Ok then... Mark Picone, Trainee Unix Administrator Deakin University, Information Technology Services Division Phone: 03 5227 8602 International: +61 3 5227 0806 Fax: 03 5227 8799 International: +61 3 5227 8799 Email: [EMAIL PROTECTED] Website: http://www.deakin.edu.au
> -----Original Message----- > From: [EMAIL PROTECTED] [mailto:owner-freebsd- > [EMAIL PROTECTED] On Behalf Of johnharten > Sent: Wednesday, 17 September 2008 8:05 AM > To: freebsd-ports@freebsd.org > Subject: Linux Help > > LINUX is not UNIX, but it's close enough > This book is old > I will try to take the concepts the book lays out and integrate them > with more recent versions. > Lecture material will be a hybrid > Chapter 1 > > Logging on to the System > Why we study UNIX/LINUX > Started in the 1970's (pre-Microsoft) > UNIX runs “everything” > The Internet, Stock Market, Movies, technological advancements, > Embedded Devices (ATM's), POS systems, the military, utility > companies, and much more. > Linux began in the early '90's > Has revolutionized the IT world > Authentication > Authentication is the process of confirming that you are who you say > you are > Logging in (authenticating) typically requires two components > Username (login name) > Password > Authentication (continued) > Authentication can also take the following forms: > Public/Private Key Pair > PGP / SSL certificates > Biometric Authentication > Fingerprints > Retinal scans > Voice matching > Authorization > Authorization is the process of determining who gets access to what. > Unlike authentication, authorization does not generally involve and > additional input. > Keeps the system secure. > Login Scenarios > Running system with the root username and password > Running system where you have the username / password of non-root user > A system with no Operating System > Scenario: No OS > If there is no operating system present, you'll need to install one > Setting the root password is part of the installation process > You'll need to create a regular user account for yourself as well > Scenarios: Have Credentials > Running system where you have a valid username / password (root or > otherwise) > At login prompt, enter valid username and hit enter > At password prompt, enter valid password and hit enter > Have Credentials (cont'd) > Whether or not you enter the username correctly, you will be prompted > for a password. > When you type your password, the characters will be masked by *'s for > security > UNIX is case-sensitive!!! > Notes on root user > root is the system administrator > root has access to all resources and there is no safety net > Use root only when necessary > Login Prompts > Command Line > Telnet / ssh / no graphical environment > Graphical > X Windows system is installed on system > Pictures of both on page 5 > Lab Work > Install Linux > Fedora 9, because we have new hardware. > > CMPSC 249: > > Introduction to UNIX/Linux > Week 2: UNIX Essentials > The Kernel > The kernel is the core of an operating system. > Source Provided for compilation. > RedHat provides RPM packages. (rpm –qa |grep –i kernel) > You don’t touch it directly. > Where is the Kernel? > The boot loader runs and loads the kernel based on the choice you > make. > Editing the boot loader entry will reveal the location of the kernel > (usually /boot/vmlinuz*) > RamDisk and kernel load and then process id 1 (init) is spawned. > User Space > User space is the visible part of the operating system. > User processes > Services / daemons > Every user space component executes system calls and spends some time > in the kernel. > > > What does the Kernel do? > Sits between programs (user space) and hardware. > Applications use syscall facilities to have the kernel perform work on > their behalf. > This work takes many forms. > > Kernel > Kernel (2) > System call facility that allows processes to use kernel functions. > Process creation and tracking > Process priority control > Swapping pages & memory management > IPC - inter-process communication > Cache, Buffer and I/O management > File creation, removal and modification and permissions > Filesystems > Log file data accumulation and flushing. > > The Filesystem > What is a filesystem? > > <geek>a method for storing and organizing computer files and the data > they contain to make it easy to find and access them.</geek> > What is a filesystem (English) > A container for data > More than just a place to keep files > Hierarchical > File Attributes > Security > > Filesystem Layout > / = System Top Level Directory. Start of the tree > Everything is hierarchical in folders underneath / (/opt, /usr, /etc) > Folders can act as mount points for local and network filesystems > Common Directories > /usr – shared, read-only libraries and binaries. > /etc – machine configuration files > /home – user directories (private) > /boot – location of necessary files for boot loader. Usually at > beginning of disk > Common Directories (2) > /dev – location of special device descriptor files > /proc – Virtual filesystem that gives insight into the kernel and > running tasks. > /var – Typical location of spool (print, mail) and log files. > /opt – optional installed packages > /dev and /proc are virtual file systems > Separation of FileSystems > Separation involves creating individual filesystems for each mount > point. > Prevents an errant process or user from filling up all filesystems and > crashing the system. > Separating everything is not practical. > > Disk Partitioning > Very similar to Windows, Linux supports carving of drives into > partitions. > During our install we created separate partitions for various > filesystems > The fdisk command can be used to create and view partitions. > To view current disks and partitions use fdisk –l > > Disk Partitioning (2) > Disk /dev/sda: 80.0 GB, 80000000000 bytes > 255 heads, 63 sectors/track, 9726 cylinders > Units = cylinders of 16065 * 512 = 8225280 bytes > Device Boot Start End Blocks Id System > /dev/sda1 * 1 32 257008+ 83 Linux > /dev/sda2 33 1307 10241437+ 83 Linux > /dev/sda3 1308 1568 2096482+ 82 Linux swap / > Solaris > /dev/sda4 1569 9726 65529135 5 Extended > /dev/sda5 1569 1600 257008+ 83 Linux > /dev/sda6 1601 2875 10241406 83 Linux > /dev/sda7 2876 9726 55030626 8e Linux LVM > [EMAIL PROTECTED] ~]# > > Disk Partitioning (3) > To match partitions to filesystems you can use df to correllate: > [EMAIL PROTECTED] ~]# df > Filesystem 1K-blocks Used Available Use% Mounted on > /dev/sda6 9920592 1463072 7945452 16% / > /dev/sda5 248895 17297 218748 8% /boot > tmpfs 479236 0 479236 0% /dev/shm > /dev/mapper/vg_reboot-lvhome > 19838052 16125436 2688616 86% /home > /dev/mapper/vg_reboot-lvtmp > 9560920 72440 9010156 1% /tmp > /dev/mapper/vg_reboot-lvusr > 6983168 4129136 2493612 63% /usr > /dev/mapper/vg_reboot-lvvar > 2951952 701460 2098124 26% /var > > LVM > LVM (Logical Volume Management Devices) > Volume Groups are essentially logical disks that can actually span > multiple devices. > /dev/mapper entries are logical volumes within a volume group. > Volume groups can consist of partitions on disks or whole disks called > physical volumes. > Physical volumes (pv’s) are raw LVM devices with a partition type of > 8e or Linux LVM. > “vgdisplay” command lists current lvm volume groups. With no options, > it prints a stanza for each defined volume group (vg) > [EMAIL PROTECTED] ~]# vgdisplay > --- Volume group --- > VG Name vg_reboot > … > > > LVM (continued) > Vgdisplay gives other useful information as well: > PE Size 32.00 MB > Total PE 1679 > Alloc PE / Size 1239 / 38.72 GB > Free PE / Size 440 / 13.75 GB > > PE’s are Physical Extent. It’s the smallest allocation size you can > have in a VG. > Total PE is the number of PE’s in the volume group. > You multiply PE size by Total PE to get the size of the vg. > Allocated PE is the total amount of PE assigned to a volume > Free PE is the total amount of PE that is available for use. > LVM Part 3 > How do you determine how many devices are contained in a vg? > Easy – you use vgdisplay –v. At the very bottom of the output will be > a stanza for each physical volume. > > --- Physical volumes --- > PV Name /dev/sda7 > PV UUID YDgW5U-0gFQ-kzBm-HJh5-0l7F-V1Cp-X7VPcj > PV Status allocatable > Total PE / Free PE 1679 / 440 > > > LVM 4 > So, we have pv’s and vg’s – what about filesystems? > You need a logical volume. Vgdisplay –v also gives information on > each lv: > --- Logical volume --- > LV Name /dev/vg_reboot/lvusr > VG Name vg_reboot > LV Status available > LV Size 6.88 GB > Current LE 220 > > Creating PV’s > Create a partition with fdisk of type 8e (Linux LVM) – BE CAREFUL not > to change an existing partition. > fdisk /dev/sda will open the fdisk menu for /dev/sda > The letter m prints help information, but you want to use n to create > a new partition. > Accept default for starting location (first available sector). > Default end will be end of disk. > The t option lets you specify type. > Use the letter w once you are sure it’s right to activate changes. > It will created /dev/sda# for you > Once done, you’ll need to execute the following command: > pvcreate /dev/sd# (where # is the partition you just > created). > Creating VG’s and LV’s > Once you have your pv, creating a volume group is as easy as > executing: > vgcreate classvg /dev/sda# (where number is the pv you just > created) > > Once you have your vg, you can create an LV contained within: > lvcreate –n classlv -L 1G classvg > > NOTE – lvcreate and lvextend accept –L for human readable size as well > as –l to specify the number of PE’s > Resizing Filesystems > Resizing filesystems can be done online with resize2fs. > Resize2fs /dev/vg_reboot/lv_to_increase – by default it will increase > to LV size > MUCH easier to resize a filesystem in an LV. > LV’s can be grown without being contiguous. > lvextend –L +5G /dev/vg_reboot/lv_to_increase will increase the > logical volume by 5G > If you need to resize a partition it can be risky because you need to > reorder your partition scheme. > Command Syntax Basics > Commands consist of three parts: > Binary Name > Options > Arguments > > Options are indicated by “-” or “—” prefixes. > indicates a single letter (Linux Style) > --indicates a word argument (BSD Style) > > Arguments take many forms. > Values for options (-L 5G) > Full paths to files or directories on which to perform actions. > “ls” command > Command: ls > Purpose: Lists files and directories (like dir) > Options: common options are –la (long listing). Also use –t to sort > by time, latest first. -r reverses the sort. > Common use: ls -latr > Arguments: directory or path name (fully qualified or in current dir). > “cat” command > Command: cat > Purpose: display contents of a file > Options: none > Arguments: filename > > “more” command > Command: more > Purpose: another way to show the contents of a file. Breaks by page, > so you can read it. Space advances a page, entery advances a line. > Options: none > Arguments: filename > > “grep” command > Command: grep > Purpose: Funny name, serious tool. Grep looks for pattern matches in > a file. > Options: –i is my favorite. Means case insensitive (remember, UNIX > is case sensitive). Also grep –v means show me everyline but the ones > that contain this text. > Arguments: filename > > > “cd” > Command: cd > Purpose: Changes your current directory. Default directory is /home/ > student. > Options: none > Arguments: Directory name > > NOTE: cd – puts you back to your old directory. > > > “mkdir” > Command: mkdir > Purpose: Creates a directory > Options: -p if you’re making a deep directory mkdir –p /dir1/dir2/dir3 > will create dir1, dir2, and dir3 > Arguments: Directory name > > “pwd” > Command: pwd > Purpose: Prints your current directory (you are here). > Options: none > Arguments: none > > “rm” > Command: rm > Purpose: Removes a file > Options: -f (means don’t ask me if I’m sure). -R means recursive – > the only way to remove a directory. > Arguments: file name > > NOTE: rm –Rf / is ALWAYS a bad idea > > “df” > Command: df > Purpose: Displays information on mounted filesystems > Options: -h (human readable. Calculates in GB, MB, or KB). > Arguments: none > > “cp” > Command: cp > Purpose: copies one file or directory to another, preserving the > original. > Options: -R for directories –p to preserve permissions > Arguments: Directory 1, directory2 > > “mv” > Command: mv > Purpose: Moves one file or a directory to another – rename. > Options: none > Arguments: file1, file2 > > “echo” > Command: echo > Purpose: prints a string to a screen > Options: none we’ll worry about right now. > Arguments: text --- if spaces, enclose in “’s > > “appropos” > Command: appropos > Purpose: English to geek translator. > Options: none > Arguments: when you know the purpose of the command, appropos might > help you find it. > > “man” > Command: man > Purpose: more information than you will ever want to know. > Options: none > Arguments: command > > Bash basics > Bash is your friend. IT is the lazy man’s shell. > If you know the command name, you can type the first few letters and > hit tab twice. It will show you your options. > Also, to use your last command (or edit it) just hit up. > Notes > Linux treats spaces as special characters. Avoid them at all costs. > If you are dead set on using them, you’ll have to escape them with \. > If you don’t know how to use a command, try man. > If you want a quick reference, 9 times out of 10 you can type the > command without any options or arguments and hit enter. It will tell > you. > > CMPSC 249: > > Introduction to UNIX/Linux > Syllabus Update > Test has been pushed from next week. > > > All the LVM material from last week’s class will not be tested on. > Tonight’s Agenda > Revisit Common Commands > > Chain Commands Together > > Redirecting Output > > Any questions from last week? > > Navigating the filesystem > Filesystem: collection of files and directories contained on a block > device. > Examples of commands to navigate the filesystem: pwd, mv, cp, ls > > > Navigating the Filesystem > The best place to start is how to figure out where you are. > To get a listing of your current directory execute: pwd > Pwd tells you where you are, and as such where your various file > operations will look for their input > The output of that command will look like: > /home/student > > Looking for files > We’ll talk about ls again in a few minutes, but in the meantime, > without any arguments or options, it serves one purpose: > To show you a list of the files in your current directory: > [EMAIL PROTECTED] ~]# ls > backup other_stuff scripty tmp.sh > French.zip questions.out test.txt uniq_q_n_a.txt > [EMAIL PROTECTED] ~]# > > / -- ain’t it cool > / is where all files start from > To get a listing of all objects in / execute the following: ls / > Note, in the command above, / serves the purpose of an argument. > The output should look something like: > [EMAIL PROTECTED] ~]# ls / > backup boot etc lib media mnt Old proc sbin srv > tmp var > bin dev home lost+found misc net opt root selinux sys > usr > > > > Looking at other directories > What happens when you want to view a filesystem that has thousands of > files in it? > ls will display it’s output and it will scroll way past the top of > your screen, and you won’t be able to see it all. > Try it --- go ahead, I’ll wait (ls /etc) > This is where you get to meet a UNIX Admin’s best friend: | > > | -- also cool > | (called “pipe”) serves the purpose of taking the output from one > command and passing it as input to another. It’s basically a chain. > By itself ls | would do nothing. However, if we introduce the more > command, our output is broken up by our screen size. > ls |more – try it. I’ll wait > You can use the Enter key to advance one line at a time, or the space > key basically does a page down, q will exit more without looking at > all the output. > Directories Gone Wild > So, this whole / thing is pretty cool, but what if I want to make my > own directory? > Easy, use mkdir > mkdir takes one argument – the destination directory. > If you do not specify a full path (mkdir /home/student/cookiemonster), > mkdir will create the directory in your current directory. > mkdir tmp will create a dir called tmp wherever you are. > Go ahead, try it…I’ll wait *use ls to validate* > Moving between directories > So, you know what your pwd is, and you know you want to get to /, but > how do you make / your current directory? > Use cd. Much like in windows/dos, cd changes your current directory. > If you fully path the directory (cd /home/etc/sysconfig/) you will be > dropped in /etc/sysconfig. > Without specifying a full path (cd cookiemonster) cd will look in your > pwd for that directory and fail if it does not exist. > Go ahead, try it. I’ll wait. > cd with no arguments > What happens if you run cd with no arguments? Where do you get > placed? > > Anyone? Try it and see. > More more > As with most commands in UNIX, more can be used in a few different > ways. > When we used more before, we piped the output of ls to more. > But what if I want to look at a file to see what is in it? > Well, you can use more in one of two ways: more /etc/termcap > Or, you can use cat to print the contents of the file and pipe it to > more: cat /etc/termcap | more > What’s the difference? > Looking only at parts of a file > Look at the beginning of a file: head /etc/termcap > Look at the end of a file: tail /etc/termcap. > These commands will show you the first ten and the last ten lines of > the file. > Or, as above, cat /etc/termcap |head and cat /etc/termcap | tail > See a pattern developing? > > I wanna see more!!! > Well, head and tail both support one option: -n > -n specifies a number of lines to see. > tail –n 20 /etc/termcap will show the last 20 lines. > Try this: tail –n 100 /etc/termcap > What do you need to do to make that all visible? tail –n 100 /etc/ > termcap/more > > Tail continued > Tail supports an option to view a file in real time: tail –f /etc/ > termcap. > Will show you the last 10 lines and then sit there. > If any new characters get added to the file –f forces tail to show > them to you as they arrive. > VERY useful for debugging and monitoring log files. > How do you interrupt it? Any ideas? > Bueller? > cp/mv > We used cp/mv in lab last week. Anyone remember how they worked? > cp was to copy, mv was to rename. Only, mv is actually a move > command, and does more than just rename. > Using cp to copy a file involves two arguments: source and > destination. > If destination is a directory, cp and mv will put their output in the > directory as opposed to pwd. > > Notes > Remember, there is a difference between fully pathing (/etc/ > sysconfig/) and shortening the path (sysconfig). > Anyone want to tell us what that difference is? > Cp supports many source files. You can do something like this: cp > file1 file2 file3 file4 backup_directory. All files will be put in > the directory: backup_directory. > Time to remove > rm = delete a file > rm –i puts rm into interactive (windows) mode. It asks for > confirmation before deleting. It’s safer to run as root. > rm accepts many arguments as files to remove. > Let’s try an experement… > touch file_1 > rm file_1 > > > What happened? Did it ask you to remove the file, or did it just do > it? > Confirmation Elimination > What do you do if rm asks you if it’s ok to remove a file every > time? > You can use rm –f (f for force). > It’s probably set up as a command alias for your user. > [EMAIL PROTECTED] ~]# alias > alias rm='rm -i‘ > > > > > > To get rid of the alias execute: unalias rm --- but this will only > work for this session. The next time you log in you’ll have to change > it. > > Removing files with wild cards > * is a wildcard, meaning it matches any character. > If you have a bunch of files to delete that all have a similar name, > you can use * to your advantage. > Let’s try: > Create the files: > touch file_1 > touch file_2 > touch file_3 > touch file_4 > Wild cards Continued > We created the files, now let’s remove them. > To do it manually would require either one long command (rm file_1 > file_2 file_3 file_4) or four separate remove commands. > Instead, we can save ourselves some work: > rm file_* > Common error scenarios > Command not found: indicates you have spelled your base command > wrong. > Using copy instead of cp > No such file or directory: indicates that one of your options is > incorrect (you’ve tried to look at a file that doesn’t exist) > ls /cookiemonster > > Errors (2) > If you get an error message, but you are sure that you’ve spelled the > command right, often times you can use man to figure out what you are > doing wrong. > man ls > Navigating man pages is similar to navigating output from more. > > Redirecting Output > We use | to move output from one program to another in a command > string. > What if we wanted our commands to output directories to a file? > There are two ways to do this. One way appends to an existing file, > the other creates a new file. > Redirecting Output > ls > ls.out – this creates a new file called ls.out. If ls.out > already existed, it would destroy the file and create a new one for > us. > ls >> ls.out – this will append the output of our command to the > existing ls.out file. If ls.out did not exist, it would create it for > us. > Line counting > Sometimes you really need to see how big a file is. The wc command > counts certain types of entries. > To see how many lines are in the /etc/termcap file, you’d use: cat / > etc/termcap | wc –l > wc –w shows how many words > wc –c shows how many characters > Without any options, wc will show you all three: > [EMAIL PROTECTED] ~]# cat /etc/termcap |wc > 19092 91266 807103 > [EMAIL PROTECTED] ~]# > Lines Chars Words > > Command Arguments > As we discussed last week, commands are made up of three pieces: > Binary > Options (Flags – preceeded with a -) > Arguments > These arguments can be combined in various ways. > > Ls Options > ls –l shows us more columns of output. > [EMAIL PROTECTED] ~]# ls > backup other_stuff scripty tmp.sh > French.zip questions.out test.txt uniq_q_n_a.txt > > > [EMAIL PROTECTED] ~]# ls -l > total 1524 > drwxr-xr-x 4 root root 4096 Dec 23 2006 backup > > Permissions Owner Group Size Time Name > > Ls Options > Using ls –la will show you hidden files as well. > Hidden files begin with a . > Other than that, the output is the same to ls –l > Using the –t option for ls sorts the files by modification time with > the latest first. > Using the –r option combined with –t reverses the time search order, > and puts oldest last. > Combining ls Options > So, we like –l, -a, -t, and –r for ls. It gives as much information > as we’d need to look at the files. How do we use them? > ls –l –a –t –r (long way) > ls –latr (lazy way) > The options are able to be combined by a single: – > > This is true for most commands you will use. > Sorting your output > By default the output you will from some commands seems to have no > real order. > Not all files are maintained in alphabetical order. > So, the powers that be developed a command to help us poor humans: > sort > Sorting 2 > The /etc/passwd file contains information on all users on the > system. > When new users get created, they get added to the bottom of the > file. > It gets out of hand quickly. So, let’s try the following: > cat /etc/passwd – It’s hard to find entries > sort /etc/passwd – much easier to read > More sorting > Looking to reverse that sort, backwards alphabetical anyone? > sort –r has got your back. > Very similar to the ls –latr (reverse time sort) > Just remember, UNIX is case sensitive, so A does not equal a out of > the box. > You can use sort –f to ignore case > > > If you have a list of files that has multiple entries, many of which > are the same, you can use sort to extract unique records: sort –u > We all need a break > Great time to break. Be back here in 10 minutes. > The sleep command > Sometimes you are writing a script that needs to allow time for > something to complete. > For that the sleep command was invented. > sleep takes one argument, and one argument only. A number. > It’s job is basically to wait that many seconds and the return control > to you. sleep 5 sleeps for 5 seconds, etc. > Passing Arguments > You can see what’s happening with arguments by issuing this command: > echo A B C D > file > cat file > > > What does this tell us? > echo is spawned with 4 arguments > echo operates and redirects its output to a file > Combining files > Suppose you have three separate files that you need in one big file. > No problem! > date >file1 > echo hello world > file2 > ls > file3 > Cat each of these files individually to see the results. > Merge them by executing: > cat file1 file2 file3 > bigfile > cat bigfile > Locating Specific lines in a file > grep is hands down one of the most useful utilities ever. > Quickly described, grep simply looks for a pattern and outputs the > match. > A quick example would be: > grep student /etc/passwd > What else can grep do? > You can use grep to find files: > ls |grep “file1” > You can use grep to match case insensitive: > grep –i student /etc/passwd > You can use grep to match everything but: > grep –v student /etc/passwd > Notes > Just a reminder, many utilities can be executed as part of a command > string: > cat /etc/passwd |grep –i student > They can also be executed independently > grep –i student /etc/passwd > > > > > Both methods produce identical output > Some utilities are dumb > cat, more, sort when executed without any arguments or options will > just sit there and stare at you blankly > Why? Because they are waiting for you to do something. > Execute the sort command and hit enter. > What happens? > Sort with nothing > You didn’t give it a file to work with. You didn’t give it a stream > to work with, so it’s waiting for you to provide the data for it to do > it’s work. > Enter the following strings: > Hello > Goodbye > 42 > AAA > Aaa > Then hold ctrl and hit D. This tells sort that it’s reached the end > of the input and it does its work. It will use your input as its > source. > Using cat to create a new file > cat > newfile > Opens the cat binary and attaches to newfile for output. > Drops to the input method we just used to get it’s contents. Enter: > Hello from cat > Then ctrl-D. > cat newfile – will show your entry. > You can also create a file with echo: echo “hello” > newfile2 > > CMPSC 249: > > Introduction to UNIX/Linux > Week 4 -- vi > > Announcements > > Test next week > We will review at the end of lecture > > > vi > The vi that we are using tonight is not actually, it’s vim (VI > iMproved). > VIM is not the same as vi, but the basics are close. Not all UNIXes > are the same when it comes to vi. > Type the following command: vi > ~ VIM - Vi IMproved > ~ > ~ version 7.0.109 > ~ by Bram Moolenaar et al. > ~ Vim is open source and freely distributable > > > What now? > ~ type :q<Enter> to exit > ~ type :help<Enter> or <F1> for on-line help > ~ type :help version7<Enter> for version info > This information is useful, because it gives you some of vi’s more > basic commands. > Use :q to exit vi. > > VI modes > VI has two basic modes: command and insert. > VI always starts in command mode. > Special mode for vi control commands. > Insert mode is the only way to actually modify text. > Insert mode > VI help > So, when you enter vi, you can use :help to navigate through the help > pages. > It’s a little quirky, and easy to get lost. I often refer to: > http://vimdoc.sourceforge.net/htmldoc/help.html > What’s vi for? > As we’ve discussed, the vast majority of UNIX programs are controlled > by text files. > VI gives you a method to edit those files. VI can: > Move around a file > Delete Text > Copy and Paste Text > Moving around (viewing) a file > The arrow keys can help you move through a file. For example, execute > the following command: vi /etc/passwd > Use the arrow keys to scroll up, down, left, and right in the file. > > Quitting vi > Ctrl-c will not work with vi. > You can manually kill the process from another window or use the q > command. > Esc + :q to exit (might not need esc, but always good to be safe) > You must ALWAYS precede commands with : > Moving 2 > In vi, there are special key sequences to perform certain tasks. > For example: > Shift + G moves to the last line in a file > $ moves to the end of the line you are currently on. > : and a number moves to that specific line (:1 moves to line 1) > > Who needs a mouse? > When not in insert mode, vi allows you to move around with more than > just arrow keys. > w allows you to move your cursor from word to word. > e moves the cursor to the end of the word > b moves back one word. > Cursor positioning is key to hitting insert mode running. > > > Searching in a file > You can use vi to search for a specific pattern in a file. > Anyone remember the command we used to search for patterns last week? > In vi you precede your search string with a / > vi /etc/passwd > Type: /root it will take you to the line that matches root. > The “n” key will move to the next match. > An uppercase N will take you to the previous match. > Searching continued > UNIX is case sensitive, remember? > Grep, sort, and other utilities have builtin ways of being case- > agnostic. > Vi is no different, although there is no special command – it’s a > setting. > You need to use the following: > Esc + :set ignorecase > Using VI to edit an existing file > Find one of the files we created last week in lab. > vi that file (vi $filename) > Go to the end of the file. > Go to the end of the line. > Append the following line: ABC123 > Save the file (esc + :wq) > cat it to ensure the text got saved. > Using vi to create a new file > Similar to some of the topics we covered last week, if you give vi an > argument that is a filename, vi will open that file. > If the filename does not exist, vi will create it for you. > Let’s execute the following command: > vi vi_file_create.txt > > > Switching out of command mode > So, you’ve opened your file, and you want to put text in it. > If this is truly a new file, vi will tell you at the bottom: > "vi_tmp_file.txt" [New File] > > But, you’re in command mode. > Type the letter h. Nothing happens. > > To get out of insert mode, type: i > > I is a vi command that switches to insert mode. Now, you can type > till your heart’s content. Go ahead, put a few lines of text in. > > Done entering text? > If you hit :q here, all changes are lost. Your file will not > created. > So, how you actually write to your file? > Four key components: > Hit the escape key. This is the way to exit command mode > : -- Indicates a command string is coming > w – is the vi command to write your changes to the file. > q – command to quit. > > > > Often the command sequence used is: Esc + :wq > Accomplishes this all in one command > Checking your input > cat vi_tmp_file.txt > See your text? > You’ve now used vi to create a file. > > What commands did you use to create files last week? > Key Concept > “vi rocks” – Ron Utsinger, Caterpillar, Inc UNIX Operations. > The Esc key switches from insert mode to command mode. > If you accidentally put your command in your text, you need to switch > to command mode. > Before you execute a command it is a good idea to ensure you’re out of > insert mode by using escape. > Functions > You’ve probably noticed some commands are not preceded by a : > These are hotkeys, or macros. I call them functions. > There are functions to perform various tasks. > Most of these functions, when preceded with a number will perform that > function that number of times. 20dd will delete 20 lines, etc. > > > > > These functions cannot be performed in insert mode. So, if you are in > insert mode you’ll need to hit escape first. > Deleting Lines > dd is the function to delete the current line. > If you precede dd with a number it will remove that many lines. From > your cursor down. > Let’s create a file. > vi tmp_file1 > 20i aaaa (enter) > Esc + :w > Now, let’s delete 1 line > Dd > Now, let’s delete 10 lines > 10dd esc + :wq > cat the file, and count the lines. How many are there? > Deleting a specific line > Knowing the line number can be advantageous. You can delete a line > based on it’s number. > For example: to delete line 5, execute: > :5d > Deleting Characters > You can delete characters with the delete key and backspace keys when > insert mode. > In command mode, there is an alternative: x > Why? Sometimes you just don’t want to go to insert mode. > vi tmp_file1 again. Delete the remaining lines (9dd) > Go to insert mode and enter: abcd > Hit escape, and move your cursor to c > Hit x > > > Move your cursor back to b, and hit 2x. What happened? > Deleting a word > Much line x can delete a character , you can delete a word by > highlighting it and executing: > dw > Undo > Yes, we realize that people make mistakes. > u executes the undo command. It will undo your previous text add or > deletion. > Hit u once – what happened? > Hit it again – what happened? > Changing a word > Vi gives you the option to change an entire word without actually > going into insert mode first. > You simply highlight any letter in the word, and enter: > cw > The existing word is replaced, and you are allowed to insert a new one > in its place. > > Changing a letter > VI also gives you the option to change just a letter without going > into insert mode first. > You simply move your cursor to the word you want to replace and > enter: > cl > The existing letter is removed and you are allowed to insert a new one > in its place. > > Yanking (copying) > VI supports functionality similar to that of copy and paste. > To copy an entire line, you type: > yy (yank) > To copy a word, you type: > yw (Yank Word) > To copy a letter, you type: > yl (Yank Letter) > As with all functions, you can precede any of the y’s above with a > number to yank that many lines/words/letters > Pasting > Pasting is accomplished via the following function: > p > Regardless if you have yanked a word, letter, or line p will paste > what is in the buffer. > You can precede p with a number to paste the line that many times. > > > A line will be pasted below your current line > > A word or letter will be pasted after your current cursor position. > > An example > vi paste_test > i for insert mode > Enter the following text: > abcd > 1234 > Move your cursor to the first line and type: > yy > then p > You should now have two lines of abcd. > > > > Example 2 > Delete the second abcd line with dd > Move back to the first line, and type yy. > Move your arrow key down to the second line and hit p. > Your file should now look like: > abcd > 1234 > abcd > You can move your cursor between yanks and pastes. > Example 3 > Delete the second abcd line with dd. > Move back to the first line and type 2yy (yank 2 lines). > Then hit p. Your file will now look like: > abcd > abcd > 1234 > 1234 > Why? Vi pastes lines below your current cursor position. > Hit u, then move down to the bottom of the file and type p, does that > look better? > Example 4 > Delete the extra lines, and move back to the top line. > Enter yy > Move to the bottom of the file and type 20p. > You should now have abcd, 1234, and then 20 more abcd’s > Creating a new line > We can use Shift G and $ to go the the last line and the end of that > line respectively. From there we can hit enter and put a new line > below the last. > Is there a better way? Yes, when you are on a line and you want to > start a new line directly below, you can use the o key to insert a new > line and begin typing. o will automatically put you into insert > mode. > A Capital O will create a new line above the cursor. > Also, capital A will automatically create a new line at the bottom of > the file and put you in insert mode. > Mistakes > So, what if you horribly mess up the file in the process of miskeying > functions? > Easy, all you have to do is quit without saving: > Esc + :q > What happened? Why won’t it let you leave? > > > > Esc + :q! (! Means force) > Reading in another file > So, you’ve opened vi, but you forgot to make a copy of your original > file before editing. > Sure, you COULD leave vi, copy it and open up vi again. > But there are better ways. > > Reading in another file – Way 1 > Way 1 is slightly dangerous in that you actually edit the live file. > You could accidentally overwrite the live file. > This violates Creighton’s rule #1: Always make a backup before you > change something. > So, assume you vi /etc/passwd, and you make changes. But you don’t > want your changes to go live yet. Easy: > Esc + :w filename will write to a new file > Reading in another file – Way 2 > Method 2 involves telling vi to input the contents of another file > with read. > Let’s try this: > vi passwd_copy > Esc + :r /etc/passwd > Esc + :wq > Cat the file – what just happened? > > > You copied the contents of /etc/passwd into your file and saved it as > passwd_copy > VI’ing a directory > This is my favorite feature of vim. Why? Because I’m lazy. > In vim, you can call vi with the argument of a directory. It will > give you a list of files in that directory, and let you pick a file. > Arrow keys and enter select. > Then, you are editing that file. > Try it, type vi /etc/ > What happens? > More navigation > Now that we know the power that numbers hold in front of functions, > get out of insert mode. > Type 15, and then the down key. > You’ve just moved 15 lines down. > See why lazy admins like vi? > View > View is a utility that is bundled with vi. > We used more, head, and tail last week to look at big files. > View gives us another method to do that. > It is basically a read-only version of vi. > You can search, use Shift G, or $, /, etc to navigate through bigger > files. > More tricks > $ moves to the end of a line > :$ moves to the last line of the file (like Shift G) > ^ moves to the beginning of the line you’re on. > w advances a word from the beginning of the word (try 3w) > b moves back one word (try 3b) > A number followed by | moves to that position in the line. > E moves word to word by the end of the word (5e) > Tricks continued > L moves the cursor to the lowest line on the screen > M moves the cursor to the middle line on the screen > H moves the cursor to the first line on the screen. > Screen Scrolling > Ctrl-D moves down one-half screen at a time > Ctrl-U moves up one half screen at a time > Ctrl-F moves down one whole screen > Ctrl-B moves up one whole screen > > _______________________________________________ > freebsd-ports@freebsd.org mailing list > http://lists.freebsd.org/mailman/listinfo/freebsd-ports > To unsubscribe, send any mail to "[EMAIL PROTECTED]"
smime.p7s
Description: S/MIME cryptographic signature
