the correct syntax. Otherwise it tests to see whether a plain file has the same name as the file the user wanted to copy to. If such a file exists, it asks the user if he/she wants to continue before proceding to copy. @c --------------------- @c Subsection @c --------------------- @menu * Switch example:: @end menu @node Switch example, , Tests and conditions, Tests and conditions @subsection Switch example: configure script @cindex @samp{uname} command @cindex System identity and @samp{uname} Here is another example which compiles a software package. This is a problem we shall return to later @xref{Make}. The problem this script tries to address is the following. There are many different versions of UNIX and they are not exactly compatible with one another. The program this file compiles has to work on any kind of UNIX, so it tries first to determine what kind of UNIX system the script is being run on by calling @samp{uname}. Then it defines a variable @samp{MAKE} which contains the path to the `make' program which will build @emph{software}. The make program reads a file called `Makefile' which contains instructions for compiling the program, but this file needs to know the type of UNIX, so the script first copies a file `Makefile.src' using @samp{sed} replace a dummy string with the real name of the UNIX. Then it calls make and sets the correct permission on the file using @samp{chmod}. @cindex @samp{sed}, search and replace @cindex Swapping text strings @cindex Make software script @cindex Build software script @cindex Compiler script @smallexample #!/bin/csh -f ################################################# # # # CONFIGURE Makefile AND BUILD software # # ################################################# set NAME = ( `uname -r -s` ) switch ($NAME[1]) case SunOS*: switch ($NAME[2]) case 4*: setenv TYPE SUN4 setenv MAKE /bin/make breaksw case 5*: setenv TYPE SOLARIS setenv MAKE /usr/ccs/bin/make breaksw endsw breaksw case ULTRIX*: setenv TYPE ULTRIX setenv MAKE /bin/make breaksw case HP-UX*: setenv TYPE HPuUX setenv MAKE /bin/make breaksw case AIX*: setenv TYPE AIX setenv MAKE /bin/make breaksw case OSF*: setenv TYPE OSF setenv MAKE /bin/make breaksw case IRIX*: setenv TYPE IRIX setenv MAKE /bin/make breaksw default: echo Unknown architecture $NAME[1] endsw # Generate Makefile from source file sed s/HOSTTYPE/$TYPE/ Makefile.src > Makefile echo "Making software. Type CTRL-C to abort and edit Makefile" $MAKE software # call make to build program chmod 755 software # set correct protection @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Loops in csh, Input from the user, Tests and conditions, C shell @section Loops in csh @cindex Loops in csh @cindex @code{foreach} @cindex @code{repeat} @cindex @code{while} @vindex foreach @vindex repeat @vindex while The C-shell has three loop structures: @samp{repeat}, @samp{while} and @samp{foreach}. We have already seen some examples of the @samp{foreach} loop. The structure of these loops is as follows @smallexample repeat @var{number-of-times} @var{command} while ( @var{test expression} ) @var{commands} end foreach @var{control-variable} ( @var{list-or-array} ) @var{commands} end @end smallexample @noindent @cindex @samp{continue} in csh @vindex continue The commands @samp{break} and @samp{continue} can be used to break out of the loops at any time. Here are some examples. @smallexample repeat 2 echo "Yo!" | write mark @end smallexample @noindent This sends the message "Yo!" to mark's terminal twice. @smallexample repeat 5 echo `echo "Shutdown time! Log out now" | wall ; sleep 30` ; halt @end smallexample @noindent This example repeats the command `echo Shutdown time...' five times at 30 second intervals, before shutting down the system. Only the superuser can run this command! Note the strange construction with `echo echo'. This is to force the repeat command to take two shell commands as an argument. (Try to explain why this works for yourself.) @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Input from the user, Extracting parts of a pathname, Loops in csh, C shell @section Input from the user @cindex Input in csh @cindex @samp{$<} operator @vindex breaksw @cindex @samp{breaksw} @smallexample # Test a user response echo "Answer y/n (yes or no)" set valid = false while ( $valid == false ) switch ( $< ) case y: echo "You answered yes" set valid = true breaksw case n: echo "You answered no" set valid = true breaksw default: echo "Invalid reponse, try again" breaksw endsw end @end smallexample @noindent Notice that it would have been simpler to replace the two lines @smallexample set valid = true breaksw @end smallexample @noindent by a single line @samp{break}. @samp{breaksw} jumps out of the switch construction, after which the @samp{while} test fails. @samp{break} jumps out of the entire while loop. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Extracting parts of a pathname, Arithmetic, Input from the user, C shell @section Extracting parts of a pathname @cindex Parts of a filename @vindex :h @vindex :t @vindex :e @vindex :r @cindex Extracting filename components A path name consists of a number of different parts: @itemize @bullet @item The path to the directory where a file is held. @item The name of the file itself. @item The file extension (after a dot). @end itemize By using one of the following modifiers, we can extract these different elements. @table @samp @item :h The path to the file @item :t The filename itself @item :e The file extension @item :r The complete file-path minus the file extension @end table Here are some examples and the results: @smallexample set f = ~/progs/c++/test.C echo $f:h /home/mark/progs/c++ echo $f:t test.C echo $f:e C echo $f:r /home/mark/progs/c++/test @end smallexample @noindent @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Arithmetic, Examples, Extracting parts of a pathname, C shell @section Arithmetic Before using these features in a real script, we need one more possibility: numerical addition, subtraction and multiplication etc. @cindex Arithemtic operations in csh @vindex @samp{+} @vindex @samp{+=} @vindex @samp{++} @vindex @samp{-} @vindex @samp{-=} @vindex @samp{--} @vindex @samp{==} equal to (compare) @vindex @samp{>} greater than @vindex @samp{<} less than @vindex @samp{||} logical OR @vindex @samp{|} OR @vindex @samp{&} AND @vindex @samp{^} XOR @vindex @samp{>>} shift @vindex @samp{<<} shift @vindex @samp{!} not @vindex @samp{=} assignment To tell the C-shell that you want to perform an operation on numbers rather than strings, you use the @samp{@@} symbol followed by a space. Then the following operations are possible. @smallexample @@ var = 45 # Assign a numerical value to var echo $var # Print the value @@ var = $var + 34 # Add 34 to var @@ var += 34 # Add 34 to var @@ var -= 1 # subtract 1 from var @@ var *= 5 # Multiply var by 5 @@ var /= 3 # Divide var by 3 (integer division) @@ var %= 3 # Remainder after dividing var by 3 @@ var++ # Increment var by 1 @@ var-- # Decrement var by 1 @@ array[1] = 5 # Numerical array @@ logic = ( $x > 6 && $x < 10) # AND @@ logic = ( $x > 6 || $x < 10) # OR @@ false = ! $var # Logical NOT @@ bits = ( $x | $y ) # Bitwise OR @@ bits = ( $x ^ $y ) # Bitwise XOR @@ bits = ( $x & $y ) # Bitwise AND @@ shifted = ( $var >> 2 ) # Bitwise shift right @@ back = ( $var << 2 ) # Bitwise shift left @end smallexample These operators are precisely those found in the C programming language. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Examples, , Arithmetic, C shell @section Examples The following script uses the operators in the last two sections to take a list of files with a given file extension (say @samp{.doc}) and change it for another (say @samp{.tex}). This is a partial solution to the limitation of not being able to do multiple renames in shell. @smallexample #!/bin/csh -f ############################################################# # # Change file extension for multiple files # ############################################################# if ($#argv < 2) then echo Syntax: chext oldpattern newextension echo "e.g: chext *.doc tex " exit 0 endif mkdir /tmp/chext.$user # Make a scratch area set newext="$argv[$#argv]" # Last arg is new ext set oldext="$argv[1]:e" echo "Old extenstion was ($oldext)"" echo "New extension ($newext) -- okay? (y/n)" switch( $< ) case y: breaksw default: echo "Nothing done." exit 0 endsw ############################################################## # Remove the last file extension from files ############################################################## @ i = 0 foreach file ($argv) @ i++ if ( $i == $#argv ) break cp $file /tmp/chext.$user/$file:r # temporary store end ############################################################### # Add .newext file extension to files ############################################################### set array = (`ls /tmp/chext.$user`) foreach file ($array) if ( -f $file.$newext ) then echo destination file $file.$newext exists. No action taken. continue endif cp /tmp/chext.$user/$file $file.$newext rm $file.$oldext end rm -r /tmp/chext.$user @end smallexample Here is another example to try to decipher. Use the manual pages to find out about @samp{awk}. This script can be written much more easily in Perl or C, as we shall see in the next chapters. It is also trivially implemented as a script in the system administration language cfengine. @smallexample #!/bin/csh -f ########################################################### # # KILL all processes owned by $argv[1] with PID > $argv[2] # ########################################################### @cindex @samp{whoami} command if ("`whoami`" != "root") then echo Permission denied exit 0 endif if ( $#argv < 1 || $#argv > 2 ) then echo Usage: KILL username lowest-pid exit 0 endif if ( $argv[1] == "root") then echo No! Too dangerous -- system will crash exit 0 endif ############################################################ # Kill everything ############################################################ if ( $#argv == 1 ) then set killarray = ( `ps aux | awk '@{ if ($1 == user) \ @{printf "%s ",$2@}@}' user=$argv[1]` ) foreach process ($killarray) kill -1 $process kill -15 $process > /dev/null kill -9 $process > /dev/null if ("`kill -9 $process | egrep -e 'No such process'`" == "") then echo "Warning - $process would not die - try again" endif end ############################################################# # Start from a certain PID ############################################################# else if ( $#argv == 2 ) then set killarray = ( `ps aux | awk '@{ if ($1 == user && $2 > uid) \ @{printf "%s ",$2@}@}' user=$argv[1] uid=$argv[2]` ) foreach process ($killarray) @cindex @samp{sleep} command kill -1 $process > /dev/null kill -15 $process sleep 2 kill -9 $process > /dev/null if ("`kill -9 $process | egrep -e 'No such process'`" == "") then echo "Warning - $process would not die - try again" endif end endif @end smallexample @noindent This program would be better written in C or Perl. @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node Bourne shell, Perl, C shell, Top @chapter Bourne shell @cindex Bourne shell Programmers who are used to C or C++ often find it easier to program in C-shell because there are strong similarities between the two. The Bourne shell is somewhat different in style, but is structured in a way which makes it better suited to more complicated script writing, especially for system administrators. Also it is closer to the kernels own exec mechanism. The Bourne shell allows subroutines and default values for parameters. Most of the system scripts in UNIX are written in the Bourne shell. The principles of the Bourne shell are largely the same as those for the C-shell, so we shall skip fairly quickly through the details. Historically, the Bourne shell came before the C shell. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * .profile:: * Variables and export:: * stdin stdout stderr:: * Arithmetic in sh:: * Scripts and arguments:: * Return codes:: * Tests in sh:: * Input in sh:: * Loops in sh:: * Procedures and traps:: * setuid setgid scripts:: * Limitations of shell:: * Ex4:: @end menu @node .profile, Variables and export, Bourne shell, Bourne shell @section .profile @vindex .profile @vindex .cshrc @cindex @file{.profile} set up in sh The @samp{.profile} file is the Bourne shell's answer to @samp{.cshrc}. This file is read by interactive @file{/bin/sh} shells on starting up. On Sun systems the file @file{/etc/profile} is also read. On @samp{HPUX} machines, the file @file{/etc/src.sh} is read. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Variables and export, stdin stdout stderr, .profile, Bourne shell @section Variables and export @cindex Global variables in sh @cindex Local variables in sh Local and global variables are both defined using the syntax @smallexample VARIABLE="Some string" VAR=13 @end smallexample @noindent It is important that there be no space between the variable and the equals sign. By default these variables are @emph{local}. To make them global (so that child processes will inherit them) we use the command @smallexample export VARIABLE @end smallexample @cindex @samp{EXPORT} command in sh @noindent This adds the variable to the process @emph{environment}. It is the analogue of making `environment variables' with @code{setenv} in C shell. The command @smallexample set -a @end smallexample @noindent changes the default so that all variables, after the command are created @emph{global}. Arrays or lists are often simulated in shell by sandwiching the colon @samp{:} symbol between items @smallexample PATH=/bin:/usr/bin:/etc:/local/bin:. LD_LIBARAY_PATH=/usr/lib:/usr/openwin/lib:/local/lib @end smallexample @noindent but there is no real facility for arrays in the Bourne shell. Note that the UNIX @samp{cut} command can be used to extract the elements of the list. Loops can also read such lists directly @xref{Loops in sh}. A Perl script can also be used. The value of a variable is given by the dollar symbol as in C-shell. It is also possible to use @emph{curly braces} around the variable name to `protect' the variable from interfering text. For example: @smallexample $ animal=worm $ echo book$animal bookworm $ thing=book $ echo $thingworm (nothing..) $ echo $@{thing@}worm bookworm @end smallexample @noindent Default values can be given to variables in the Bourne shell. The following commands illustrate this. @smallexample echo $@{var-"No value set"@} echo $@{var="Octopus"@} echo $@{var+"Forced value"@} echo $@{var?"No such variable"@} @end smallexample @noindent The first of these prints out the contents of @samp{$var}, if it is defined. If it is not defined the variable is substituted for the string "No value set". The value of @samp{var} is not changed by this operation. It is only for convenience. The second command has the same effect as the first, but here the value of @samp{$var} is actually changed to "Octopus" if @samp{$var} is not set. The third version is slightly peculiar. If @samp{$var} is @emph{already} set, its value will be forced to be "Forced value", otherwise it is left undefined. Finally the last instance issues an error message "No such variable" if @samp{$var} is not defined. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node stdin stdout stderr, Arithmetic in sh, Variables and export, Bourne shell @section Stdin, stdout and stderr In the Bourne shell, the standard input/output files are referred to by numbers rather than by names. @cindex Standard I/O in sh @table @emph @item stdin File number 0 @item stdout File number 1 @item stderr File number 2 @end table @noindent The default routes for these files can be changed by @emph{redirection}. The redirection commands are more complicated than in C-shell, but they are also more flexible. Here is a comparison. @cindex Redirecting stdio in sh @cindex @samp{1>} in sh @cindex @samp{2>} in sh @cindex @samp{2>&1} in sh @smallexample sh csh Description command > file command > file Stdout to file command 1> file command > file Stdout to file command 2> errs (No analogue) Stderr only to file errs command 1> file 2>&1 command >& file stdout and stderr to file command 1> file 2> errs (No analogue) stdout to file, stderr to errs @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Arithmetic in sh, Scripts and arguments, stdin stdout stderr, Bourne shell @section Arithmetic in sh @cindex Arithemtic in sh Arithmetic is performed entirely `by proxy'. There are no internal arithmetic operators as in the C-shell. To evaluate an expression we call the @samp{expr} command or the @samp{bc} precision calculator. Here are some examples of @samp{expr} @smallexample a=`expr $a+1` # increment a a=`expr 4 + 10 \* 5` # 4+10*5 check = `expr $a \> $b` # true=1, false=0. True if $a > $b @end smallexample @noindent @samp{expr} is very sensitive to spaces and backslash characters. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Scripts and arguments, Return codes, Arithmetic in sh, Bourne shell @section Scripts and arguments Scripts are created by making an executable file which begins with the sequence of characters @cindex Scripts in sh, making @cindex Making scripts in sh @cindex @samp{#!program} sequence @smallexample #!/bin/sh @end smallexample @noindent Although we didn't discuss it before, this construction is quite general: any executable file which begins with a sequence @smallexample #!myprogram -option @end smallexample @noindent will cause the shell to attempt to execute @smallexample myprogam -option filename @end smallexample @noindent where @var{filename} is the name of the file. @cindex Command line arguments in sh If a script is to accept arguments then these can be referred to as @samp{ $1 $2 $3..$9}. There is a logical limit of nine arguments to a script, but in practice it is possibile to get around this limitation. @samp{$0} is the name of the script itself. Here is a simple script in the Bourne shell which prints out all its arguments. @cindex @code{for} loop in sh @smallexample #!/bin/sh # # Print all arguments (version 1) # for arg in $* do echo Argument $arg done echo Total number of arguments was $# @end smallexample @noindent The @samp{$*} symbol stands for the entire list of arguments (like @samp{$argv} in C-shell) and @samp{$#} is the total number of arguments (like @samp{$#argv} in C-shell). Another way of achieving the same is to use the @samp{shift} command. We shall meet this again in the Perl programming language. @samp{shift} takes the first argument from the argument list and deletes it, moving all of the other arguments down one number -- this is how we can handle long lists of arguments in @samp{sh}. @cindex @code{while} loop in sh @cindex @code{shift} operator on strings @smallexample #!/bin/sh # # Print all arguments (version 2) # while ( true ) do arg=$1; shift; echo $arg was an argument; if [ $# -eq 0 ]; then break fi done @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Return codes, Tests in sh, Scripts and arguments, Bourne shell @section Return codes @cindex Return codes @cindex Decisions and return codes in sh All programs which execute in UNIX return a value through the C @samp{return} command. There is a convention that a return value of zero (0) means that everything went well, whereas any other value implies that some error occurred. The return value is usually the value returned in @samp{errno}, the extenal error variable in C. Shell scripts can test for these values either by placing the command directly inside an `if' test, or by testing the variable @samp{$?} which is always set to the return code of the last command. Some examples are given following the next two sections. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Tests in sh, Input in sh, Return codes, Bourne shell @section Tests and conditionals @cindex Tests in sh The Bourne shell has the usual array of tests. They are written as follows. Notice that @samp{test} is itself not a part of the shell, but is a program which works out conditions and provides a return code. See the manual page on @samp{test} for more details. @vindex @samp{-f} in sh @vindex @samp{-d} in sh @vindex @samp{-r} in sh @vindex @samp{-w} in sh @vindex @samp{-x} in sh @vindex @samp{-s} in sh @vindex @samp{-g} in sh @vindex @samp{-u} in sh @vindex @samp{=} in sh @vindex @samp{!=} in sh @vindex @samp{-eq} in sh @vindex @samp{-ne} in sh @vindex @samp{-gt} in sh @vindex @samp{-lt} in sh @vindex @samp{-ge} in sh @vindex @samp{-le} in sh @vindex @samp{!} in sh @vindex @samp{-a} in sh @vindex @samp{-o} in sh @table @code @item test -f @var{file} True if the file is a plain file @item test -d @var{file} True if the file is a directory @item test -r @var{file} True if the file is readable @item test -w @var{file} True if the file is writable @item test -x @var{file} True if the file is executable @item test -s @var{file} True if the file contains something @item test -g @var{file} True if setgid bit is set @item test -u @var{file} True if setuid bit is set @item test s1 = s2 True if strings s1 and s2 are equal @item test s1 != s2 True if strings s1 and s2 are unequal @item test @var{x} -eq @var{y} True if the integers @var{x} and @var{y} are numerically equal @item test @var{x} -ne @var{y} True if integers are not equal @item test @var{x} -gt @var{y} True if @var{x} is greater than @var{y} @item test @var{x} -lt @var{y} True if @var{x} is less than @var{y} @item test @var{x} -ge @var{y} True if @var{x}>=@var{y} @item test @var{x} -le @var{y} True if @var{x} <= @var{y} @item ! Logical NOT operator @item -a Logical AND @item -o Logical OR @end table Note that an alternate syntax for writing these commands if to use the square brackets, instead of writing the word test. @cindex @samp{test} in sh @cindex @samp{[]} for test in sh @smallexample [ $x -lt $y ] "==" test $x -lt $y @end smallexample @cindex @code{if..then..else..fi} in sh The conditional structures have the following syntax. @smallexample if @var{unix-command} then @var{command} else @var{commands} fi @end smallexample @noindent The @samp{else} clause is, of course, optional. As noted before, the first unix command could be @emph{anything}, since every command has a return code. The result is TRUE if it evaluates to @emph{zero} and false otherwise (in contrast to the conventions in most languages). Multiple tests can be made using @smallexample if @var{unix-command} then @var{commands} elif @var{unix-command} then @var{commands} elif @var{unix-command} then @var{commands} else @var{commands} fi @end smallexample @noindent where @samp{elif} means `else-if'. The equivalent of the C-school's @samp{switch} statement is a more Pascal-like @samp{case} structure. @smallexample case @var{unix-command-or-variable} in wildcard1) @var{commands} ;; wildcard2) @var{commands} ;; wildcard3) @var{commands} ;; esac @end smallexample @noindent This structure uses the wildcards to match the output of the command or variable in the first line. The first pattern which matches gets executed. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Input in sh, Loops in sh, Tests in sh, Bourne shell @section Input from the user in sh In shell you can read the value of a variable using the @samp{read} command, with syntax @smallexample read @var{variable} @end smallexample @noindent This reads in a string from the keyboard and terminates on a newline character. Another way to do this is to use the @samp{input} command to access a particular logical device. The keyboard device in the current terminal is @file{/dev/tty}, so that one writes @smallexample @var{variable} = `line < /dev/tty` @end smallexample @noindent which fetches a single line from the user. @cindex Input in sh Here are some examples of these commands. First a program which asks yes or no... @smallexample #!/bin/sh # # Yes or no # echo "Please answer yes or no: " answer=`line < /dev/tty` case $answer in y* | Y* | j* | J* ) echo YES!! ;; n* | N* ) echo NO!! ;; *) echo "Can't you answer a simple question?" esac echo The end @end smallexample @noindent Notice the use of pattern matching and the @samp{|} `OR' symbol. @smallexample #!/bin/sh # # Kernel check # if test ! -f /vmunix # Check that the kernel is there! then echo "This is not BSD unix...hmmm" if [ -f /hp-ux ] then echo "It's a Hewlett Packard machine!" fi elif [ -w /vmunix ] then echo "HEY!! The kernel is writable my me!"; else echo "The kernel is write protected." echo "The system is safe from me today." fi @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Loops in sh, Procedures and traps, Input in sh, Bourne shell @section Loops in sh @cindex Loops in sh @cindex @samp{while} in sh The loop structures in the Bourne shell have the following syntax. @smallexample while @var{unix-command} do commands done @end smallexample @noindent The first command will most likely be a test but, as before, it could in principle be any UNIX command. The @samp{until} loop, reminiscent of BCPL, carries out a task until its argument evaluates to TRUE. @smallexample until @var{unix-command} do commands done @end smallexample @cindex @samp{until} @noindent Finally the @samp{for} structure has already been used above. @smallexample for @var{variable} in @var{list} do @var{commands} done @end smallexample Often we want to be able to use an array of values as the list which @code{for} parses, but Bourne shell has no array variables. This problem is usually solved by making a long string separated by, for example, colons. For example, the @code{$PATH} variable has the form @smallexample PATH = /usr/bin:/bin:/local/gnu/bin @end smallexample @cindex @samp{IFS} variable in sh @cindex Redefining list separator in sh @cindex Loops and list separators @noindent Bourne shell allows us to split such a string on whatever character we wish. Normally the split is made on spaces, but the variable @samp{IFS} can be defined with a replacement. To make a loop over all directories in the command path we would therefore write @smallexample IFS=: for name in $PATH; do @var{commands} done @end smallexample @noindent The best way to gain experience with these commands is through some examples. @smallexample #!/bin/sh # # Get text from user repeatedly # echo "Type away..." while read TEXT do echo You typed $TEXT if [ "$TEXT" = "quit" ]; then echo "(So I quit!)" exit 0 fi done echo "HELP!" @end smallexample @noindent @cindex @key{CTRL-D} and EOF @cindex End of file @key{CTRL-D} This very simple script is a typical use for a while-loop. It gets text repeatedly until the user type `quit'. Since read never returns `false' unless an error occurs or it detects an EOF (end of file) character @key{CTRL-D}, it will never exit without some help from an `if' test. If it does receive a @key{CTRL-D} signal, the script prints `HELP!'. @smallexample #!/bin/sh # # Watch in the background for a particular user # and give alarm if he/she logs in # # To be run in the background, using & # if [ $# -ne 1 ]; then echo "Give the name of the user as an argument" > /dev/tty exit 1 fi echo "Looking for $1" until users | grep -s $1 do sleep 60 done echo "!!! WAKE UP !!!" > /dev/tty echo "User $1 just logged in" > /dev/tty @end smallexample This script uses @samp{grep} in `silent mode' (-s option). i.e. grep never writes anything to the terminal. The only thing we are interested in is the return code the piped command produces. If `grep' detects a line containing the username we are interested in, then the result evaluates to TRUE and the sleep-loop exits. Our final example is the kind of script which is useful for a system administrator. It transfers over the Network Information Service database files so that a slave server is up to date. All we have to do is make a list of the files and place it in a @samp{for} loop. The names used below are the actual names of the NIS maps, well known to system administrators. @smallexample #!/bin/sh # # Update the NIS database maps on a client server. This program # shouldn't have to be run, but sometimes things go wrong and we # have to force a download from the main sever. # PATH=/etc/yp:/usr/etc/yp:$PATH MASTER=myNISserver for map in auto.direct auto.master ethers.byaddr ethers.byname\ group.bygid group.byname hosts.byaddr hosts.byname\ mail.aliases netgroup.byhost netgroup.byuser netgroup\ netid.byname networks.byaddr networks.byname passwd.byname\ passwd.byuid priss.byname protocols.byname protocols.bynumber\ rpc.bynumber services.byname services usenetgroups.byname; do ypxfr $1 -h $MASTER $map done @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Procedures and traps, setuid setgid scripts, Loops in sh, Bourne shell @section Procedures and traps @cindex Procedures and subroutines in sh @cindex Signal handler in sh @cindex Interrupt handler in sh @cindex Traps in sh One of the worthy features of the Bourne shell is that it allows you to define @emph{subroutines} or @emph{procedures}. Subroutines work just like subroutines in any other programming language. They are executed in same shell (not as a sub-process). Here is an interesting program which demonstrates two useful things at the same time. First of all, it shows how to make a hierachical subroutine structure using the Bourne shell. Secondly, it shows how the @samp{trap} directive can be used to trap signals, so that Bourne shell programs can exit safely when they are killed or when CTRL-C is typed. @smallexample #!/bin/sh # # How to make a signal handler in Bourne Shell # using subroutines # ##################################################### # Level 2 ##################################################### ReallyQuit() @{ while true do echo "Do you really want to quit?" read answer case $answer in y* | Y* ) return 0;; *) echo "Resuming..." return 1;; esac done @} ##################################################### # Level 1 ##################################################### SignalHandler() @{ if ReallyQuit # Call a function then exit 0 else return 0 fi @} ##################################################### # Level 0 : main program ##################################################### trap SignalHandler 2 15 # Trap kill signals 2 and 15 echo "Type some lines of text..." while read text do echo "$text - CTRL-C to exit" done @end smallexample Note that the logical tree structure of this program is upside down (the highest level comes at the bottom). This is because all subroutines must be defined before they are used. This example concludes our brief survey of the Bourne shell. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node setuid setgid scripts, Limitations of shell, Procedures and traps, Bourne shell @section setuid and setgid scripts @cindex SetUID scripts @cindex Root privileges The superuser @samp{root} is the only privileged user in UNIX. All other users have only restricted access to the system. Usually this is desirable, but sometimes it is a nuisance. A setuid script is a script which has its @emph{setuid-bit} set. When such a script is executed by a user, it is run with all the rights and privileges of the owner of the script. All of the commands in the script are executed as the owner of the file and not with the user-id of the person who ran the script. If the owner of the setuid script is @samp{root} then the commands in the script are run with @emph{root privileges}! Setuid scripts are clearly a touchy security issue. When giving away one's rights to another user (especially those of @samp{root}) one is tempting hackers. Setuid scripts should be @emph{avoided}. A setgid program is almost the same, but only the group id is set to that of the owner of the file. Often the effect is the same. An example of a setuid program is the @samp{ps} program. @samp{ps} lists all of the processes running in the kernel. In order to do this it needs permission to access the private data structures in the kernel. By making @samp{ps} setgid root, it allows ordinary users to be able to read as much as the writers of @samp{ps} thought fit, but no more. Naturally, only the superuser can make a file setuid or setgid root. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Limitations of shell, Ex4, setuid setgid scripts, Bourne shell @section Summary: Limitations of shell programming @cindex Limitations of shell programs To summarize the last two long and oppressive chapters we shall take a step back from the details and look at what we have achieved. The idea behind the shell is to provide a user interface, with access to the system's facilities at a simple level. In the 70's user interfaces were not deisgned to be user-friendly. The UNIX shell is not particularly use friendly, but it is very powerful. Perhaps it would have been enough to provide only commands to allow users to write C programs. Since all of the system functions are available from C, that would certainly allow everyone to do what anything that UNIX can do. But shell programming is much more @emph{immediate} than C. It is an environment of @emph{frequently used tools}. Also for quick programming solutions: C is a compiled language, whereas the shell is an interpreter. A quick shell program can solve many problems in no time at all, without having to compile anything. Shell programming is only useful for `quick and easy' programs. To use it for anything serious is an abuse. Programming difficult things in shell is clumsy, and it is difficult to get returned-information (like error messages) back in a useful form. Besides, shell scripts are slow compared to real programs since they involve starting a new program for each new command. These difficulties are solved partly by Perl, which we shall consider next -- but in the final analysis, real programs of substance need to be written in C. Contrary to popular belief, this is not more difficult than programming in the shell -- in fact, many things are much simpler, because all of the shell commands originated as C functions. The shell is an extra layer of the UNIX onion which we have to battle our way through to get where we're going. Sometimes it is helpful to be shielded from @emph{low level} details -- sometimes it is a @emph{hindrance}. In the remaining chapters we shall consider more involved programming needs. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Ex4, , Limitations of shell, Bourne shell @section Exercises @enumerate @item Write an improved @samp{which} command in C-shell. @item Make a counter program which records in a file how many times you log in to your account. You can call this in your .cshrc file. @item Make a Bourne shell script to kill all the processes owned by a particular user. (Note, that if you are not the superuser, you cannot kill processes owned by other users.) @item Write a script to replace the @samp{rm} command with something safer. Think about a way of implementing @samp{rm} so that it is possible to get deleted files back again in case of emergencies. This is not possible using the normal @samp{rm} command. Hint: save files in a hidden directory @samp{.deleted}. Make your script delete files in the @samp{.deleted} directory if they are older than a week, so that you don't fill up the disk with rubbish. @item Suppose you have a bunch of files with a particular file-extension: write a script in csh to change the extension to something else. e.g. to change *.C into *.c. Give the old and new extensions as arguments to the script. @item Write a program in sh to search for files in the current directory which contain a certain string. e.g. search for all files which contain the word "if". Hint: use the "find" command. @item Use the manual pages to find out about the commands @samp{at}, @samp{batch} and @samp{atq}. Test these commands by executing the shell command @samp{date} at some time of your choice. Use the @samp{-m} option so that the result of the job is mailed to you. @cindex @samp{batch} command @cindex @samp{at} command @item Write a script in sh or csh to list all of the files bigger than a certain size starting from the current directory, and including all subdirectories. This kind of program is useful for system administrators when a disk becomes full. @end enumerate @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node Perl, WWW and CGI programming, Bourne shell, Top @chapter Perl @cindex Perl So far, we have been looking at shell programming for performing fairly simple tasks. Now let's extend the idea of shell programming to cover more complex tasks like systems programming and network communications. Perl is a language which was designed to retain the immediateness of shell languages, but at the same time capture some of the flexibility of C. Perl is an acronym for @emph{Practical extraction and report language}. In this chapter, we shall not aim to teach Perl from scratch -- the best way to learn it is to use it! Rather we shall concentrate on demonstrating some principles. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * sed awk cut paste:: * Perl structure:: * Perl variables:: * Loops and conditionals:: * Files in perl:: * Perl subroutines:: * die:: * stat():: * Perl examples:: * Pattern matching:: * search and replace:: * WWW example:: * WWW example 2:: * Other functions:: * Perl summary:: * Ex5:: * Proj5:: @end menu @node sed awk cut paste, Perl structure, Perl, Perl @section Sed and awk, cut and paste @cindex @samp{sed} batch editor @cindex @samp{awk} pattern extractor @cindex @samp{cut} @cindex @samp{paste} One of the reasons for using Perl is that it is extremely good at textfile handling--one of the most important things for UNIX users, and particularly useful in connection with CGI script processing on the World Wide Web. It has simple built-in constructs for searching and replacing text, storing information in arrays and retrieving them in sorted form. All of the these things have previously been possible using the UNIX shell commands @smallexample sed awk cut paste @end smallexample @noindent but these commands were designed to work primarily in the Bourne shell and are a bit @samp{awk}ward to use for all but the simplest applications. @table @samp @item sed is a stream editor. It takes command line instructions, reads input from the stream @code{stdin} and produces output on @emph{stdout} according to those instructions. @samp{sed} works line by line from the start of a textfile. @item awk is a pattern matching and processing language. It takes a textfile and reads it line by line, matching @emph{regular expressions} and acting on them. @samp{awk} is powerful enough to have conditional instructions like @samp{if..then..else} and uses C's @samp{printf} construction for output. @item cut Takes a line of input and cuts it into @emph{fields}, separated by some character. For instance, a normal line of text is a string of words separated by spaces. Each word is a different field. @samp{cut} can be used, for instance, to pick out the third column in a table. Any character can be specified as the separator. @item paste is the logical opposite of cut. It concatenates @math{n} files, and makes each line in the file into a column of a table. For instance, @samp{paste one two three} would make a table in which the first column consisted of all lines in @samp{one}, the second of all lines in @samp{two} and the third of all lines in @samp{three}. If one file is longer than the others, then some columns have blank spaces. @end table Perl unifies all of these operations and more. It also makes them much simpler. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Perl structure, Perl variables, sed awk cut paste, Perl @section Program structure To summarize Perl, we need to know about the structure of a Perl program, the conditional constructs it has, its loops and its variables. In the latest versions of Perl (Perl 5), you can write object oriented programs of great complexity. We shall not go into this depth, for the simple reason that Perl's strength is not as a general programming language but as a specialized language for textfile handling. The syntax of Perl is in many ways like the C programming language, but there are important differences. @itemize @bullet @item Variables do not have @emph{types}. They are interpreted in a context sensitive way. The operators which acts upon variables determine whether a variable is to be considered a string or as an integer etc. @cindex Perl variables and types @cindex Types in perl @item Although there are no types, Perl defines @emph{arrays} of different kinds. There are three different kinds of array, labelled by the symbols @samp{$}, @samp{@@} and @samp{%}. @cindex Arrays in perl @item Perl keeps a number of standard variables with special names e.g. @samp{$_ @@ARGV} and @samp{%ENV}. Special attention should be paid to these. They are very important! @cindex Argument vector in perl @cindex Command line arguments in perl @cindex Environment variables in perl @item The shell reverse apostrophe notation @code{`command`} can be used to execute UNIX programs and get the result into a Perl variable. @cindex @samp{`..`} in perl @end itemize Here is a simple `structured hello world' program in Perl. Notice that subroutines are called using the @samp{&} symbol. There is no special way of marking the main program -- it is simply that part of the program which starts at line 1. @smallexample #!/local/bin/perl # # Comments # &Hello(); &World; # end of main sub Hello @{ print "Hello"; @} sub World @{ print "World\n"; @} @end smallexample @noindent The parentheses on subroutines are optional, if there are no parameters passed. Notice that each line must end in a semi-colon. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Perl variables, Loops and conditionals, Perl structure, Perl @section Perl variables @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Scalar variables:: * The default scalar variable:: * Array variables:: * Array commands:: * Associated arrays:: * Array example:: @end menu @node Scalar variables, The default scalar variable, Perl variables, Perl variables @subsection Scalar variables @cindex Scalar variables in perl @cindex Strings in perl @cindex Perl, strings and scalar In Perl, variables do not have to be declared before they are used. Whenever you use a new symbol, Perl automatically adds the symbol to its symbol table and initializes the variable to the empty string. It is important to understand that there is no practical difference between zero and the empty string in perl -- except in the way that you, the user, choose to use it. Perl makes no distinction between strings and integers or any other types of data -- except when it wants to interpret them. For instance, to compare two variables as strings is not the same as comparing them as integers, even if the string contains a textual representation of an integer. Take a look at the following program. @smallexample #!/local/bin/perl # # Nothing! # print "Nothing == $nothing\n"; print "Nothing is zero!\n" if ($nothing == 0); if ($nothing eq "") @{ print STDERR "Nothing is really nothing!\n"; @} $nothing = 0; print "Nothing is now $nothing\n"; @end smallexample @noindent The output from this program is @smallexample Nothing == Nothing is zero! Nothing is really nothing! Nothing is now 0 @end smallexample @cindex Interpretation of values in perl There are several important things to note here. First of all, we never declare the variable `nothing'. When we try to write its value, perl creates the name and associates a NULL value to it i.e. the empty string. There is no error. Perl knows it is a variable because of the @samp{$} symbol in front of it. All @emph{scalar} variables are identified by using the dollar symbol. @cindex @samp{eq} and @samp{==} in perl @vindex @samp{eq} @vindex @samp{==} Next, we compare the value of @samp{$nothing} to the integer `0' using the integer comparison symbol @samp{==}, and then we compare it to the empty string using the string comparison symbol @samp{eq}. Both tests are true! That means that the empty string is interpreted as having a numerical value of zero. In fact @emph{any string} which does not form a valid integer number has a numerical value of zero. Finally we can set @samp{$nothing} explicitly to a valid integer string zero, which would now pass the first test, but fail the second. As extra spice, this program also demonstrates two different ways of writing the @samp{if} command in perl. @c ..................................................... @c SUBSECTION @c ..................................................... @node The default scalar variable, Array variables, Scalar variables, Perl variables @subsection The default scalar variable. The special variable @samp{$_} is used for many purposes in Perl. It is used as a buffer to contain the result of the last operation, the last line read in from a file etc. It is so general that many functions which act on scalar variables work by default on @samp{$_} if no other argument is specified. For example, @smallexample print; @end smallexample @noindent is the same as @smallexample print $_; @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node Array variables, Array commands, The default scalar variable, Perl variables @subsection Array (vector) variables @cindex Arrays (normal) in perl The complement of scalar variables is arrays. An array, in Perl is identified by the @samp{@@} symbol and, like scalar variables, is allocated and initialized dynamically. @smallexample @@array[0] = "This little piggy went to market"; @@array[2] = "This little piggy stayed at home"; print "@@array[0] @@array[1] @@array[2]"; @end smallexample @noindent The index of an array is always understood to be a number, not a string, so if you use a non-numerical string to refer to an array element, you will always get the zeroth element, since a non-numerical string has an integer value of zero. An important array which every program defines is @smallexample @@ARGV @end smallexample @noindent This is the argument vector array, and contains the commands line arguments by analogy with the C-shell variable @samp{$argv[]}. Given an array, we can find the last element by using the @samp{$#} operator. For example, @smallexample $last_element = $ARGV[$#ARGV]; @end smallexample @noindent Notice that each element in an array is a scalar variable. The @samp{$#} cannot be interpreted directly as the number of elements in the array, as it can in the C-shell. You should experiment with the value of this quantity -- it often necessary to add 1 or 2 to its value in order to get the behaviour one is used to in the C-shell. Perl does not support multiple-dimension arrays directly, but it is possible to simulate them yourself. (See the Perl book.) @cindex Argument vector in perl @cindex Command line arguments in perl @c ..................................................... @c SUBSECTION @c ..................................................... @node Array commands, Associated arrays, Array variables, Perl variables @subsection Special array commands The @samp{shift} command acts on arrays and returns and removes the first element of the array. Afterwards, all of the elements are shifted down one place. So one way to read the elements of an array in order is to repeatedly call @samp{shift}. @smallexample $next_element=shift(@@myarray); @end smallexample @cindex @samp{shift} and arrays in perl @cindex @samp{shift} and arrays @noindent Note that, if the array argument is omitted, then @samp{shift} works on @samp{@@ARGV} by default. @cindex @samp{split} command @cindex @samp{split} and arrays @cindex Arrays and @samp{split} Another useful function is @samp{split}, which takes a string and turns it into an array of strings. @samp{split} works by choosing a character (usually a space) to delimit the array elements, so a string containing a sentence separated by spaces would be turned into an array of words. The syntax is @smallexample @@array = split; # works with spaces on $_ @@array = split(pattern,string); # Breaks on pattern ($v1,$v2...) = split(pattern,string); # Name array elements with scalars @end smallexample @noindent In the first of these cases, it is assumed that the variable @samp{$_} is to be split on whitespace characters. In the second case, we decide on what characterthe split is to take place and on what string the function is to act. For instance @smallexample @@new_array = split(":","name:passwd:uid:gid:gcos:home:shell"); @end smallexample @noindent The result is a seven element array called @samp{@@new_array}, where @samp{$new_array[0]} is @samp{name} etc. @noindent In the final example, the left hand side shows that we wish to capture elements of the array in a named set of scalar variables. If the number of variables on the lefthand side is fewer than the number of strings which are generated on the right hand side, they are discarded. If the number on the left hand side is greater, then the remainder variables are empty. @c ..................................................... @c SUBSECTION @c ..................................................... @node Associated arrays, Array example, Array commands, Perl variables @subsection Associated arrays @cindex Arrays (associated) in perl One of the very nice features of Perl is the ability to use one string as an index to another string in an array. For example, we can make a short encyclopaedia of zoo animals by constructing an associative array in which the keys (or indices) of the array are the names of animals, and the contents of the array are the information about them. @smallexample $animals@{"Penguin"@} = "A suspicious animal, good with cheese crackers..."; $animals@{"dog"@} = "Plays stupid, but could be a cover..."; if ($index eq "fish") @{ $animals@{$index@} = "Often comes in square boxes. Very cold."; @} @end smallexample @noindent An entire associated array is written @samp{%array}, while the elements are @samp{$array@{$key@}}. Perl provides a special associative array for every program called @samp{%ENV}. This contains the @emph{environment variables} defined in the parent shell which is running the Perl program. For example @smallexample print "Username = $ENV@{"USER"@}\n"; $ld = "LD_LIBRARY_PATH"; print "The link editor path is $ENV@{$ld@}\n"; @end smallexample @cindex Environment variables in perl @noindent To get the current path into an ordinary array, one could write, @smallexample @@path_array= split(":",$ENV@{"PATH"@}); @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node Array example, , Associated arrays, Perl variables @subsection Array example program Here is an example which prints out a list of files in a specified directory, in order of their UNIX protection bits. The @emph{least} protected file files come first. @smallexample #!/local/bin/perl # # Demonstration of arrays and associated arrays. # Print out a list of files, sorted by protection, # so that the least secure files come first. # # e.g. arrays # arrays *.C # ############################################################ print "You typed in ",$#ARGV+1," arguments to command\n"; if ($#ARGV < 1) @{ print "That's not enough to do anything with!\n"; @} while ($next_arg = shift(@@ARGV)) @{ if ( ! ( -f $next_arg || -d $next_arg)) @{ print "No such file: $next_arg\n"; next; @} ($dev,$ino,$mode,$nlink,$uid,$gid,$rdev,$size) = stat($next_arg); $octalmode = sprintf("%o",$mode & 0777); $assoc_array@{$octalmode@} .= $next_arg. " : size (".$size."), mode (".$octalmode.")\n"; @} print "In order: LEAST secure first!\n\n"; foreach $i (reverse sort keys(%assoc_array)) @{ print $assoc_array@{$i@}; @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Loops and conditionals, Files in perl, Perl variables, Perl @section Loops and conditionals @cindex For loops in perl @cindex @code{for} loop in perl @cindex @code{foreach} loop in perl Here are some of the most commonly used decision-making constructions and loops in Perl. The following is not a comprehensive list -- for that, you will have to look in the Perl bible: @emph{Programming Perl}, by Larry Wall and Randal Schwartz. The basic pattern follows the C programming language quite closely. In the case of the @samp{for} loop, Perl has both the C-like version, called @samp{for} and a @samp{foreach} command which is like the C-shell implementation. @cindex @samp{if} in perl @cindex @samp{unless} in perl @cindex @samp{while} in perl @cindex @samp{do..while} in perl @smallexample if (@var{expression}) @{ block; @} else @{ block; @} command if (@var{expression}); unless (@var{expression}) @{ block; @} else @{ block; @} while (@var{expression}) @{ block; @} do @{ block; @} while (@var{expression}); for (@var{initializer}; @var{expression}; @var{statement}) @{ block; @} foreach @var{variable}(@var{array}) @{ block; @} @end smallexample @noindent In all cases, the @samp{else} clauses may be omitted. Strangely, perl does not have a @samp{switch} statement, but the Perl book describes how to make one using the features provided. @c ..................................................... @c SUBSECTION @c ..................................................... @menu * The for loop:: * The foreach loop:: * Iterating over elements in arrays:: * Iterating over lines in a file:: @end menu @node The for loop, The foreach loop, Loops and conditionals, Loops and conditionals @subsection The for loop The for loop is exactly like that in C or C++ and is used to iterate over a numerical index, like this: @smallexample for ($i = 0; $i < 10; $i++) @{ print $i, "\n"; @} @end smallexample @cindex For loop @c ..................................................... @c SUBSECTION @c ..................................................... @node The foreach loop, Iterating over elements in arrays, The for loop, Loops and conditionals @subsection The foreach loop The foreach loop is like its counterpart in the C shell. It is used for reading elements one by one from a regular array. For example, @smallexample foreach $i ( @@array ) @{ print $i, "\n"; @} @end smallexample @cindex Foreach loop @c ..................................................... @c SUBSECTION @c ..................................................... @node Iterating over elements in arrays, Iterating over lines in a file, The foreach loop, Loops and conditionals @subsection Iterating over elements in arrays One of the main uses for @samp{for} type loops is to iterate over successive values in an array. This can be done in two ways which show the essential difference between @code{for} and @code{foreach}. If we want to fetch each value in an array in turn, without caring about numerical indices, the it is simpest to use the @code{foreach} loop. @cindex Iteration over arrays @smallexample @@array = split(" ","a b c d e f g"); foreach $var ( @@array ) @{ print $var, "\n"; @} @end smallexample @noindent This example prints each letter on a separate line. If, on the other hand, we are interested in the index, for the purposes of some calculation, then the @code{for} loop is preferable. @smallexample @@array = split(" ","a b c d e f g"); for ($i = 0; $i <= $#array; $i++) @{ print $array[$i], "\n"; @} @end smallexample @noindent Notice that, unlike the for-loop idiom in C/C++, the limit is @samp{$i <= $#array}, i.e. `less than or equal to' rather than `less than'. This is because the @samp{$#} operator does not return the number of elements in the array but rather the last element. Associated arrays are slightly different, since they do not use numerical keys. Instead they use a set of strings, like in a database, so that you can use one string to look up another. In order to iterate over the values in the array we need to get a list of these strings. The @code{keys} command is used for this. @vindex keys @cindex Associated arrays, iteration @smallexample $assoc@{"mark"@} = "cool"; $assoc@{"GNU"@} = "brave"; $assoc@{"zebra"@} = "stripy"; foreach $var ( keys %assoc ) @{ print "$var , $assoc@{$var@} \n"; @} @end smallexample @noindent The order of the keys is not defined in the above example, but you can choose to sort them alphabetically by writing @smallexample foreach $var ( sort keys %assoc ) @end smallexample @noindent instead. @c ..................................................... @c SUBSECTION @c ..................................................... @node Iterating over lines in a file, , Iterating over elements in arrays, Loops and conditionals @subsection Iterating over lines in a file Since Perl is about file handling we are very interested in reading files. Unlike C and C++, perl likes to read files line by line. The angle brackets are used for this, @xref{Files in perl}. Assuming that we have some file handle @samp{}, for instance @samp{}, we can always read the file line by line with a while-loop like this. @smallexample while ($line = ) @{ print $line; @} @end smallexample @noindent Note that @code{$line} includes the end of line character on the end of each line. If you want to remove it, you should add a `chop' command: @smallexample while ($line = ) @{ chop $line; print "line = ($line)\n"; @} @end smallexample @cindex Iterating over files @cindex Files, iterating over lines @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Files in perl, Perl subroutines, Loops and conditionals, Perl @section Files in perl @cindex Files in perl @cindex Standard I/O in perl @cindex @samp{open} command in perl @cindex @samp{close} command in perl Opening files is straightforward in Perl. Files must be opened and closed using -- wait for it -- the commands @samp{open} and @samp{close}. You should be careful to close files after you have finished with them -- especially if you are writing @emph{to} a file. Files are buffered and often large parts of a file are not actually written until the @samp{close} command is received. Three files are, of course, always open for every program, namely @samp{STDIN}, @samp{STDOUT}and @samp{STDERR}. Formally, to open a file, we must obtain a file descriptor or file handle. This is done using @samp{open}; @cindex @samp{<>} filehandle in perl @cindex File handles in perl @smallexample open (file_descrip,"Filename"); @end smallexample @noindent The angular brackets @samp{<..>} are used to read from the file. For example, @smallexample $line = ; @end smallexample @noindent reads one line from the file associated with @samp{file_descrip}. Let's look at some examples of filing opening. Here is how we can implement UNIX's @samp{cut} and @samp{paste} commands in perl: @cindex Cut as a perl script @smallexample #!/local/bin/perl # # Cut in perl # # Cut second column while (<>) @{ @@cut_array = split; print "@@cut_array[1]\n"; @} @end smallexample @noindent This is the simplest way to open a file. The empty file descriptor @samp{<>} tells perl to take the argument of the command as a filename and open that file for reading. This is really short for @samp{while($_=)} with the standard input redirected to the named file. @cindex Paste as a perl script The @samp{paste}program can be written as follows: @smallexample #!/local/bin/perl # # Paste in perl # # Two files only, syntax : paste file 1file2 # open (file1,"@@ARGV[0]") || die "Can't open @@ARGV[0]\n"; open (file2,"@@ARGV[1]") || die "Can't open @@ARGV[1]\n"; while (($line1 = ) || ($line2 = )) @{ chop $line1; chop $line2; print "$line1 $line2\n"; # tab character between @} @end smallexample Here we see more formally how to read from two separate files at the same time. Notice that, by putting the read commands into the test-expression for the @samp{while} loop, we are using the fact that @samp{<..>} returns a non-zero (true) value unless we have reached the end of the file. To write and append to files, we use the shell redirection symbols inside the @samp{open} command. @smallexample open(fd,"> filename"); # open file for writing open(fd,">> filename"); # open file for appending @end smallexample @noindent We can also open a pipe from an arbitrary UNIX command and receive the output of that command as our input: @smallexample open (fd,"/bin/ps aux | "); @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Simplest perl program:: * == and eq:: * chop:: @end menu @node Simplest perl program, == and eq, Files in perl, Files in perl @subsection A simple perl program Let us now write the simplest perl program which illustrates the way in which perl can save time. We shall write it in three different ways to show what the short cuts mean. Let us implement the @samp{cat} command, which copies files to the standard output. The simplest way to write this is perl is the following: @smallexample #!/local/bin/perl while (<>) @{ print; @} @end smallexample @noindent Here we have made heavy use of the many default assumptions which perl makes. The program is simple, but difficult to understand for novices. First of all we use the default file handle @code{<>} which means, take one line of input from a default file. This object returns true as long as it has not reached the end of the file, so this loop continues to read lines until it reaches the end of file. The default file is standard input, unless this script is invoked with a command line argument, in which case the argument is treated as a filename and perl attempts to open the argument-filename for reading. The @code{print} statement has no argument telling it what to print, but perl takes this to mean: print the default variable @samp{$_}. We can therefore write this more explicitly as follows: @smallexample #!/local/bin/perl open (HANDLE,"$ARGV[1]"); while () @{ print $_; @} @end smallexample @noindent Here we have simply filled in the assumptions explicitly. The command @samp{} now reads a single line from the named file-handle into the default variable @samp{$_}. To make this program more general, we can elimiate the defaults entirely. @smallexample #!/local/bin/perl open (HANDLE,"$ARGV[1]"); while ($line=) @{ print $line; @} @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node == and eq, chop, Simplest perl program, Files in perl @subsection == and @samp{eq} Be careful to distinguish between the comparison operator for integers @samp{==} and the corresponding operator for strings @samp{eq}. These do not work in each other's places so if you get the wrong comparison operator your program might not work and it is quite difficult to find the error. @c ..................................................... @c SUBSECTION @c ..................................................... @node chop, , == and eq, Files in perl @subsection chop @cindex Truncating strings in perl @cindex Perl, truncating strings @cindex @samp{chop} command in perl The command @samp{chop} cuts off the last character of a string. This is useful for removing newline characters when reading files etc. The syntax is @smallexample chop; # chop $_; chop $scalar; # remove last character in $scalar @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Perl subroutines, die, Files in perl, Perl @section Perl subroutines @cindex Subroutines in perl Subroutines are indicated, as in the example above, by the ampersand @samp{&} symbol. When parameters are passed to a Perl subroutine, they are handed over as an array called @samp{@@_}. Which is analogous to the @samp{$_} variable. Here is a simple example: @smallexample #!/local/bin/perl $a="silver"; $b="gold"; &PrintArgs($a,$b); # end of main sub PrintArgs @{ ($local_a,$local_b) = @@_; print "$local_a, $local_b\n"; @} @end smallexample @cindex Local variables in perl @cindex Parameters in perl functions @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node die, stat(), Perl subroutines, Perl @section die - exit on error @cindex Exiting on errors in perl @cindex @samp{die} @cindex Errors in perl When a program has to quit and give a message, the @samp{die} command is normally used. If called without an argument, Perl generates its own message including a line number at which the error occurred. To include your own message, you write @smallexample die "My message...."; @end smallexample @noindent If the string is terminated with a @samp{\n} newline character, the line number of the error is not printed, otherwise Perl appends the line number to your string. When opening files, it is common to see the syntax: @smallexample open (filehandle,"Filename") || die "Can't open..."; @end smallexample @noindent The logical @samp{OR} symbol is used, because @samp{open} returns true if all goes well, in which case the right hand side is never evaluated. If @samp{open} is false, then die is executed. You can decide for yourself whether or not you think this is good programming style -- we mention it here because it is common practice. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node stat(), Perl examples, die, Perl @section The @code{stat()} idiom The unix library function @code{stat()} is used to find out information about a given file. This function is available both in C and in Perl. In perl, it returns an array of values. Usually we are interested in knowing the access permissions of a file. @code{stat()} is called using the syntax @smallexample @@array = stat ("@var{filename}"); @end smallexample @noindent or alternatively, using a named array @smallexample ($device,$inode,$mode) = stat("@var{filename}"); @end smallexample The value returned in the @var{mode} variable is a bit-pattern, @xref{Protection bits}. The most useful way of treating these bit patterns is to use octal numbers to interpret their meaning. To find out whether a file is readable or writable to a group of users, we use a programming idiom which is very common for dealing with bit patterns: first we define a mask which zeroes out all of the bits in the mode string except those which we are specifically interested in. This is done by defining a mask value in which the bits we want are set to 1 and all others are set to zero. Then we AND the mask with the mode string. If the result is different from zero then we know that all of the bits were also set in the mode string. As in C, the bitwise AND operator in perl is called @samp{&}. For example, to test whether a file is writable to other users in the same group as the file, we would write the following. @smallexample $mask = 020; # Leading 0 means octal number ($device,$inode,$mode) = stat("@var{file}"); if ($mode & $mask) @{ print "File is writable by the group\n"; @} @end smallexample @noindent Here the 2 in the second octal number means "write", the fact that it is the second octal number from the right means that it refers to "group". Thus the result of the if-test is only true if that particular bit is true. We shall see this idiom in action below. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Perl examples, Pattern matching, stat(), Perl @section Perl example programs @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Passwd example program:: * Fork example:: * Database example:: @end menu @node Passwd example program, Fork example, Perl examples, Perl examples @subsection The passwd program and @samp{crypt()} function Here is a simple implementation of the UNIX @samp{passwd} program in Perl. @cindex @file{passwd} file @cindex Encryption @cindex @samp{stty} and switching off term echo @cindex Terminal echo and @samp{stty} @vindex @code{crypt()} @smallexample #!/local/bin/perl # # A perl version of the passwd program. # # Note - the real passwd program needs to be much more # secure than this one. This is just to demonstrate the # use of the crypt() function. # ############################################################# print "Changing passwd for $ENV@{'USER'@} on $ENV@{'HOST'@}\n"; system 'stty','-echo'; print "Old passwd: "; $oldpwd = ; chop $oldpwd; ($name,$coded_pwd,$uid,$gid,$x,$y,$z,$gcos,$home,$shell) = getpwnam($ENV@{"USER"@}); if (crypt($oldpwd,$coded_pwd) ne $coded_pwd) @{ print "\nPasswd incorrect\n"; exit (1); @} $oldpwd = ""; # Destroy the evidence! print "\nNew passwd: "; $newpwd = ; print "\nRepeat new passwd: "; $rnewpwd = ; chop $newpwd; chop $rnewpwd; if ($newpwd ne $rnewpwd) @{ print "\n Incorrectly typed. Password unchanged.\n"; exit (1); @} @vindex @code{rand()} $salt = rand(); $new_coded_pwd = crypt($newpwd,$salt); print "\n\n$name:$new_coded_pwd:$uid:$gid:$gcos:$home:$shell\n"; @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node Fork example, Database example, Passwd example program, Perl examples @subsection Example with @samp{fork()} The following example uses the @samp{fork} function to start a daemon which goes into the background and watches the system to which process is using the greatest amount of CPU time each minute. A pipe is opened from the BSD @samp{ps} command. @vindex @code{fork()} @cindex Forking new processes @smallexample #!/local/bin/perl # # A fork() demo. This program will sit in the background and # make a list of the process which uses the maximum CPU average # at 1 minute intervals. On a quiet BSD like system this will # normally be the swapper (long term scheduler). # $true = 1; $logfile="perl.cpu.logfile"; print "Max CPU logfile, forking daemon...\n"; if (fork()) @{ exit(0); @} while ($true) @{ open (logfile,">> $logfile") || die "Can't open $logfile\n"; open (ps,"/bin/ps aux |") || die "Couldn't open a pipe from ps !!\n"; $skip_first_line = ; $max_process = ; close(ps); print logfile $max_process; close(logfile); sleep 60; ($a,$b,$c,$d,$e,$f,$g,$size) = stat($logfile); if ($size > 500) @{ print STDERR "Log file getting big, better quit!\n"; exit(0); @} @} @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node Database example, , Fork example, Perl examples @subsection Example reading databases @cindex Database support @cindex User database support Here is an example program with several of the above features demonstrated simultaneously. This following program lists all users who have home directories on the current host. If the home area has sub-directories, corresponding to groups, then this is specified on the command line. The word `home' causes the program to print out the home directories of the users. @smallexample #!/local/bin/perl ################################################################## # # allusers - list all users on named host, i.e. all # users who can log into this machine. # # Syntax: allusers group # allusers mygroup home # allusers myhost group home # # NOTE : This command returns only users who are registered on # the current host. It will not find users which cannot # be validated in the passwd file, or in the named groups # in NIS. It assumes that the users belonging to # different groups are saved in subdirectories of # /home/hostname. # ################################################################## &arguments(); die "\n" if ( ! -d "/home/$server" ); $disks = `/bin/ls -d /home/$server/$group`; foreach $home (split(/\s/,$disks)) @{ open (LS,"cd $home; /bin/ls $home |") || die "allusers: Pipe didn't open"; while () @{ $exists = ""; ($user) = split; ($exists,$pw,$uid,$gid,$qu,$cm,$gcos,$dir)=getpwnam($user); if ($exists) @{ if ($printhomes) @{ print "$dir\n"; @} else @{ print "$user\n"; @} @} @} close(LS); @} ######################################################## sub arguments @{ $printhomes = 0; $group = "*"; $server = `/bin/hostname`; chop $server; foreach $arg (@@ARGV) @{ if (substr($arg,0,1) eq "u") @{ $group = $arg; next; @} if ($arg eq "home") @{ $printhomes = 1; next; @} $server= $arg; #default is to interpret as a server. @} @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Pattern matching, search and replace, Perl examples, Perl @section Pattern matching and extraction Perl has regular expression operators for identifying patterns. The operator @smallexample /@var{regular expression}/ @end smallexample @noindent returns true of false depending on whether the regular expression matches the contents of @code{$_}. For example @smallexample if (/perl/) @{ print "String contains perl as a substring"; @} if (/(Sat|Sun)day/) @{ print "Weekend day...."; @} @end smallexample The effect is rather like the @code{grep} command. To use this operator on other variables you would write: @smallexample $variable =~ /@var{regexp}/ @end smallexample Regular expression can contain parenthetic sub-expressions, e.g. @smallexample if (/(Sat|Sun)day (..)th (.*)/) @{ $first = $1; $second = $2; $third = $3; @} @end smallexample @noindent in which case perl places the objects matched by such sub-expressions in the variables @code{$1}, @code{$2} etc. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node search and replace, WWW example, Pattern matching, Perl @section Searching and replacing text @cindex @samp{sed} as a perl script @cindex Pattern matching in perl @cindex Pattern replacement in perl @noindent The @samp{sed}-like function for replacing all occurances of a string is easily implemented in Perl using @smallexample while () @{ s/$search/$replace/g; print output; @} @end smallexample @noindent This example replaces the string inside the default variable. To replace in a general variable we use the operator @samp{=~}, with syntax: @smallexample $variable =~ s/@var{search}/@var{replace}/ @end smallexample Here is an example of some of this operator in use. The following is a program which searches and replaces a string in several files. This is useful program indeed for making a change globally in a group of files! The program is called `file-replace'. @smallexample #!/local/bin/perl ############################################################## # # Look through files for findstring and change to newstring # in all files. # ############################################################## # # Define a temporary file and check it doesn't exist # $outputfile = "tmpmarkfind"; unlink $outputfile; # # Check command line for list of files # if ($#ARGV < 0) @{ die "Syntax: file-replace [file list]\n"; @} print "Enter the string you want to find (Don't use quotes):\n\n:"; $findstring=; chop $findstring; print "Enter the string you want to replace with (Don't use quotes):\n\n:"; $replacestring=; chop $replacestring; # print "\nFind: $findstring\n"; print "Replace: $replacestring\n"; print "\nConfirm (y/n) "; $y = ; chop $y; if ( $y ne "y") @{ die "Aborted -- nothing done.\n"; @} else @{ print "Use CTRL-C to interrupt...\n"; @} # # Now shift default array @@ARGV to get arguments 1 by 1 # while ($file = shift) @{ if ($file eq "file-replace") @{ print "Findmark will not operate on itself!"; next; @} # # Save existing mode of file for later # ($dev,$ino,$mode)=stat($file); open (INPUT,$file) || warn "Couldn't open $file\n"; open (OUTPUT,"> $outputfile") || warn "Can't open tmp"; $notify = 1; while () @{ if (/$findstring/ && $notify) @{ print "Fixing $file...\n"; $notify = 0; @} s/$findstring/$replacestring/g; @cindex Searching and replacing in perl (example) print OUTPUT; @} close (OUTPUT); # # If nothing went wrong (if outfile not empty) # move temp file to original and reset the # file mode saved above # @vindex -z in perl if (! -z $outputfile) @{ rename ($outputfile,$file); @vindex @samp{rename} in perl chmod ($mode,$file); @} else @{ print "Warning: file empty!\n."; @} @} @end smallexample @noindent Similarly we can search for lines containing a string. Here is the grep program written in perl @smallexample #!/local/bin/perl # # grep as a perl program # # Check arguments etc while (<>) @{ print if (/$ARGV[1]/); @} @end smallexample @noindent The operator @samp{/@var{search-string}/} returns true if the search string is a substring of the default variable @code{$_}. To search an arbitrary string, we write @smallexample .... if (@var{teststring} =~ /@var{search-string}/); @end smallexample @noindent Here @var{teststring} is searched for occurrances of @var{search-string} and the result is true if one is found. @cindex Pattern matching in perl In perl you can use regular expressions to search for text patterns. Note however that, like all regular expression dialects, perl has its own conventions. For example the dollar sign does not mean "match the end of line" in perl, instead one uses the @samp{\n} symbol. Here is an example program which illustrates the use of regular expressions in perl: @smallexample #!/local/bin/perl # # Test regular expressions in perl # # NB - careful with \ $ * symbols etc. Use '' quotes since # the shell interprets these! # open (FILE,"regex_test"); $regex = $ARGV[$#ARGV]; print "Looking for $ARGV[$#ARGV] in file...\n"; while () @{ if (/$regex/) @{ print; @} @} # # Test like this: # # regex '.*' - prints every line (matches everything) # regex '.' - all lines except those containing only blanks # (. doesn't match ws/white-space) # regex '[a-z]' - matches any line containing lowercase # regex '[^a-z]' - matches any line containg something which is # not lowercase a-z # regex '[A-Za-z]' - matches any line containing letters of any kind # regex '[0-9]' - match any line containing numbers # regex '#.*' - line containing a hash symbol followed by anything # regex '^#.*' - line starting with hash symbol (first char) # regex ';\n' - match line ending in a semi-colon # @end smallexample @noindent Try running this program with the test data on the following file which is called @file{regex_test} in the example program. @smallexample # A line beginning with a hash symbol JUST UPPERCASE LETTERS just lowercase letters Letters and numbers 123456 123456 A line ending with a semi-colon; Line with a comment # COMMENT... @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node WWW example, WWW example 2, search and replace, Perl @section Example: convert mail to WWW pages Here is an example program which you could use to automatically turn a mail message of the form @smallexample From: Newswire To: Mail2html Subject: Nothing happened On the 13th February at kl. 09:30 nothing happened. No footprints were found leading to the scene of a terrible murder, no evidence of a struggle .... etc etc @end smallexample @noindent into an html-file for the world wide web. The program works by extracting the message body and subject from the mail and writing html-commands around these to make a web page. The subject field of the mail becomes the title. The other headers get skipped, since the script searches for lines containing the sequence "colon-space" or @samp{: }. A regular expression is used for this. @smallexample #!/local/bin/perl # # Make HTML from mail # &BeginWebPage(); &ReadNewMail(); &EndWebPage(); ########################################################## sub BeginWebPage @{ print "\n"; print "\n"; @} ########################################################## sub EndWebPage @{ print "\n"; print "\n"; @} ########################################################## sub ReadNewMail @{ while (<>) @{ if (/Subject:/) # Search for subject line @{ # Extract subject text... chop; ($left,$right) = split(":",$_); print "

$right

\n"; next; @} elsif (/.*: .*/) # Search for - anything: anything @{ next; # skip other headers @} print; @} @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node WWW example 2, Other functions, WWW example, Perl @section Generate WWW pages automagically The following program scans through the password database and build a standardized html-page for each user it finds there. It fills in the name of the user in each case. Note the use of the @samp{<<} operator for extended input, already used in the context of the shell, @xref{Pipes and redirection}. This allows us to format a whole passage of text, inserting variables at strategic places, and avoid having to the @code{print} over many lines. @smallexample #!/local/bin/perl # # Build a default home page for each user in /etc/passwd # # #################################################################### # Level 0 (main) #################################################################### $true = 1; $false = 0; # First build an associated array of users and full names setpwent(); while ($true) @{ ($name,$passwd,$uid,$gid,$quota,$comment,$fullname) = getpwent; $FullName@{$name@} = $fullname; print "$name - $FullName@{$name@}\n"; last if ($name eq ""); @} print "\n"; # Now make a unique filename for each page and open a file foreach $user (sort keys(%FullName)) @{ next if ($user eq ""); print "Making page for $user\n"; $outputfile = "$user.html"; open (OUT,"> $outputfile") || die "Can't open $outputfile\n"; &MakePage; close (OUT); @} #################################################################### # Level 1 #################################################################### sub MakePage @{ print OUT < $FullName@{$user@}'s Home Page

$FullName@{$user@}'s Home Page

Hi welcome to my home page. In case you hadn't got it yet my name is: $FullName@{$user@}... I study at Høgskolen i Oslo. ENDMARKER @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Other functions, Perl summary, WWW example 2, Perl @section Other supported functions Perl has very many functions which come directly from the C library. To give a taster, a few are listed here. The Perl book contains a comprehensive description of these. @table @emph @item Fork The standard UNIX fork command for spawning new processes. @item Sockets Support for network socket communication. @item Directories Directory opening and handling routines. @item Databases Reading from the password files and the host databases is supported through the standard C functions @samp{getpasswdbyname} etc. dressed up to look like Perl. @item Crypt The password encryption function. @item Regexp Regular expressions and pattern matching, search and replace functions as in @samp{sed}. @item Operators Perl has the full set of C's logical operators. @item File testing Tests from the shell like @samp{if (-f file)}. @end table @noindent Here are some of the most frequently used functions @table @code @item chmod Change the file mode of a file. e.g. @code{chmod 755,@var{filename}} @item chdir Change the current working directory. e.g. @code{chdir /etc} @item stat Get info about permissions, ownership and type of a file. @item open Open a file for reading, @samp{>} writing, @samp{|} as a pipe. @item close Close an open file handle. @item system Execute a shell command as a child process. e.g. @code{system "ls";} @item split Split a string variable into an array of elements, by searching for a special character (space or @samp{:} etc.) e.g. @code{@@array=split(":",$string)}. @item rename Rename a file. e.g. @code{rename @var{old name} @var{new-name}} @item mkdir Make a new directory. @code{mkdir @var{newdir}} @item shift Read the first element of an array and delete it, shifting all the array elements down by one. (e.g. @code{$first=shift(@@array);)}. @item chop Chops off the last character of a string. Often used for deleting the end-of-line character when reading from a file. @item oct Interprets a number as octal (converts to decimal). e.g. @code{$decimal = oct(755);} @item kill Send a kill signal to a list of processes. e.g. @code{kill -9, @var{pid1},@var{pid2}...} @end table @noindent You should explore Perl's possibilities yourself. Perl is a good alternative to the shell which has much of the power of C and is therefore ideal for simple and more complex system programming tasks. If you intend to be a system administrator for UNIX systems, you could do much worse than to read the Perl book and learn Perl inside out. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Perl summary, Ex5, Other functions, Perl @section Summary The Practical Extraction and Report Language is a powerful tool which goes beyond shell programming, but which retains much of the immediateness of shell programming in a more formal programming environment. The success of Perl has led many programmers to use it exclusively. In the next section, I would like to argue that programming directly in C is not much harder. In fact it has advantages in the long run. The power of Perl is that it is as @emph{immediate} as shell programming. If you are inexperienced, Perl is a little easier than C because many features are ready programmed into the language, but with time one also builds up a repertoire of C functions which can do the same tricks. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Ex5, Proj5, Perl summary, Perl @section Exercises @enumerate @item Write a progam which prints out all of its arguments alphabetically together with the first and the last, and the number of arguments. @item Write a program which prints out the pathname of the home directory for a given user. The user's login name should be given as an argument. @item Write a program called `search-replace' which looks for a given string in a list of files and replaces it with a new string. You should be able to specify a list of files using ordinary unix wildcards. e.g. @samp{search-replace search-string replace-string *.text}. This is a dangerous operation! What if the user types the strings incorrectly? How can you may the program safer? @item Write a program which opens a pipe from @samp{ps} and computes the total cpu-time used by each user. Print the results in order of maximum to minimum. Hint: use an associated array to store the information. @item Write a program which forks and goes into the background. Make the program send you mail when some other user of your choice logs in. Use sleep to check only every few minutes. @item Open a pipe from @samp{find} and collect statistics over how many files there are in all of your sub-directories. @end enumerate @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Proj5, , Ex5, Perl @section Project Write a program which checks the `sanity' of your UNIX system. @enumerate @item Check that the password file /etc/passwd is not writable by general users. @item Check that the processes @samp{cron} and @samp{sendmail} are running. @item Check that, if the file @samp{/etc/exports} or @file{/etc/dfs/dfstab} exists, the @emph{nfsd} daemon is running. @item Check that if the filesystem table @samp{/etc/fstab} (or its equivalent on non-BSD systems) contains NFS mounted filesystems, the @samp{biod} or @samp{nfsiod} daemon is running. @item Check that the file @samp{/etc/resolv.conf} contains the correct domain name. It may or may not be the same as that returned by the shell command @samp{domainname}. If it is not the same, you should print the message `NIS domain has different name to DNS domain'. @end enumerate @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node WWW and CGI programming, C programming, Perl, Top @chapter WWW and CGI programming CGI stands for the Common Gateway Interface. It is the name given to scripts which can be executed from within pages of the world wide web. Although it is possible to use any language in CGI programs (hence the word `common'), the usual choice is Perl, because of the ease with which Perl can handle text. The CGI interface is pretty unintelligent, in order to be as general as possible, so we need to do a bit of work in order to make scripts work. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Permissions:: * Protocols:: * HTML coding:: * Perl and the web:: * PHP and the web:: @end menu @node Permissions, Protocols, WWW and CGI programming, WWW and CGI programming @section Permissions The key thing about the WWW which often causes a lot of confusion is that the W3 service runs with a user ID of `nobody'. The purpose of this is to ensure that nobody has the right to read or write files unless they are opened very explicitly by the user who owns them. In order for files to be readable on the WWW, they must have file mode @samp{644} and they must lie in a directory which has mode @samp{755}. In order for a CGI program to be executable, it must have permission @samp{755} and in order for such a program to write to a file in a user's directory, it must be possible for the file to be created (if necessary) and everyone must be able to write to it. That means that files which are written to by the WWW must have mode @samp{666} and must either exist already or lie in a directory with permission @samp{777}@footnote{You could also set the sticky bit @samp{1777} in order to prevent malicious users from deleting your file.}. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Protocols, HTML coding, Permissions, WWW and CGI programming @section Protocols CGI script programs communicate with W3 browsers using a very simple protocol. It goes like this: @itemize @bullet @item A web page sends data to a script using the `forms' interface. Those data are concatenated into @emph{a single line}. The data in separate fields of a form are separated by @samp{&} signs. New lines are replaced by the text @samp{%0D%0A}, which is the DOS ASCII representation of a newline, and spaces are replaced by @samp{+} symbols. @item A CGI script reads this single line of text on the standard input. @item The CGI script replies to the web browser. The first line of the reply @emph{must} be a line which is tells the browser what mime-type the data are sent in. Usually, a CGI script replies in HTML code, in which case the first line in the reply must be: @smallexample Content-type: text/html @end smallexample @cindex CGI protocol @cindex Mime types in W3 @noindent This must be followed by a blank line. @end itemize @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node HTML coding, Perl and the web, Protocols, WWW and CGI programming @section HTML coding of forms To start a CGI program from a web page we use a @emph{form} @cindex Forms in HTML which is a part of the HTML code enclosed with the parentheses @smallexample
...
@end smallexample @noindent The method `post' means that the data which get typed into this form will be piped into the CGI program via its standard input. The `action' specifies which program you want to start. Note that @cindex Script aliases in W3 you cannot simply use the absolute path of the file, for security reasons. You must use something called a `script alias' to tell the web browser where to find the program. If you do not have a script alias defined for you personally, then you need to get one from your system administrator. By using a script alias, no one from outside your site can see where your files are located, only that you have a `cgi-bin' area somewhere on your system. Within these parentheses, you can arrange to collect different kinds of input. The simplest kind of input is just a button which starts the CGI program. This has the form @smallexample @end smallexample @noindent This code creates a button. When you click on it the program in your `action' string gets started. More generally, you will want to create input boxes where you can type in data. To create a single line input field, you use the following syntax: @smallexample @end smallexample @noindent This creates a single line text field of width 40 characters. This is not the limit on the length of the string which can be typed into the field, only a limit on the amount which is visible at any time. It is for visual formatting only. The NAME field is used to identify the data in the CGI script. The string you enter here will be sent to the CGI script in the form @samp{variable-name=value of input...}. Another type of input is a text area. This is a larger box where one can type in text on several lines. The syntax is: @smallexample

@end smallexample @noindent The reset button clears the form. When the submit button is pressed, the CGI program is activated. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Perl and the web, PHP and the web, HTML coding, WWW and CGI programming @section Perl and the web @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Interpreting the data:: * A complete guestbook example:: @end menu @node Interpreting the data, A complete guestbook example, Perl and the web, Perl and the web @subsection Interpreting data from forms To interpret and respond to the data in a form, we must write a program which satsifies the protocol above, @xref{Protocols}. We use perl as a script langauge. The simplest valid CGI script is the following: @smallexample #!/local/bin/perl # # Reply with proper protocol # print "Content-type: text/html\n\n"; # # Get the data from the form ... # $input = ; # # ... and echo them back # print $input, "\n Done! \n"; @end smallexample @noindent Although rather banal, this script is a useful starting point for CGI programming, because it shows you just how the input arrives at the script from the HTML form. The data arrive all in a single, enormously long line, full of funny characters. The first job of any script is to decode this line. Before looking at how to decode the data, we should make an important point about the protocol line. If a web browser does not get this `Content-type' line from the CGI script it returns with an error: @example 500 Server Error The server encountered an internal error or misconfiguration and was unable to complete your request. Please contact the server administrator, and inform them of the time the error occurred, and anything you might have done that may have caused the error. Error: HTTPd: malformed header from script www/cgi-bin/comment.pl @end example @noindent Before finishing your CGI script, you will probably ecounter this error several times. A common reason for getting the error is a syntax error in your script. If your program contains an error, the first thing a browser gets in return is not the `Content-type' line, but an error message. The browser does not pass on this error message, it just prints the uninformative message above. If you can get the above script to work, then you are ready to decode the data which are sent to the script. The first thing is to use perl to split the long line into an array of lines, by splitting on @samp{&}. We can also convert all of the @samp{+} symbols back into spaces. The script now looks like this: @smallexample #!/local/bin/perl # # Reply with proper protocol # print "Content-type: text/html\n\n"; # # Get the data from the form ... # $input = ; # # ... and echo them back # print "$input\n\n\n"; $input =~ s/\+/ /g; # # Now split the lines and convert # @@array = split('&',$input); foreach $var ( @@array ) @{ print "$var\n"; @} print "Done! \n"; @end smallexample @noindent We now have a series of elements in our array. The output from this script is something like this: @smallexample variable1=Mark+Burgess&variable2=%0D%0AI+just+called+to+say+ @var{(wrap)} ....%0D%0A...hey+pig%2C+nothing%27s+working+out+the+way+I+planned variable1=Mark Burgess variable2=%0D%0AI just called to say @var{(wrap)} ....%0D%0A...hey pig%2Cnothing%27s working out the way I planned Done! @end smallexample @noindent As you can see, all control characters are converted into the form @samp{%XX}. We should now try to do something with these. Since we are usually not interested in keeping new lines, or any other control codes, we can simply null-out these with a line of the form @smallexample $input =~ s/%..//g; @end smallexample @noindent The regular expression @samp{%..} matches anything beginning with a percent symbol followed by two characters. The resulting output is then free of these symbols. We can then separate the variable contents from their names by splitting the input. Here is the complete code: @smallexample #!/local/bin/perl # # Reply with proper protocol # print "Content-type: text/html\n\n"; # # Get the data from the form ... # $input = ; # # ... and echo them back # print "$input\n\n\n"; $input =~ s/%..//g; $input =~ s/\+/ /g; @@array = split('&',$input); foreach $var ( @@array ) @{ print "$var
"; @} print "

\n"; ($name,$variable1) = split("variable1=",$array[0]); ($name,$variable2) = split("variable2=",$array[1]); print "
var1 = $variable1
"; print "
var2 = $variable2
"; print "
Done! \n"; @end smallexample @noindent and the output @cartouche @smallexample variable1=Mark+Burgess&variable2=%0D%0AI+just+called+to+say @var{(wrap)} +....%0D%0A...hey+pig%2C+nothing%27s+working+out+the+way+I+planned variable1=Mark Burgess variable2=I just called to say .......hey pig nothings working @var{(wrap)} out the way I planned var1 = Mark Burgess var2 = I just called to say .......hey pig nothings working out @var{(wrap)} the way I planned Done! @end smallexample @end cartouche @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node A complete guestbook example, , Interpreting the data, Perl and the web @subsection A complete guestbook example in perl Let us now use this technique to develop a guest book aplication. Based on the code above, analyze the following code. @smallexample #!/local/bin/perl #################################################################### # # Guest book # #################################################################### $guestbook_page = "/iu/nexus/ud/mark/www/tmp/cfguest.html"; $tmp_page = "/iu/nexus/ud/mark/www/tmp/guests.tmp"; $remote_host = $ENV@{"REMOTE_HOST"@}; print "Content-type: text/html\n\n"; print "

\n"; print "Thank you for submitting your comment!

\n"; print "best wishes,

"; print "-Mark


"; print "Return to menu\n"; $input = ; $input =~ s/%..//g; $input =~ s/\+/ /g; @@array = split('&',$input); ($skip,$name) = split("var1=",$array[0]); ($skip,$message) = split("var2=",$array[1]); if (! open (PAGE, $guestbook_page)) @{ print "Content-type: text/html\n\n"; print "couldn't open guestbook page file!"; @} if (! open (TMP, "+>$tmp_page")) @{ print "Content-type: text/html\n\n"; print "couldn't open temporary output file!"; @} while ($line = ) @{ if ($line =~ /

Number of entries: (..)/) @{ $entry_no = $1; $entry_no++; $line = "

Number of entries: $entry_no

\n"; @} if ($line =~ //) @{ $date = `date +"%A, %b %d %Y"`; print TMP "Entry $date from host: $remote_host\n

\n"; print TMP "From: $name\n

\n"; print TMP $message; print TMP "\n

\n"; @} print TMP "$line"; @} close PAGE; close TMP; if (! rename ($tmp_page, $guestbook_page)) @{ print "Oops! Rename operation failed!\n"; @} chmod (0600, $guestbook_page); @end smallexample This script works by reading through the old guest book file, opening a new copy of the guest book file and appending a new messages at the end. The end of the message section (not counting the @samp{} tags) is marked by a comment line. @smallexample @end smallexample @noindent Note that a provisional guest book file has to exist in the first place. The script writes to a new file and then swaps the new file for the old one. The guest book file looks something like this: @smallexample Comments

My guest book

Entry no. Wednesday, Feb 28 1996 from host: dax

From: Mark.Burgess@@iu.hioslo.no

Just to start the ball rolling....

Entry no. Tuesday, Mar 26 1996 from host: enterprise.subspace.net

From: spock@@enterprise

Registering a form of energy never before encountered.

Mark Burgess - Mark.Burgess@@iu.hioslo.no @end smallexample The directory in which this file lies needs to be writable to the user nobody (the WWW user) and the files within need to be deletable by nobody but no one else. Some users try to make guest book scripts setuid-themselves in order to overcome the problem that httpd runs with uid nobody, but this opens many security issues. In short it is asking for trouble. Unfortunately an ordinary user cannot use @code{chown} in order to give access only to the WWW user nobody, so this approach needs the cooperation of the system administrator. Nevertheless this is the most secure approach. Try to work through this example step for step. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node PHP and the web, , Perl and the web, WWW and CGI programming @section PHP and the web The PHP 3 language makes the whole business of web programming rather simpler than perl. It hides the business of translating variables from forms into new variables in a CGI program and it even allows you to embed active code into you HTML pages. PHP has special support for querying data in an SQL database like MySQL or Oracle. PHP documentation lives at @uref{http://www.php.net}. @menu * Embedded PHP:: * PHP and forms:: @end menu @node Embedded PHP, PHP and forms, PHP and the web, PHP and the web @subsection Embedded PHP PHP code can be embedded inside HTML pages provided your WWW server is configurered with PHP support. PHP code lives inside a tag with the general form @smallexample @end smallexample @noindent For example, we could use this to import one file into another and print out a table of numbers: @smallexample "; @} ?> @end smallexample This makes it easy to generate WWW pages with a fixed visual layout: @smallexample \n"; print ""; print "

"$title

"; print "$comment"; print "
\n"; include $contents; print ("
\n"); print ("\n"); print ("\n"); @end smallexample @noindent Variables are easily set by calling PHP code in the form of a CGI program from a form. @node PHP and forms, , Embedded PHP, PHP and the web @subsection PHP and forms PHP is particularly good at dealing with forms, as a CGI scripting langauge. Consider the following form: @smallexample
Name:
Email:
Preferred language:
@end smallexample This produces a page into which one types a name and email address and chooses a language from a list of three possible choices. When the user clicks on a button marked by the file @file{image.gif} the form is posted. Here is a program which unravels the data sent to the CGI program: @smallexample #!/local/bin/php "; echo "

$title

"; echo "$comment"; echo "
\n"; ### echo "Your name is $personal[name]

"; echo "Your email is $personal[email]

"; echo "Language options: "; echo " "; for ($i = 0; strlen($language[$i]) > 0; $i++) @{ echo ""; @} if ($language[0] == "Norwegian") @{ echo "Hei alle sammen

"; @} else @{ echo "Greetings everyone, this page will be in English

"; @} echo "

Variable language[$i] = $language[$i]
"; ### echo ("
\n"); echo ("\n"); echo ("\n"); ?> @end smallexample @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node C programming, Network programming, WWW and CGI programming, Top @chapter C programming @emph{This section is not meant to teach you C. It is a guide to using C in UNIX and it is assumed that you have a working knowledge of the language. See the GNU C-Tutorial for an introduction to basics.} @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Shell or C?:: * C program structure:: * UNIX calls and POSIX:: * Compiling C:: * Make:: * argv argc envp:: * Environment variables in C:: * C files and dirs:: * File test macros:: * Process control:: * A more secure popen:: * Traps and signals:: * Regular expressions in C:: * DES encryption:: * ioctl:: * Database example in C:: * lex and yacc:: * Ex 6:: @end menu @node Shell or C?, C program structure, C programming, C programming @section Shell or C? @cindex C programming In the preceding chapters we have been looking at ways to get simple programming tasks done. The immediateness of the script languages is a great advantage when we just want to get a job done as quickly as possible. Scripts lend themselves to simple system administration tasks like file processing, but they do not easily lend themselves to more serious programs. @cindex Wrappers Although some system administrators have grown to the idea that shell programming is easier, I would argue that this is not really true. First of all, most of the UNIX shell commands are just wrapper programs for C function calls. Why use the wrapper when you can use the real thing? Secondly, the C function calls return data in pointers and structures which are very easy to manipulate, whereas piping the output of shell programs into others can be a very messy and awkward way of working. Here are some of the reasons why we also need a more traditional programming language like C. @enumerate @item The shell languages do not allow us to create an acceptable user-interface, like X-windows, or the curses (cursor manipulation) library. They are mainly intended for file-processing. (Though recently the @emph{Tk} library has provided a way of creating user interfaces in Tcl and Perl.) @cindex Curses @cindex X windows @cindex Tk library @item Shell commands read their input line-by-line. Not all input is generated in this simple way -- we also need to be able to read through lines i.e. the concept of a @emph{data stream}. @item More advanced data structures are needed for most applications, such as linked lists and binary trees, acyclic graphs etc. @item Compilers help to sort out simple typographical and logical errors by compile-time checking source code. @item Compiled code is faster than interpreted code. @item Many tools have been written to help in the programming of C code (dbx, lex, yacc etc.). @cindex @samp{lex} @cindex @samp{yacc} @end enumerate @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node C program structure, UNIX calls and POSIX, Shell or C?, C programming @section C program structure @c ..................................................... @c SUBSECTION @c ..................................................... @menu * C structure:: * Macros and declarations:: * Several files:: @end menu @node C structure, Macros and declarations, C program structure, C program structure @subsection The form of a C program A C program consists of a set of function, beginning with the main program: @smallexample main () /* This is a comment */ @{ Commands ... @} @end smallexample The source code of a C program can be divided into several text files. C compiles all functions separately; the linker @code{ld} joins them all up at the end. This means that we can plan out a strategy for writing large programs in a clear and efficient manner. NOTE: C++ style comments @samp{//...} are not allowed by most C compilers. @c ..................................................... @c SUBSECTION @c ..................................................... @node Macros and declarations, Several files, C structure, C program structure @subsection Macros and declarations Most Unix systems now have ANSI C compatible compilers, but this has not always been the case. Most UNIX programs written in a version of C which is older than the ANSI standard, so you will need an appreciation of old Kernighan and Ritchie C conventions for C programming. See for example my C book. An obvious difference between ANSI C and K&R C is that the C++ additions to the language are not included. Here are some useful points to remember. @cindex ANSI C @cindex Kernighan and Ritchie C @itemize @bullet @item K&R C does not allow @samp{const} data, it uses the C preprocessor with @samp{#define} instead. i.e. intead of @smallexample const int blah = 1; @end smallexample @noindent use @smallexample #define blah 1 @end smallexample @noindent Remember that the hash symbol @samp{#} must be the first character on a line under UNIX. @item K&R C doesn't use function prototypes or declarations of the form: @smallexample void function (char *string, int a, int b) @{ @} @end smallexample @noindent Instead one writes: @smallexample void function (string, a, b) char *string; int a,b; @{ @} @end smallexample @end itemize @c ..................................................... @c SUBSECTION @c ..................................................... @node Several files, , Macros and declarations, C program structure @subsection Several files Most unix programs are very large and are split up into many files. Remember, when you split up programs into several files, you must declare variables as @samp{extern} in file A if they are really declared in file B. in which you want to use them. This tells the compiler that it should not try to create local storage for the variable, because this was already done in another file. @cindex Splitting C into many files. @cindex @code{extern} variables @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node UNIX calls and POSIX, Compiling C, C program structure, C programming @section A note about UNIX system calls and standards @cindex POSIX standard Most of the system calls in UNIX return data in the form of @samp{struct} variables. Sometimes these are structures used by the operating system itself -- in other cases they are just put together so that programmers can handle a packet of data in a convenient way. If in doubt, you can find the definitions of these structures in the relevant include files under @file{/usr/include}. Since UNIX comes in many flavours the system calls are not always compatible and may have different options and arguments. Because of this there is a number of standardizing organizations for UNIX. One of them is POSIX which is an organization run by the major UNIX vendors. Programs written for UNIX are now expected to be POSIX compliant. This is not something you need to think about at the level of this course, but you should certainly remember that there exist programming standards and that these should be adhered to. The aim is to work towards a single standard UNIX. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Compiling C, Make, UNIX calls and POSIX, C programming @section Compiling: @samp{cc}, @samp{ld} and @file{a.out} The C compiler on the unix system is traditionally called @samp{cc} and has always been a traditional part of every Unix environment. Recently several Unix vendors have stopped including the C compiler as a part of their operating systems and instead sell a compiler separately. Fortunately there is a public domain Free Software version of the compiler called @samp{gcc} (the GNU C compiler). We shall use this in all the examples. To compile a program consisting of several files of code, we first compile all of the separate pieces without trying to link them. There are therefore two stages: first we turn @samp{.c} files into @samp{.o} files. This compiles code but does not fix any address references. Then we link all @samp{.o} files into the final executable, including any libraries which are used. @cindex Compiling programs @cindex Multiple C files, compiling Let's suppose we have files @file{a.c}, @file{b.c} and @file{c.c}. We write: @smallexample gcc -c a.c b.c c.c @end smallexample This creates files @samp{a.o}, @samp{b.o} and @samp{c.o}. Next we link them into one file called `myprog'. @smallexample gcc -o myprog a.o b.o c.o @end smallexample @noindent If the naming option @samp{-o myprog} is not used, the link @samp{ld} uses the default name @emph{a.out} for the executable file. @cindex @file{a.out} @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Libraries and LD_LIBRARY_PATH:: * Include files:: * Shared and static libraries:: * Important paths:: @end menu @node Libraries and LD_LIBRARY_PATH, Include files, Compiling C, Compiling C @subsection Libraries and @samp{LD_LIBRARY_PATH} @vindex @samp{LD_LIBRARY_PATH} @cindex libc @cindex libm @cindex libcurses @cindex Library path for C loader @cindex @samp{-L} option to @code{cc} The resulting file is called @samp{myprog} and includes references only to the standard library @samp{libc}. If we wish to link in the math library @samp{libm} or the cursor movement library @samp{libcurses} -- or in general, a library called @samp{libBLAH} , we need to use the @samp{-l} directive. @smallexample gcc -o myprog files.o -lm -lcurses -lBLAH @end smallexample @noindent The compiler looks for a suitable library in all of the directories listed in the environment variable @samp{LD_LIBRARY_PATH}. Alternatively we can add a directory to the search path by using the @samp{-L}. option: @smallexample gcc -o myprog files.o -L/usr/local/lib -lm -lcurses -lBLAH @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node Include files, Shared and static libraries, Libraries and LD_LIBRARY_PATH, Compiling C @subsection Include files @cindex Include files Normally the compiler looks for include files only in the directory @file{/usr/include}. We can add further paths to search using the @samp{-I} option. @smallexample gcc -o myprog file.c -I/usr/local/include -I/usr/local/X11/include @end smallexample @cindex @samp{-I} option to @code{cc} @cindex Include file search path Previously, Unix libraries have been in @file{a.out} code format, but recent releases of unix have gone over to a more efficient and flexible format called ELF (executable and linking format). @c ..................................................... @c SUBSECTION @c ..................................................... @node Shared and static libraries, Important paths, Include files, Compiling C @subsection Shared and static libraries @cindex Static linking @cindex Shared libraries Libraries are collections of C functions which the operating system creators have written for our convenience. The source code for such a library is just the source for a collection of functions -- there is no @code{main} program. There are two kinds of library used by modern operating systems: @emph{archive libraries} or static libraries and @code{shared libraries} or dynamical libraries. An archive library has a name of the form @smallexample lib@var{name}.a @end smallexample @noindent When an archive library is linked to a program, it is appended lock, stock and barrel to the program code. This uses a lot of disk space and makes the size of the compiled program very large. Shared libraries (shared objects @samp{so} or shared archives @samp{sa} generally have names of the form) @smallexample lib@var{name}.so lib@var{name}.sa @end smallexample @noindent often with version numbers appended. When a program is linked with a shared library the code is not appended to the program. Instead pointers to the shared objects are created and the library is loaded at runtime, thus avoiding the problem of having to store the library effectively multiple times on the disk. To make an archive library we compile all of the functions we wish to include in the library @smallexample gcc -c function1.c function2.c ... @end smallexample @noindent and then join the files using the @samp{ar} command. @smallexample ar rcv libMYLIB.a function1.o ar rcv libMYLIB.a function2.o @end smallexample @noindent To make a shared library one provides an option to the linker program. The exact method is different in different operating systems, so you should look at the manual page for @code{ld} on your system. Under SunOS 4 we take the object files @samp{*.o} and run @smallexample ld -o libMYLIB.so.1.1 -assert pure-text *.o @end smallexample Under HPUX, we write @smallexample ld -b -o libMYLIB.so.1.1 *.o @end smallexample With the GNU linker, you write @smallexample ld -shared -o libMYLIB.so.1.1 *.o @end smallexample NOTE: when you add a shared library to the system under SunOS or GNU/Linux you must run the command @samp{ldconfig}, making sure that the path to the library is included in @samp{LD_LIBRARY_PATH}. SunOS and GNU/Linux use a cache file @file{/etc/ld.so.cache} to keep current versions of libraries. GNU/Linux also uses a configuration file called @file{/etc/ld.so.conf}. @cindex @file{ld.so.cache} @cindex @file{ldconfig} @c ..................................................... @c SUBSECTION @c ..................................................... @node Important paths, , Shared and static libraries, Compiling C @subsection Knowing about important paths: directory structure It is important to understand how the C compiler finds the files it needs. We have already mentioned the @samp{-I} and @samp{-L} options to the compilation command line. In general, all system include files can be found in the directory @file{/usr/include} and subdirectories of this directory. All system libraries can be found in @file{/usr/lib}. Many packages build their own libraries and keep the relevant files in separate directories so that if the system gets reinstalled, they do not get deleted. This is true for example of the X-windows system. The include and library files for this are typically kept in directories which look something like @file{/usr/local/X11R5/include} and @file{/usr/X11R6/lib}. That means that we need to give all of this information to the compiler. Compiling a program becomes a complicated task in many cases so we need some kind of script to help us perform the task. The Unix tool @code{make} was designed for this purpose. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Make, argv argc envp, Compiling C, C programming @section Make Nowadays compilers are often sold with fancy user environments driven by menus which make it easier to compile programs. Unix has similar environments but all of them use shell-based command line compilation beneath the surface. That is because UNIX programmers are used to writing large and complex programs which occupy many directories and subdirectories. Each directory has to be adapted or configured to fit the particular flavour of Unix system it is being compiled upon. Interactive user environments are very poor at performing this kind of service. UNIX solves the problem of compiling enormous @emph{trees} of software (such as the unix system itself!) by using a compilation language called @samp{make}. Such language files can be generated automatically by scripts, allowing very complex programs to configure and compile themselves from a single control script. @cindex Compiling huge programs @c ..................................................... @c SUBSECTION @c ..................................................... @menu * Compiling large programs:: * Makefiles:: * Suffix rules for C++:: @end menu @node Compiling large programs, Makefiles, Make, Make @subsection Compiling large projects Typing lines like @smallexample cc -c file1.c file2.c ... cc -o target file1.o .... @end smallexample @noindent repeatedly to compile a complicated program can be a real nuisance. One possibility would therefore be to keep all the commands in a script. This could waste a lot of time though. Suppose you are working on a big project which consists of many lines of source code -- but are editing only one file. You really only want to recompile the file you are working on and then relink the resulting object file with all of the other object files. Recompiling the other files which hadn't changed would be a waste of time. But that would mean that you would have to change the script each time you change what you need to compile. A better solution is to use the @samp{make} command. @samp{make} was designed for precisely this purpose. To use @samp{make}, we create a file called @file{Makefile} in the same directory as our program. @samp{make} is a quite general program for building software. It is not specifically tied to the C programming language--- it can be used in any programming language. A @samp{make} configuration file, called a @file{Makefile}, contains rules which describe how to compile or build all of the pieces of a program. For example, even without telling it specifically, @code{make} knows that in order to go from @samp{prog.c} to @samp{prog.o} the command @samp{cc -c prog.c} must be executed. A Makefile works by making such associations. The Makefile contains a list of all of the files which compose the program and rules as to how to get to the finished product from the source. The idea is that, to compile a program, we just have to type @kbd{make}. @samp{make} then reads the Makefile and compiles all of the parts which need compiling. It does not recompile files which have not changed since the last compilation! How does it do this? @samp{make} works by comparing the time-stamp on the file it needs to create with the time-stamp on the file which is to be compiled. If the compiled version exists and is newer than its source then the source does not need to be recompiled. To make this idea work in practice, @samp{make} has to know how to go through the steps of compiling a program. Some default rules are defined in a global configuration file, e.g. @smallexample /usr/include/make/default.mk @end smallexample @noindent Let's consider an example of what happens for the the three files @file{a.c}, @file{b.c} and @file{c.c} in the example above -- and let's not worry about what the Makefile looks like yet. The first time we compile, only the `.c' files exist. When we type @samp{make}, the program looks at its rules and finds that it has to make a file called `myprog'. To make this it needs to execute the command @smallexample gcc -o myprog a.o b.o c.o @end smallexample @noindent So it looks for `a.o' etc and doesn't find them. It now goes to a kind of subroutine and looks to see if it has any rules for making files called `.o' and it discovers that these are made by compiling with the @samp{gcc -c} option. Since the files do not exist, it does this. Now the files `a.o b.o c.o' exist and it jumps back to the original problem of trying to make `myprog'. All the files it needs now exist and so it executes the command and builds `myprog'. If we now edit `a.c', and type @samp{make} once again -- it goes through the same procedure as before but now it finds all of the files. So it compares the dates on the files -- if the source is newer than the result, it recompiles. By using this recursive method, @samp{make} only compiles those parts of a program which need compiling. @c ..................................................... @c SUBSECTION @c ..................................................... @node Makefiles, Suffix rules for C++, Compiling large programs, Make @subsection Makefiles @cindex Suffix rules in Makefiles @cindex Dependencies in Makefiles To write a Makefile, we have to tell @samp{make} about @emph{dependencies}. The dependencies of a file are all of those files which are required to build it. Thus, the dependencies of `myprog' are `a.o', `b.o' and `c.o'. The dependencies of `a.o' are simply `a.c', the dependencies of `b.o' are `b.c' and so on. A Makefile consists of rules of the form: @smallexample target : dependencies @key{TAB} rule; @end smallexample @noindent The target is the thing we want to build, the dependenices are like subroutines to be executed first if they do not exist. Finally the rule is to be executed if all if the dependenices exist; it takes the dependencies and turns them into the target. There are two important things to remember: @itemize @bullet @item The file names must start on the first character of a line. @item There must be a @key{TAB} character at the beginning of every rule or action. If there are spaces instead of tabs, or no tab at all, @samp{make} will signal an error. This bizarre feature can cause a lot of confusion. @end itemize @noindent Let's look at an example Makefile for a program which consists of two course files @file{main.c} and @file{other.c} and which makes use of a library called @file{libdb} which lies in the directory @file{/usr/local/lib}. Our aim is to build a program called @code{database}: @smallexample # # Simple Makefile for `database' # # First define a macro OBJ = main.o other.o CC = gcc CFLAGS = -I/usr/local/include LDFLAGS = -L/usr/local/lib -ldb INSTALLDIR = /usr/local/bin # # Rules start here. Note that the $@@ variable becomes the name of the # executable file. In this case it is taken from the $@{OBJ@} variable # database: $@{OBJ@} $@{CC@} -o $@@ $@{OBJ@} $@{LDFLAGS@} # # If a header file changes, normally we need to recompile everything. # There is no way that make can know this unless we write a rule which # forces it to rebuild all .o files if the header file changes... # $@{OBJ@}: $@{HEADERS@} # # As well as special rules for special files we can also define a # "suffix rule". This is a rule which tells us how to build all files # of a certain type. Here is a rule to get .o files from .c files. # The $< variable is like $? but is only used in suffix rules. # .c.o: $@{CC@} -c $@{CFLAGS@} $< ####################################################################### # Clean up ####################################################################### # # Make can also perform ordinary shell command jobs # "make tidy" here performs a cleanup operation # clean: rm -f $@{OBJ@} rm -f y.tab.c lex.yy.c y.tab.h rm -f y.tab lex.yy rm -f *% *~ *.o rm -f mconfig.tab.c mconfig.tab.h a.out rm -f man.dvi man.aux man.log man.toc rm -f cfengine.tar.gz cfengine.tar cfengine.tar.Z make tidy rm -f cfengine install: $@{INSTALLDIR@}/database cp database $@{INSTALLDIR@}/database @end smallexample @noindent The Makefile above can be invoked in several ways. @smallexample make make database make clean make install @end smallexample @noindent If we simple type @samp{make} i.e. the first of these choices, @samp{make} takes the first of the rules it finds as the object to build. In this case the rule is `database', so the first two forms above are equivalent. On the other hand, if we type @smallexample make clean @end smallexample @noindent then execution starts at the rule for `clean', which is normally used to remove all files except the original source code. Make `install' causes the compiled program to be installed at its intended destination. @samp{make} uses some special variables (which resemble the special variables used in Perl -- but don't confuse them). The most useful one is @samp{$@@} which represents the current @emph{target} -- or the object which @samp{make} would like to compile. i.e. as @samp{make} checks each file it would like to compile, @samp{$@@} is set to the current filename. @table @code @item $@@ This evaluates to the current target i.e. the name of the object you are currently trying to build. It is normal to use this as the final name of the program when compiling @item $? This is used only outside of suffix rules and means the name of all the files which must be compiled in order to build the current target. @smallexample target: file1.o file2.o @key{TAB} cc -o $@@ $? @end smallexample @item $< This is only used in suffix rules. It has the same meaning as @samp{$?} but only in suffix rules. It stands for the pre-requisite, or the file which must be compiled in order to make a given object. @vindex $@@ in make @vindex $? in make @vindex $< in make @end table Note that, because @samp{make} has some default rules defined in its configuration file, a single-file C program can be compiled very easily by typing @smallexample make filename.c @end smallexample @noindent This is equivalent to @smallexample cc -c filename.c cc -o filename filename.o @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node Suffix rules for C++, , Makefiles, Make @subsection New suffix rules for C++ Standard rules for C++ are not often built into UNIX systems at the time of writing, but we can create them in our own Makefiles very easily. Here we shall use the GNU compiler g++'s conventions for C++ files. Here is a sample Makefile for using C++. Note that the @samp{.SUFFIXES} command must be used to declare new endings or file extensions. @cindex C++ suffix rules @cindex Make rules for C++ @smallexample ################################################################## # # This is the Makefile for g++ # ################################################################## OBJ = cpp-prog.o X.o Y.o Z.o CCPLUS = g++ .SUFFIXES: .C .o .h # # Program Rules # filesys: $@{OBJ@} $(CCPLUS) -o filesys $(OBJ) # # Extra dependencies on the header file # (if the header file changes, we need to rebuild *.o) # cpp-prog.o: filesys.h X.o: filesys.h Y.o: filesys.h Z.o: filesys.h # # Suffix rules # .C.o: $(CCPLUS) -c $< @end smallexample The general rule here tells @code{make} that a @file{.o} file can be created from a @file{.C} file by executing the command @samp{$(CCPLUS) -c}. (This is identical to the C case, exept for the name of the compiler). The extra dependencies tell @code{make} that, if we change the header file @file{filesys.h}, then we must recompile all the files which read in @file{filesys.h}, since this could affect all of these. Finally, the highest level rule says that to make @file{filesys} from the @file{.o} files, we have to run @samp{$(CCPLUS) -o filesys *.o}. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node argv argc envp, Environment variables in C, Make, C programming @section The @code{argv}, @code{argc} and @code{envp} paramters When we write C programs which reads command line arguments, they are fed to us as an array of strings called the argument vector. The mechanisms for the C-shell and Perl are derived from the C argument vector. To read in the command line, we write @cindex The arguement vector in C @cindex Command line arguments in C @vindex @code{argv} in C @vindex @code{argc} in C @cindex @code{envp} in C @cindex Environment variables in C @smallexample main (argc,argv,envp) int argc; char *argv[], *envp[]; @{ printf ("The first argument was %s\n",argv[1]); @} @end smallexample @noindent Argument zero is the name of the program itself and @samp{argv[argc-1]} is the last argument. The above definitions are in Kernighan and Ritchie C style. In ANSI C, the arguments can be declared using prototype: @smallexample main (int argc, char **argv) @{ @} @end smallexample @noindent The array of strings @samp{envp[]} is a list of values of the @emph{environment variables} of the system, formatted by @smallexample NAME=value @end smallexample @noindent This gives C programmers access to the shell's global environment. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Environment variables in C, C files and dirs, argv argc envp, C programming @section Environment variables in C @cindex Environment variables in C @cindex @code{getenv() function} @vindex @code{getenv()} In addition to the @samp{envp} vector, it is possible to access the environment variables through the call @samp{getenv()}. This is used as follows; suppose we want to access the shell environment variable @samp{$HOME}. @smallexample char *string; string = getenv("HOME"); @end smallexample @noindent @samp{string} is now a pointer to static but public data. You should not use @samp{string} as if it were you're own property because it will be used again by the system. Copy it's contents to another string before using the data. @smallexample char buffer[500]; strcpy (buffer,string); @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node C files and dirs, File test macros, Environment variables in C, C programming @section Files and directories All of the regular C functions from the standard library are available to Unix programmers. The standard functions only address the issue of reading and writing to files however, they do not deal with operating system specific attributes such as file permissions and file types. Nor is there a mechanisms for obtaining lists of files within a directory. The reason for these omissions is that they are operating system dependent. To find out about these other attributes POSIX describes some standard Unix system calls. @c ..................................................... @c SUBSECTION @c ..................................................... @menu * opendir and readir:: * stat:: * lstat and readlink:: @end menu @node opendir and readir, stat, C files and dirs, C files and dirs @subsection opendir, readdir @cindex @code{opendir} command @cindex @code{closedir} command @cindex @code{readdir} command @cindex dirent directory interface Files and directories are handled by functions defined in the header file @file{dirent.h}. In earlier UNIX systems the file @file{dir.h} was used -- and the definitions were slightly different, but not much. To get a list of files in a directory we must open the directory and read from it -- just like a file. (A directory is just a file which contains data on its entries). The commands are @smallexample opendir closedir readdir @end smallexample @noindent See the manual pages for dirent. These functions return pointers to a @code{dirent} structure which is defined in the file @file{/usr/include/dirent.h}. Here is an example @code{ls} command which lists the contents of the directory @file{/etc}. This header defines a structure @smallexample struct dirent @{ off_t d_off; /* offset of next disk dir entry */ unsigned long d_fileno; /* file number of entry */ unsigned short d_reclen; /* length of this record */ unsigned short d_namlen; /* length of string in d_name */ char d_name[255+1]; /* name (up to MAXNAMLEN + 1) */ @}; @end smallexample @noindent which can be used to obtain information from the directory nodes. @smallexample #include #include main () @{ DIR *dirh; struct dirent *dirp; static char mydir[20] = "/etc"; if ((dirh = opendir(mydir)) == NULL) @{ perror("opendir"); return; @} for (dirp = readdir(dirh); dirp != NULL; dirp = readdir(dirh)) @{ printf("Got dir entry: %s\n",dirp->d_name); @} closedir(dirh); @} @end smallexample @noindent Notice that reading from a directory is like reading from a file with @code{fgets()}, but the entries are filenames rather than lines of text. @c ..................................................... @c SUBSECTION @c ..................................................... @node stat, lstat and readlink, opendir and readir, C files and dirs @subsection @code{stat()} @cindex Information about file properties @cindex Statistics about a file @cindex @code{stat()} @cindex @code{lstat()} To determine the file properties or @emph{statistics} we use the function call @samp{stat()} or its corollory @samp{lstat()}. Both these functions find out information about files (permissions, owner, filetype etc). The only difference between them is the way in which they treat symbolic links. If @samp{stat} is used on a symbolic link, it stats the file the link points to rather than the link itself. If @samp{lstat} is used, the data refer to the link. Thus, to detect a link, we must use @samp{lstat}, @xref{lstat and readlink}. The data in the `stat' structure are defined in the file @file{/usr/include/sys/stat.h}. Here are the most important structures. @smallexample struct stat @{ dev_t st_dev; /* device number*/ ino_t st_ino; /* file inode */ mode_t st_mode; /* permission */ short st_nlink; /* Number of hardlinks to file */ uid_t st_uid; /* user id */ gid_t st_gid; /* group id */ dev_t st_rdev; off_t st_size; /* size in bytes */ time_t st_atime; /* time file last accessed */ time_t st_mtime; /* time file contents last modified */ time_t st_ctime; /* time last attribute change */ long st_blksize; long st_blocks; @}; @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @node lstat and readlink, , stat, C files and dirs @subsection lstat and readlink @cindex @code{readlink()} @cindex Links, where do they point? @cindex Links in C The function @samp{stat()} treats symbolic links as though they were the files they point to. In other words, if we use @samp{stat()} to read a symbolic link, we end up reading the file the link points to and not the link itself--- we never see symbolic links. To avoid this problem, there is a different version of the stat function called @samp{lstat()} which is identical to @samp{stat()} except that it treats links as links and @emph{not} as the files they point to. This means that we can test whether a file is a symbolic link, only if we use @samp{lstat()}. (See the next paragraph.) Once we have identified a file to be a symbolic link, we use the @samp{readlink()} function to obtain the name of the file the link points to. @smallexample #define bufsize 512 char buffer[bufsize]; readlink("/path/to/file",buffer,bufsize); @end smallexample @noindent The result is returned in the string buffer. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node File test macros, Process control, C files and dirs, C programming @section @code{stat()} test macros @cindex Macros for stat @cindex Permissions, determining in C @cindex File type, determining in C As we have already mentioned, the Unix mode bits contain not only information about what permissions a file has, but also bits describing the type of file -- whether it is a directory or a link etc. There are macros defined in UNIX to extract this information from the @samp{st_mode} member of the @samp{stat} structure. They are defined in the @samp{stat.h} headerfile. Here are some examples. @smallexample #define S_ISBLK(m) /* is block device */ #define S_ISCHR(m) /* is character device */ #define S_ISDIR(m) /* is directory */ #define S_ISFIFO(m) /* is fifo pipe/socket */ #define S_ISREG(m) /* is regular (normal) file */ #define S_ISLNK(m) /* is symbolic link */ /* Not POSIX */ #define S_ISSOCK(m) /* is a lock */ #define S_IRWXU /* rwx, owner */ #define S_IRUSR /* read permission, owner */ #define S_IWUSR /* write permission, owner */ #define S_IXUSR /* execute/search permission, owner */ #define S_IRWXG /* rwx, group */ #define S_IRGRP /* read permission, group */ #define S_IWGRP /* write permission, grougroup */ #define S_IXGRP /* execute/search permission, group */ #define S_IRWXO /* rwx, other */ #define S_IROTH /* read permission, other */ #define S_IWOTH /* write permission, other */ #define S_IXOTH /* execute/search permission, other */ @end smallexample These return true or false when acting on the mode member. Here is an example @xref{readdir example}. @smallexample struct stat statvar; stat("file",&statvar); /* test return values */ if (S_ISDIR(statvar.st_mode)) @{ printf("Is a directory!"); @} @end smallexample @c ..................................................... @c SUBSECTION @c ..................................................... @menu * readdir example:: @end menu @node readdir example, , File test macros, File test macros @subsection Example filing program The following example program demonstrates the use of the directory functions in @code{dirent} and the @code{stat} function call. @smallexample /********************************************************************/ /* */ /* Reading directories and `statting' files */ /* */ /********************************************************************/ #include #include #include #include #define DIRNAME "/." #define bufsize 255 /********************************************************************/ main () @{ DIR *dirh; struct dirent *dirp; struct stat statbuf; char *pathname[bufsize]; char *linkname[bufsize]; if ((dirh = opendir(DIRNAME)) == NULL) @{ perror("opendir"); exit(1); @} for (dirp = readdir(dirh); dirp != NULL; dirp = readdir(dirh)) @{ if (strcmp(".",dirp->d_name) == 0 || strcmp("..",dirp->d_name) == 0) @{ continue; @} if (strcmp("lost+found",dirp->d_name) == 0) @{ continue; @} sprintf(pathname,"%s/%s",DIRNAME,dirp->d_name); if (lstat(pathname,&statbuf) == -1) /* see man stat */ @{ perror("stat"); continue; @} if (S_ISREG(statbuf.st_mode)) @{ printf("%s is a regular file\n",pathname); @}; if (S_ISDIR(statbuf.st_mode)) @{ printf("%s is a directory\n",pathname); @} if (S_ISLNK(statbuf.st_mode)) @{ bzero(linkname,bufsize); /* clear string */ readlink(pathname,linkname,bufsize); printf("%s is a link to %s\n",pathname,linkname); @} printf("The mode of %s is %o\n\n",pathname,statbuf.st_mode & 07777); @} closedir(dirh); @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Process control, A more secure popen, File test macros, C programming @section Process control, @code{fork()}, @code{exec()}, @code{popen()} and system There is a number of ways in which processes can interact with one another and in which we can control their behaviour. We shall not go into great detail in this course, only provide examples for reference. The UNIX @samp{fork()} function is used to create child processes. This is the basis of all `heavyweight' multitasking under unix. Here is a simple example of fork in which we start a child process from within a program and wait for it to finish. Note that the code for the parent and the child is is the same file. The only thing that distinguishes parent from child is the value returned by the fork function. When @samp{fork()} is called, it duplicates the entire current process so that two parallel processes are then running. The only difference between these is that the child process (the copy) gets a return value of zero from @samp{fork()}, whereas the parent gets a return value equal to the process identifier (pid) of the child. This value can be used by the parent to send messages or to wait for the child. Here we show a simple example in which the @samp{wait(NULL)} command is used to wait for the last child spawned by the parent. @smallexample /**************************************************************/ /* */ /* A brief demo of the UNIX process duplicator fork(). */ /* */ /**************************************************************/ #include /***************************************************************/ main () @{ int pid, cid; pid = getpid(); printf ("Fork demo! I am the parent (pid = %d)\n",pid); if (! fork()) @{ cid = getpid(); printf ("I am the child (cid = %d) of (pid=%d)\n",cid,pid); ChildProcess(); exit(0); @} printf("Parent waiting here for the child...\n"); wait(NULL); printf("Child finished, parent quitting too!\n"); @} /**************************************************************/ ChildProcess() @{ int i; for (i = 0; i < 10; i++) @{ printf ("%d...\n",i); sleep(1); @} @} @end smallexample Another possibility is that we might want to execute a program and wait to find out what the result of the program is before continuing. There are two ways to do this. The first is a variation on the theme above and uses @code{fork()}. Let's create a function which runs a shell command from within a C program, and determines its return value. We make the result a boolean (integer) value, so that the function returns `true' if the shell command exits normally, @xref{Return codes}. @smallexample if (ShellCommandReturnsZero(@var{shell-command})) @{ printf ("Command %s went ok\n",@var{shell-command}); @} @end smallexample @noindent To do this we first have to fork a new process and then use one of the @code{exec} commands to load a new code image on top of the new process. @cindex Starting shell commands from C This sounds complicated, but it is necessary because of the way unix handles processes. If we had no use for the return value, we could simply execute a shell command using the @code{system("shell command")} function, (which does all this for us) but when @code{system()} exits, we can only tell if the command was executed successfully or unsuccessfully---we learn nothing about what actually failed (the shell or command which was executed under the shell?) If we require detailed information about what happened to the child process then we need to do the following. @smallexample #include #include /* Send complete command as a string */ /* including all arguments */ ShellCommandReturnsZero(comm) char *comm; @{ int status, i, argc; pid_t pid; char arg[maxshellargs][bufsize]; char **argv; /* Build argument array for execv call*/ for (i = 0; i < maxshellargs; i++) @{ bzero (arg[i],bufsize); @} argc = SplitCommand(comm,arg); if ((pid = fork()) < 0) @{ FatalError("Failed to fork new process"); @} else if (pid == 0) /* child */ @{ argv = malloc((argc+1)*sizeof(char *)); for (i = 0; i < argc; i++) @{ argv[i] = arg[i]; @} argv[i] = (char *) NULL; if (execv(arg[0],argv) == -1) @{ yyerror("script failed"); perror("execvp"); exit(1); @} @} else /* parent */ @{ if (wait(&status) != pid) @{ printf("Wait for child failed\n"); perror("wait"); return false; @} else @{ if (WIFSIGNALED(status)) @{ printf("Script %s returned: %s\n",comm,WTERMSIG(status)); return false; @} if (! WIFEXITED(status)) @{ return false; @} if (WEXITSTATUS(status) == 0) @{ return true; @} else @{ return false; @} @} @} @} /*******************************************************************/ SplitCommand(comm,arg) char *comm, arg[maxshellargs][bufsize]; @{ char *sp; int i = 0, j; char buff[bufsize]; for (sp = comm; *sp != NULL; sp++) @{ bzero(buff,bufsize); if (i >= maxshellargs-1) @{ yyerror("Too many arguments in embedded script"); FatalError("Use a wrapper"); @} while (*sp == ' ' || *sp == '\t') @{ sp++; @} switch (*sp) @{ case '\"': sscanf (++sp,"%[^\"]",buff); break; case '\'': sscanf (++sp,"%[^\']",buff); break; default: sscanf (sp,"%s",buff); break; @} for (j = 0; j < bufsize; j++) @{ arg[i][j] = buff[j]; @} sp += strlen(arg[i]); i++; @} return (i); @} @end smallexample @noindent In this example, the script waits for the exit signal from the child process before continuing. The return value from the child is available from the wait function with the help of a set of macros defined in @file{/usr/include/sys/wait.h}. The value is given by @code{WTERMSIG(status)}. @cindex Waiting for child processes @cindex @file{wait.h} @cindex @code{WTERMSIG(status)} In the final example, we can open a pipe to a process directly in a C program as though it were a file, by using the function @code{popen()}. @cindex @code{popen()} @cindex Pipes in C @cindex Opening a pipe in C Pipes may be opened for reading or for writing, in exactly the same way as a file is opened. The child process is automatically synchronized with the parent using this method. Here is a program which opens a unix command for reading (both stdout and stderr) from the child process are piped into the program. Notice that the syntax used in this call is that used by the Bourne shell, since this is build deeply into the unix execution design. @smallexample #define bufsize 1024 FILE *pp; char VBUFF[bufsize]; ... if ((pp = popen( "/sbin/mount -va 2<&1","r")) == NULL) @{ printf("Failed to open pipe\n"); return errorcode; @} while (!feof(pp)) @{ fgets(VBUFF,bufsize,pp); /* Just write the output to stdout */ printf ("Pipe read: %s\n",VBUFF); @} pclose(pp); @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node A more secure popen, Traps and signals, Process control, C programming @section A more secure @code{popen()} One problem with the @code{popen()} system call is that it uses a shell to execute the command it obtains a pipe to. In the past this has been used to allow Unix security breaches, using a so-called @code{IFS} attack which can trick the shell into executing a program with the name of the first node in the directory of the executable. For instance,if the pipe was to open the program @file{/bin/ps}, this coudl be tricked into executing a program in the current working directory of the process called @file{bin} with argument @samp{ps}. The solution is not to use a shell at all, but to replace @code{popen()} with a version which calls @code{exec()} directly. Here is a safe version from the source code of cfengine: @smallexample #define bufsize 4096 #define maxshellargs 20 pid_t *CHILD; int MAXFD = 20; /* Max number of simultaneous pipes */ /***************************************************************/ FILE *cfpopen(command, type) char *command, *type; @{ char arg[maxshellargs][bufsize]; int i, argc, pd[2]; char **argv; pid_t pid; FILE *pp = NULL; if ((*type != 'r' && *type != 'w') || (type[1] != '\0')) @{ errno = EINVAL; return NULL; @} if (CHILD == NULL) /* first time */ @{ if ((CHILD = calloc(MAXFD,sizeof(pid_t))) == NULL) @{ return NULL; @} @} if (pipe(pd) < 0) /* Create a pair of descriptors to this process */ @{ return NULL; @} if ((pid = fork()) == -1) @{ return NULL; @} if (pid == 0) @{ switch (*type) @{ case 'r': close(pd[0]); /* Don't need output from parent */ if (pd[1] != 1) @{ dup2(pd[1],1); /* Attach pp=pd[1] to our stdout */ dup2(pd[1],2); /* Merge stdout/stderr */ close(pd[1]); @} break; case 'w': close(pd[1]); if (pd[0] != 0) @{ dup2(pd[0],0); close(pd[0]); @} @} for (i = 0; i < MAXFD; i++) @{ if (CHILD[i] > 0) @{ close(CHILD[i]); @} argc = SplitCommand(command,arg); argv = (char **) malloc((argc+1)*sizeof(char *)); if (argv == NULL) @{ FatalError("Out of memory"); @} for (i = 0; i < argc; i++) @{ argv[i] = arg[i]; @} argv[i] = (char *) NULL; if (execv(arg[0],argv) == -1) @{ sprintf(OUTPUT,"Couldn't run %s",arg[0]); CfLog(cferror,OUTPUT,"execv"); @} _exit(1); @} @} else @{ switch (*type) @{ case 'r': close(pd[1]); if ((pp = fdopen(pd[0],type)) == NULL) @{ return NULL; @} break; case 'w': close(pd[0]); if ((pp = fdopen(pd[1],type)) == NULL) @{ return NULL; @} @} CHILD[fileno(pp)] = pid; return pp; @} @} /***************************************************************/ cfpclose(pp) FILE *pp; @{ int fd, status; pid_t pid; Debug("cfpclose(pp)\n"); if (CHILD == NULL) /* popen hasn't been called */ @{ return -1; @} fd = fileno(pp); if ((pid = CHILD[fd]) == 0) @{ return -1; @} CHILD[fd] = 0; if (fclose(pp) == EOF) @{ return -1; @} Debug("cfpopen - Waiting for process %d\n",pid); #ifdef HAVE_WAITPID while(waitpid(pid,&status,0) < 0) @{ if (errno != EINTR) @{ return -1; @} @} return status; #else if (wait(&status) != pid) @{ return -1; @} else @{ if (WIFSIGNALED(status)) @{ return -1; @} if (! WIFEXITED(status)) @{ return -1; @} return (WEXITSTATUS(status)); @} #endif @} /*******************************************************************/ /* Command exec aids */ /*******************************************************************/ SplitCommand(comm,arg) char *comm, arg[maxshellargs][bufsize]; @{ char *sp; int i = 0, j; char buff[bufsize]; for (sp = comm; sp < comm+strlen(comm); sp++) @{ bzero(buff,bufsize); if (i >= maxshellargs-1) @{ CfLog(cferror,"Too many arguments in embedded script",""); FatalError("Use a wrapper"); @} while (*sp == ' ' || *sp == '\t') @{ sp++; @} switch (*sp) @{ case '\0': return(i-1); case '\"': sscanf (++sp,"%[^\"]",arg[i]); break; case '\'': sscanf (++sp,"%[^\']",arg[i]); break; case '`': sscanf (++sp,"%[^`]",arg[i]); break; default: sscanf (sp,"%s",arg[i]); break; @} sp += strlen(arg[i]); i++; @} return (i); @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Traps and signals, Regular expressions in C, A more secure popen, C programming @section Traps and signals Processes can receive signals from the UNIX kernel at any time. Some of these signals terminate the execution of the program. This can cause problems if the program is in the middle of critical activity such as writing to a file. For that reason we can trap signals and provide our own routine for handling them in a special way. A signal handler is made by calling the function @samp{signal()} for each signal and by specifying a pointer to a function which will be called in the event of a signal. For example: @smallexample main () @{ int HandleSignal(); signal(SIGTERM,HandleSignal); @} HandleSignal() @{ /* Tidy up and exit cleanly */ exit(0); @} @end smallexample @samp{SIGTERM} is the usual signal sent by the command @samp{kill}. There are many other signals which can be sent to programs. Here is list. You have to decide for yourself whether or not you want to provide your own signal handling function. To ignore a signal, you write @smallexample signal(SIG@var{type},SIG_IGN); @end smallexample @noindent To remove a signal handler and re-activate a signal, you write @smallexample signal(SIG@var{type},SIG_DFL); @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Regular expressions in C, DES encryption, Traps and signals, C programming @section Regular expressions A regular expression is a pattern for matching strings of text. We have met regular expressions earlier in connection with the shell and Perl. Naturally these earlier encounters have their roots in C functions for handling expressions. A regular expression is used by first `compiling' it into a convenient data structure. Then a matching function is used to compare the expression with a test string. In this example program we show how a regular expression typed in as an argument to the program is found within strings of input entered on the keyboard. @smallexample #include #include main (argc,argv) int argc; char **argv; @{ char buffer[1024]; regex_t rx; regmatch_t match; size_t nmatch = 1; if (regcomp(&rx, argv[1], REG_EXTENDED) != 0) @{ perror("regcomp"); return; @} while (!feof(stdin)) @{ fgets(buffer,1024,stdin); if (regexec(&rx,buffer,1,&match,0) == 0) @{ printf("Matched:(%s) at %d to %d",buffer,match.rm_so,match.rm_eo); @} @} regfree(&rx); @} @end smallexample Here is an example of its use. The output of the program is in @emph{italics} @smallexample % a.out xyz this is a string another string an xyz string @emph{Matched: (an xyz string ) at 3 to 6} another xyz zyxxyz string @emph{Matched: (another xyz xyz string ) at 8 to 11} % a.out 'xyz|abc' This is a string An abc string @emph{Matched: (An abc string ) at 3 to 6} Or an xyz string @emph{Matched: (Or an xyz string ) at 6 to 9} @end smallexample If you don't want the match data set @code{&pm} to @code{NULL}. To get an exact match rather than a substring check that the bounds are 0 and @code{strlen(argv[1])-1}. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node DES encryption, ioctl, Regular expressions in C, C programming @section DES encryption Encryption with the SSLeay library, compile with command @smallexample gcc crypto.c -I/usr/local/ssl/include -L/usr/local/ssl/lib -lcrypto @end smallexample Example of normal triple DES encryption which works only on an 8-byte buffer: @smallexample /*****************************************************************************/ /* */ /* File: crypto.c */ /* */ /* Compile with: gcc program.c -lcrypto (SSLeay) */ /* */ /*****************************************************************************/ #include #include #define bufsize 1024 /* Note how this truncates to 8 characters */ main () @{ char in[bufsize],out[bufsize],back[bufsize]; des_cblock key1,key2,key3,seed = @{0xFE,0xDC,0xBA,0x98,0x76,0x54,0x32,0x10@}; des_key_schedule ks1,ks2,ks3; strcpy(in,"1 2 3 4 5 6 7 8 9 a b c d e f g h i j k"); des_random_seed(seed); des_random_key(key1); des_random_key(key2); des_random_key(key3); des_set_key((C_Block *)key1,ks1); des_set_key((C_Block *)key2,ks2); des_set_key((C_Block *)key3,ks3); des_ecb3_encrypt((C_Block *)in,(C_Block *)out,ks1,ks2,ks3,DES_ENCRYPT); printf("Encrypted [%s] into [%s]\n",in,out); des_ecb3_encrypt((C_Block *)out,(C_Block *)back,ks1,ks2,ks3,DES_DECRYPT); printf("and back to.. [%s]\n",back); @} @end smallexample Triple DES, chaining mode, for longer strings (which must be a multiple of 8 bytes): @smallexample /*****************************************************************************/ /* */ /* File: crypto.c */ /* */ /* Compile with: gcc program.c -lcrypto (SSLeay) */ /* */ /*****************************************************************************/ #include #include #define bufsize 1024 /* This can be used on arbitrary length buffers */ main () @{ char in[bufsize],out[bufsize],back[bufsize],workvec[bufsize]; des_cblock key1,key2,key3,seed = @{0xFE,0xDC,0xBA,0x98,0x76,0x54,0x32,0x10@}; des_key_schedule ks1,ks2,ks3; strcpy(in,"1 2 3 4 5 6 7 8 9 a b c d e f g h i j k l m n o p q r s t u v w x y z"); des_random_seed(seed); des_random_key(key1); des_random_key(key2); des_random_key(key3); des_set_key((C_Block *)key1,ks1); des_set_key((C_Block *)key2,ks2); des_set_key((C_Block *)key3,ks3); /* This work vector can be intialized t anything ...*/ memset(workvec,0,bufsize); des_ede3_cbc_encrypt((C_Block *)in,(C_Block *)out,(long)strlen(in), ks1,ks2,ks3,(C_Block *)workvec,DES_ENCRYPT); printf("Encypted [%s] into [something]\n",in); /* .. but this must be initialized the same as above */ memset(workvec,0,bufsize); /* Note that the length is the original length, not strlen(out) */ des_ede3_cbc_encrypt((C_Block *)out,(C_Block *)back,(long)strlen(in), ks1,ks2,ks3,(C_Block *)workvec,DES_DECRYPT); printf("and back to.. [%s]\n",back); @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node ioctl, Database example in C, DES encryption, C programming @section Device control: ioctl @cindex @code{ioctl()} @vindex ioctl() The C function @samp{ioctl} (I/O control) is used to send special control commands to devices like the disk and the network interface. The syntax of the function is @smallexample int ioctl(fd, request, arg) int fd, request; long arg; @end smallexample @noindent The first parameter is normally as device handle or socket descriptor. The second is a control parameter. Lists of valid control parameters are normally defined in the system `include' files for a particular device. They are device and system dependent so you need a local manual and som detective work to find out what they are. The final parameter is a pointer to a variable which receives return data from the device. @samp{ioctl} commands are device specific, by their nature. The commands for the ethernet interface device are only partially standardized, for example. We could read the ethernet device (which is called `le0' on a Sun workstation), using the following command: @smallexample # include /* Typical includes for internet */ # include # include # include # include # include # include # include struct ifreq IFR; int sk; struct sockaddr_in sin; strcpy(IFR.ifr_name,"le0"); IFR.ifr_addr.sa_family = AF_INET; if ((sk = socket(AF_INET,SOCK_DGRAM,IPPROTO_IP)) == -1) @{ perror("socket"); exit(1); @} if (ioctl(sk,SIOCGIFFLAGS, (caddr_t) &IFR) == -1) @{ perror ("ioctl"); exit(1); @} @end smallexample We shall not go into the further details of @samp{ioctl}, but simply note its role in system programming. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Database example in C, lex and yacc, ioctl, C programming @section Database example (Berkeley db) @smallexample DBT key,value; DB *dbp; DBC *dbcp; db_recno_t recno; if ((errno = db_open(CHECKSUMDB,DB_BTREE, DB_CREATE, 0664, NULL, NULL, &dbp)) != 0) @{ sprintf(OUTPUT,"cfd: couldn't open checksum database %s\n",CHECKSUMDB); CfLog(cferror,OUTPUT,"db_open"); return false; @} bzero(&value,sizeof(value)); bzero(&key,sizeof(key)); key.data = filename; key.size = strlen(filename)+1; value.data = dbvalue; value.size = sizeof(dbvalue); if ((errno = dbp->del(dbp,NULL,&key,0)) != 0) @{ CfLog(cferror,"","db_store"); @} key.data = filename; key.size = strlen(filename)+1; if ((errno = dbp->put(dbp,NULL,&key,&value,0)) != 0) @{ CfLog(cferror,"put failed","db->put"); @} if ((errno = dbp->get(dbp,NULL,&key,&value,0)) == 0) @{ /* Not found ... */ return; @} dbp->close(dbp,0); @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node lex and yacc, Ex 6, Database example in C, C programming @section Text parsing tools: @samp{lex} and @samp{yacc} @cindex Lexer @cindex Parser @cindex @samp{yacc} @emph{This section is a taster only. You only need to know what lex and yacc are, not how they work.} @samp{lex} and @samp{yacc} are two tools for the C programmer who wishes to make a text parser. A text parser is a program which reads a text file and interprets the symbols in it. Every programming language must include a text parser, for instance. The @samp{yacc} (yet another compiler compiler) program generates C code which parses a textfile, given a description of the syntax rules for the file. In other words, we define the logical structure of the text file, according to the way we wish to interpret it and give the rules to @samp{yacc}. @samp{yacc} produces C code from this which does the job. @samp{lex} is a `lexer'. It is normally used together with @samp{yacc}. @samp{lex} @emph{tokenizes} or identifies symbols in a file. What that means is that it reads in a file and matches @emph{types} of string in the file which are defined in terms of regular expressions by the programmer, and returns symbolic values for those strings. Although @samp{lex} can be used by independently of @samp{yacc}, it is normally used to identify the different @emph{types} of string which define the syntax of a file. For example, suppose @samp{yacc} was parsing a C program. On the beginning of a line, it might expect to find either a variable name or a preprocessor symbol. A variable name is just a string consisting of characters from the set @samp{0-9a-Z_}, whereas a preprocessor command always starts with the character @samp{#}. @samp{yacc} passes control to @samp{lex} which reads the file and matches the first object on the line. If it finds a variable, it returns to @samp{yacc} a @emph{token} which is a number or value corresponding to `variable'. Similarly, if it finds a preprocessor command, it returns a token for that. If it doesn't match either type it returns something else and @samp{yacc} signals a syntax error. Here is a @samp{yacc} file which parses a file consisting of lines of the form a+b, where $a$ and $b$ are numbers -- any other syntax is incorrect. We could have used this later in the example program for the client-server example, @xref{Sockets}. You can learn more about lex and yacc in @emph{"Lex and Yacc", J. Levine, T. Mason and D. Brown, O'Reilly and Assoc.} @smallexample %@{ /*******************************************************************/ /* */ /* PARSER for a + b protocol */ /* */ /* The section between the single %'s gets copied verbatim into */ /* the resulting C code yacc generates -- including this comment! */ /* */ /*******************************************************************/ #include extern char *yytext; %@} %token NUMBER PLUS %% specification: @{ yyerror("Warning: invalid statement");@} | statement; statement: NUMBER PLUS NUMBER; @end smallexample @noindent The lexer to go with this parser generates the tokens NUMBER and PLUS used by @samp{yacc}: @smallexample %@{ /*******************************************************************/ /* */ /* LEXER for a + b protocol */ /* */ /* Returns token types NUMBER and PLUS to yacc, one at a time */ /* */ /*******************************************************************/ #include "y.tab.h" /* yacc produces this -- need this line! */ %@} number [0-9]+ plus [+] %% number @{ return NUMBER; @} plus @{ return PLUS; @} . @{ return yytext[0]; @} %% /* EOF */ @end smallexample The main program which uses @samp{yacc} and @samp{lex} looks like this: @smallexample extern FILE *yyin; main () @{ if ((yyin = fopen("My_Input_File","r")) == NULL) /* Open file */ @{ printf("Can't open file\n"); exit (1); @} while (!feof(yyin)) @{ yyparse(); @} fclose (yyin); @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Ex 6, , lex and yacc, C programming @section Exercises @enumerate @item Write a daemon program with a signal handler which makes a log of the heaviest (maximum cpu) process running, every five minutes. The program should exit if the log file becomes greater than 5-kbytes. @item Rewrite in C the perl program which lists all the files in the current directory containing a certain string. @item Write a version of `more' which prints control characters safely. See the @samp{cat -e} command. @item Write a Makefile to create a shared library from a number of object files. @end enumerate @c ********************************************************************** @c CHAPTER @c ********************************************************************** @node Network programming, Programming idioms, C programming, Top @chapter Network Programming Client-server communication is the basis of modern operating system technology. The Unix socket mechanism makes stream-based communication virtually transparent. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @menu * Sockets:: * Multithreading a server:: * System databases:: * DNS:: * C and NFS:: * Ex 7:: @end menu @node Sockets, Multithreading a server, Network programming, Network programming @section Socket streams Analogous to filestreams are @emph{sockets} or TCP/IP @emph{network connections}. A socket is a two-way (read/write) pseudo-file node. An open socket stream is like an open file-descriptor. Berkeley sockets are part of the standard C library. There are two main kinds of socket: TCP/IP sockets and Unix domain sockets. Unix sockets can be used to provide local interprocess communication using a filestream communication protocol. TCP/IP sockets open file descriptors across the network. @cindex Sockets @cindex TCP/IP @cindex @code{INADDR_ANY} A TCP/IP socket is a file stream associated with an IP address and a port number. We write to a socket descriptor just as with a file descriptor, either with @code{write()} or using @code{send()}. When sending binary data over a network we have to be careful about machine level representations of data. Operating systems (actually the hardware they run on) fall into two categories known as @emph{big endian} and @emph{little endian}. The names refer to the @emph{byte-order} of numerical representations. @cindex Byte order @cindex Network byte order @cindex Big endian @cindex Little endian The names indicate how large integers (which require say 32 bits or more) are stored in memory. Little endian systems store the least significant byte first, while big endian systems store the most significant byte first. For example, the representation of the number 34,677,374 has either of these forms. @smallexample ----------------------------------- Big | 2 | 17 | 34 | 126 | ----------------------------------- ----------------------------------- Little | 126 | 34 | 17 | 2 | ----------------------------------- @end smallexample Obviously if we are transferring data from one host to another, both hosts have to agree on the data representation otherwise there would be disastrous consequences. This means that there has to be a common standard of @emph{network byte ordering}. For example, Solaris (SPARC hardware) uses network byte ordering (big endian), while GNU/Linux (Intel hardware) uses the opposite (little endian). This means that Intel systems have to convert the format every time something is transmitted over the network. Unix systems provide generic functions for converting between host-byteorder and network-byteorder for small and long integer data: @smallexample htonl, htons, ntohl, ntohs @end smallexample @noindent Here we list two example programs which show how to make a client-server pair. The server enters a loop, and listens for connections from any clients (the generic address @samp{INADDR_ANY} is a wildcard for any address on the current local network segment). The client program sends requests to the server as a protocol in the form of a string of the type @samp{a + b}. Normally @samp{a} and @samp{b} are numbers, in which case the server returns their sum to the client. If the message has the special form @samp{halt + *}, where the star is arbitrary, then the server shuts down. Any other form of message results in an error, which the server signals to the client. The basic structure of the client-server components in terms of system calls is this: @smallexample Client: socket() @var{Create a socket} connect() @var{Contact a server socket (IP + port)} while (?) @{ send() @var{Send to server} recv() @var{Receive from server} @} Server: socket() @var{Create a socket} bind() @var{Associates the socket with a fixed address} listen() @var{Create a listen queue} while() @{ reply=accept() @var{Accept a connection request} recv() @var{Receive from client} send() @var{Send to client} @} @end smallexample @smallexample /**********************************************************************/ /* */ /* The client part of a client-server pair. This simply takes two */ /* numbers and adds them together, returning the result to the client */ /* */ /* Compiled with: */ /* cc server.c */ /* */ /* User types: */ /* 3 + 5 */ /* a + b */ /* halt + server */ /**********************************************************************/ #include #include #include #include #include #define PORT 9000 /* Arbitrary non-reserved port */ #define HOST "nexus.iu.hioslo.no" #define bufsize 20 /**********************************************************************/ /* Main */ /**********************************************************************/ main (argc,argv) int argc; char *argv[]; @{ struct sockaddr_in cin; struct hostent *hp; char buffer[bufsize]; int sd; if (argc != 4) @{ printf("syntax: client a + b\n"); exit(1); @} @cindex @code{gethostbyname()} if ((hp = gethostbyname(HOST)) == NULL) @{ perror("gethostbyname: "); exit(1); @} memset(&cin,0,sizeof(cin)); /* Another way to zero memory */ cin.sin_family = AF_INET; cin.sin_addr.s_addr = ((struct in_addr *)(hp->h_addr))->s_addr; cin.sin_port = htons(PORT); printf("Trying to connect to %s = %s\n",HOST,inet_ntoa(cin.sin_addr)); @cindex @code{socket()} if ((sd = socket(AF_INET,SOCK_STREAM,0)) == -1) @{ perror("socket"); exit(1); @} @cindex @code{connect()} if (connect(sd,&cin,sizeof(cin)) == -1) @{ perror("connect"); exit(1); @} sprintf(buffer,"%s + %s",argv[1],argv[3]); @cindex @code{send()} if (send(sd,buffer,strlen(buffer),0) == -1) @{ perror ("send"); exit(1); @} @cindex @code{recv()} if (recv(sd,buffer,bufsize,0) == -1) @{ perror("recv"); exit (1); @} printf ("Server responded with %s\n",buffer); close (sd); unlink("./socket"); @} @end smallexample @smallexample /**********************************************************************/ /* */ /* The server part of a client-server pair. This simply takes two */ /* numbers and adds them together, returning the result to the client */ /* */ /* Compiled with: */ /* cc server.c */ /* */ /**********************************************************************/ #include #include #include #include #include #define PORT 9000 #define bufsize 20 #define queuesize 5 #define true 1 #define false 0 /**********************************************************************/ /* Main */ /**********************************************************************/ main () @{ struct sockaddr_in cin; struct sockaddr_in sin; struct hostent *hp; char buffer[bufsize]; int sd, sd_client, addrlen; memset(&sin,0,sizeof(sin)); /* Another way to zero memory */ sin.sin_family = AF_INET; sin.sin_addr.s_addr = INADDR_ANY; /* Broadcast address */ sin.sin_port = htons(PORT); if ((sd = socket(AF_INET,SOCK_STREAM,0)) == -1) @{ perror("socket"); exit(1); @} @cindex @code{bind()} if (bind(sd,&sin,sizeof(sin)) == -1) /* Must have this on server */ @{ perror("bind"); exit(1); @} @cindex @code{listen()} if (listen(sd,queuesize) == -1) @{ perror("listen"); exit(1); @} @cindex @code{accept()} while (true) @{ if ((sd_client = accept(sd,&cin,&addrlen)) == -1) @{ perror("accept"); exit(1); @} if (recv(sd_client,buffer,sizeof(buffer),0) == -1) @{ perror("recv"); exit(1); @} if (!DoService(buffer)) @{ break; @} if (send(sd_client,buffer,strlen(buffer),0) == -1) @{ perror("send"); exit(1); @} close (sd_client); @} close (sd); printf("Server closing down...\n"); @} /**************************************************************/ DoService(buffer) char *buffer; /* This is the protocol section. Here we must */ /* check that the incoming data are sensible */ @{ int a=0,b=0; printf("Received: %s\n",buffer); sscanf(buffer,"%d + %d\n",&a,&b); if (a > 0 && b> 0) @{ sprintf(buffer,"%d + %d = %d",a,b,a+b); return true; @} else @{ if (strncmp("halt",buffer,4) == 0) @{ sprintf(buffer,"Server closing down!"); return false; @} else @{ sprintf(buffer,"Invalid protocol"); return true; @} @} @} @end smallexample @noindent In the example we use `streams' to implement a typical input/output behaviour for C. A stream interface is a so-called reliable protocol. @cindex Reliable socket protocol There are other kinds of sockets too, called @emph{unrealiable}, or UDP sockets. Features to notice on the server are that we must bind to a specific address. The client is always implicitly bound to an address since a socket connection always originates from the machine on which the client is running. On the server however we want to know which addresses we shall be receiving requests from. In the above example we use the generic wildcard address @samp{INADDR_ANY} which means that any host can connect to the server. Had we been more specific, we could have limited communication to two machines only. By calling @samp{listen()} we set up a queue for incoming connections. Rather than forking a separate process to handle each request we set up a queue of a certain depth. If we exceed this depth then new clients rtying to connect will be refused connection. The @samp{accept} call is the mechanism which extracts a `reply handle' from the socket. Using the handle obtained from this call we can reply to the client without having to open a special socket explicitly. An improved server side connection can be setup, reading the service name from @file{/etc/services} and setting reusable socket options to avoid busy signals, like this: @smallexample struct sockaddr_in cin, sin; struct servent *server; int sd, addrlen = sizeof(cin); int portnumber, yes=1; if ((server = getservbyname(@var{service-name},"tcp")) == NULL) @{ CfLog(cferror,"Couldn't get cfengine service","getservbyname"); exit (1); @} bzero(&cin,sizeof(cin)); /* Service returns network byte order */ sin.sin_port = (unsigned short)(server->s_port); sin.sin_addr.s_addr = INADDR_ANY; sin.sin_family = AF_INET; if ((sd = socket(AF_INET,SOCK_STREAM,0)) == -1) @{ CfLog(cferror,"Couldn't open socket","socket"); exit (1); @} if (setsockopt (sd, SOL_SOCKET, SO_REUSEADDR, (char *) &yes, sizeof (int)) == -1) @{ CfLog(cferror,"Couldn't set socket options","sockopt"); exit (1); @} if (bind(sd,(struct sockaddr *)&sin,sizeof(sin)) == -1) @{ @} /* etc */ @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Multithreading a server, System databases, Sockets, Network programming @section Multithreading a server All the arguments must be collected into a struct, since only one argument pointer can be sent to the pthread functions. @smallexample #include SpawnCfGetFile(args) struct cfd_thread_arg *args; @{ pthread_t tid; void *CfGetFile(); pthread_attr_init(&PTHREADDEFAULTS); pthread_attr_setdetachstate(&PTHREADDEFAULTS,PTHREAD_CREATE_DETACHED); if (pthread_create(&tid,&PTHREADDEFAULTS,CfGetFile,args) != 0) @{ CfLog(cferror,"pthread_create failed","create"); CfGetFile(args); @} pthread_attr_destroy(&PTHREADDEFAULTS); @} /***************************************************************/ void *CfGetFile(args) struct cfd_thread_arg *args; @{ pthread_mutex_t mutex; if (pthread_mutex_lock(&mutex) != 0) @{ CfLog(cferror,"pthread_mutex_lock failed","pthread_mutex_lock"); free(args->replyfile); /* from strdup in each thread */ DeleteConn(args->connect); free((char *)args); return NULL; @} ACTIVE_THREADS++; /* Global variable */ if (pthread_mutex_unlock(&mutex) != 0) @{ CfLog(cferror,"pthread_mutex_unlock failed","unlock"); @} /* send data */ if (pthread_mutex_lock(&mutex) != 0) @{ CfLog(cferror,"pthread_mutex_lock failed","pthread_mutex_lock"); return; @} ACTIVE_THREADS--; if (pthread_mutex_unlock(&mutex) != 0) @{ CfLog(cferror,"pthread_mutex_unlock failed","unlock"); @} #endif return NULL; @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node System databases, DNS, Multithreading a server, Network programming @section System databases @cindex NIS @cindex Network information service @cindex Network databases @cindex Database maps The C library calls which query the databases are, amongst others, @smallexample getpwnam @var{get password data by name} getpwuid @var{get password data by uid} getgrnam @var{get group data by name} gethostent @var{get entry in hosts database} getnetgrent @var{get entry in netgroups database} getservbyname @var{get servive by name} getservbyport @var{get service by port} get protobyname @var{get protocol by name} @end smallexample @noindent For a complete list and how to use these, see the UNIX manual. @cindex @code{getpwnam()} @cindex @code{getpwuid()} @cindex @code{getgrnam()} @cindex @code{gethostent()} @cindex @code{getnetgrent()} @cindex @code{getservbyname()} @cindex @code{getservbyport()} The following example shows how to read the password file of the system. The functions used here can be used regardless of whether the network information service (NIS) is in use. The data are returned in a structure which is defined in @samp{/usr/include/pwd.h}. @smallexample /******************************************************************/ /* */ /* Read the passwd file by name and sequentially */ /* */ /******************************************************************/ #include #include main () @{ uid_t uid; struct passwd *pw; uid = getuid(); pw = getpwuid(uid); printf ("Your login name is %s\n",pw->pw_name); printf ("Now here comes the whole file!\n\n"); setpwent(); while (getpwent()) @{ printf ("%s:%s:%s\n",pw->pw_name,pw->pw_gecos,pw->pw_dir); @} endpwent(); @} @end smallexample @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node DNS, C and NFS, System databases, Network programming @section DNS - The Domain Name Service @cindex DNS @cindex The domain name service @cindex BIND @cindex Berkeley Internet Name Domain (BIND) The second network database service is that which converts host and domain names into IP numbers and vice versa. This is the domain name service, usually implemented by the BIND (Berkeley Internet Name Domain) software. The information here concerns version 4.9 of this software. @c ..................................................... @c SUBSECTION @c ..................................................... @menu * gethostbyname:: @end menu @node gethostbyname, , DNS, DNS @subsection gethostbyname() This is perhaps the most important function form hostname lookup. @samp{gethostbyname()} gets its information either from files, NIS or DNS. Its behaviour is configured by the files mentioned above, @xref{DNS}. It is used to look up the IP address of a named host (including domain name if DNS is used). On the configurable systems described above, the full list of servers is queried until a reply is obtained. The order in which the different services are queried is important here since DNS returns a @emph{fully qualified name} @cindex Fully qualified name (host name plus domain name) whereas NIS and the @file{/etc/hosts} file database return only a hostname. gethostbyname returns data in the form of a pointer to a static data structure. The syntax is @smallexample #include @cindex @code{gethostbyname()} struct hostent *hp; hp = gethostbyname("myhost.domain.country") @end smallexample @noindent The resulting structure varies on different implementations of UNIX, but the `old BSD standard' is of the form: @smallexample struct hostent @{ char *h_name; /* official name of host */ char **h_aliases; /* alias list */ int h_addrtype; /* host address type */ int h_length; /* length of address */ char **h_addr_list; /* list of addresses from name server */ @}; #define h_addr h_addr_list[0] /* address, for backward compatiblity */ @end smallexample @noindent The structure contains a list of addresses and or aliases from the nameserver. The interesting quantity is usually extracted by means of the macro @samp{h_addr} whcih gives the first value in the address list, though officially one should examine the whole list now. This value is a pointer which can be converted into a text form by the following hideous type transformation: @smallexample #include #include #include struct sockaddr_in sin; cin.sin_addr.s_addr = ((struct in_addr *)(hp->h_addr))->s_addr; printf("IP address = %s\n",inet_ntoa(cin.sin_addr)); @end smallexample See the client program in the first section of this chapter for an example of its use. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node C and NFS, Ex 7, DNS, Network programming @section C support for NFS The support for NFS mounting in the standard C library is through two sources. NFS is based on the Sun's RPC system, so the basic calls are only instances of standard RPC protocols. @cindex NFS and C support The C functions in the standard input/output library can be used to access NFS filesystems. Since NFS imitates the UNIX filesystem as closely as possible, NFS filesystems can be mounted in exactly the same way as ordinary filesystems. Unfortunately, the C functions which perform the mount operation in UNIX and depressingly non-standard. They differ on almost every implementation of UNIX. The basic function which mounts a filesystem, in @samp{mount} (see @kbd{man (2) mount}). The mount table is stored in a file @emph{/etc/mtab} on BSD systems (again the name varies wildly from UNIX to UNIX, mnttab on HPUX for instance). The file @emph{/etc/rmtab} on an NFS server contains a list of remote-mounted filesystems which are mounted by remote clients. C functions exist which can read the filesystem tables and place the resulting data in C struct types. Alas, these struct defintions are also quite different on different systems. See @file{/usr/include/sys/mount.h}, so the user wishing to write system-independent code is confounded at the lowest level. @c ----------------------------------------------------------------------- @c SECTION @c ----------------------------------------------------------------------- @node Ex 7, , C and NFS, Network programming @section Exercises @enumerate @item Use @samp{gethostbyname()} to make a simple program like @samp{nslookup} which gives the internet address of a named host. @item Modify the client server example above to make a `remote ls' command called @samp{rls}. You should be able to use the syntax @smallexample rls (options) hostname:/path/to/file @end smallexample @end enumerate @c ********************************************************************** @c APPENDIX @c ********************************************************************** @node Programming idioms, Variable Index, Network programming, Top @appendix Summary of programming idioms. @table @emph @item True and false @smallexample # C shell True - non-zero/non-empty value False - zero or null string # Bourne shell True - 0 returned by shell command False - non-zero returned by shell command ( Note that "test" converts from C shell style to Bourne shell) # Perl True - non-zero/non-empty value False - zero or null string /* C */ True - non zero integer False - zero integer @end smallexample @item Input from tty @smallexample # C shell $< # Bourne shell line read # Perl /* C */ scanf @end smallexample @item Redirection of I/O @smallexample # C shell @var{command} > file @var{command} >& file @var{command} >> file @var{command1} | @var{command2} # Bourne shell @var{command} > file @var{command} > file 2>&1 @var{command} >> file @var{command1} | @var{command2} # Perl open (HANDLE,">file") open (HANDLE,">file 2>&1") open (HANDLE,">>file") open (HANDLE,"command1 |") open (HANDLE,"| command2") /* C */ fopen ("file","w"); printf(..) fopen ("file","w"); printf(..); fprintf(stderr,..) fopen ("file","a"); printf(..) popen ("@var{command1}","r") popen ("@var{command2}","w") @end smallexample @item Loops and tests @smallexample /* C */ Shell foreach end if then else endif while end switch case breaksw endsw repeat # Bourne shell while do done if then else fi until do done case in esac for in do done # Perl while if then else for unless else foreach until do while do until /* C */ while if then else do while switch case for @end smallexample @item Arguments from command line @smallexample # C shell $argv[] $#argv # Bourne Shell $1, $2, $3... $* $# # Perl $ARGV[] $#ARGV /* C */ char argv[][] int argc @end smallexample @item Arithmetic @smallexample # C shell @ a = $b + $c # Bourne shell a = `expr $b + $c` # Perl $a = $b + $c; /* C */ a = b + c; @end smallexample @item Numerical comparison @smallexample # C shell if ( $x == $y ) then endif # Bourne shell if [ $x -eq $y ]; then fi # Perl if ( $x == $y ) @{ @} /* C */ if ( x == y ) @{ @} @end smallexample @item String comparison @smallexample # C shell if ( $x == $y ) then endif # Bourne shell if [ $x = $y ]; then fi # Perl if ( $x eq $y ) then @{ @} /* C */ if (strcmp(x,y) == 0) @{ @} @end smallexample @item Opening a file @smallexample # C shell, Bourne shell - cannot be done (pipes only) # Perl open (READ_HANDLE,"filename"); open (WRITE_HANDLE,"> filename"); open (APPEND_HANDLE,">> filename"); /* C */ FILE *fp; fp = fopen ("file","r"); fp = fopen ("file","w"); fp = fopen ("file","a"); @end smallexample @item Opening a directory @smallexample # C shell foreach dir ( @var{directory}/* ) ... end # Bourne shell for dir in @var{directory}/* ; do ... done # Perl opendir (HANDLE,"@var{directory}") || die; while ($entry = readdir(HANDLE)) @{ @} closedir(HANDLE); # C #include DIR *dirh; struct dirent *dirp; if ((dirh = opendir(name)) == NULL) @{ perror("opendir") exit(1); @} for (dirp = readdir(dirh); dirp != NULL; dirp = readdir(dirh)) @{ ... /* dirp->d_name points to child */ @} closedir(dirh); @end smallexample @item Testing file types @smallexample # C shell if ( -f file ) # plain file if ( -d file ) # directory # Bourne shell if [ -f file ] # plain file if [ -d file ] # directory # Perl if ( -f file ) # plain file if ( -d file ) # directory if ( -l file ) # symbolic link /* C */ #include struct stat statvar; stat("file", &statvar); if (S_ISREG(statvar.mode)) /* plain file */ if (S_ISDIR(statvar.mode)) /* directory */ lstat("file", &statvar); if (S_ISLNK(statvar.mode)) /* symbolic link */ @end smallexample @end table @c ********************************************************************** @c INDEX @c ********************************************************************** @node Variable Index, Concept Index, Programming idioms, Top @unnumbered Command and Variable Index @printindex vr @node Concept Index, , Variable Index, Top @unnumbered Concept Index @printindex cp @contents @bye