|
5
|
Function |
Returns |
|
a(x) |
The arc tangent of x |
|
c(x) |
The cosine of x |
|
e(x) |
e raised to the x power |
|
j(n,x) |
The Bessel function of n and x, where n is an integer and x is any real number |
|
l(x) |
The natural logarithm of x |
|
s(x) |
The sine of x |
To use these math functions, you must invoke bc with the -l option, as follows:
$ bc -l
Calculating with dc
As mentioned earlier, the desk calculator, dc, uses RPN, so unless you're comfortable with that notation, you should stick with bc. Also, dc does not provide a built-in programming language, built-in math functions, or the capability to define
functions. It can, however, take its input from a file.
If you are familiar with stack-oriented calculators, you'll find that dc is an excellent tool. It can do all the calculations that bc can and it also lets you manipulate the stack directly.
To display values, you must enter the p command. For example, to add and print the sum of 5 and 9, enter
5 9 +p 14
See your UNIX reference manual (different versions use different titles), or if you have them, view the on-line man pages for details on dc.
Finding the Date and Time
When used by an ordinary user, the date command does only one thing: it displays the date and the time. The system administrator can use date to set the date and the time.
Here's how you use date to find out the current date and time:
$ date Sat Sep 28 1:45:58 EDT 1991
This is the simplest form of executing the date command, but not the most useful. This command has many options that let you extract any part of the usual output and display that part alone or with other parts.
First, look at the individual options. Unlike most UNIX commands, which usually have single letters as options, date's options are strings of characters. The option string begins with a plus sign (+) and each option is preceded with a percent sign (%).
You can include ordinary text in the option string, as you will see in the next example.
TIP: You should enclose the option string in either single or double quotation marks so that the shell won't interpret any characters in the ordinary text portion that it believes are special.
The a option outputs the abbreviated name of the day, and the A option provides the unabbreviated name:
$ date +%a Sat $ date +%A Saturday
The b option outputs the abbreviated name of the month, and the B option provides the unabbreviated name:
$ date +%b Sep $ date +%B September
The d option provides the day of the month in two-digit numeric form, and the e option outputs the day of the month with a space preceding the day for days 1 to 9. In the following example, the date is the 28th day of the month:
$ date +%d 28 $ date +%e 28
If you execute these two commands again early the next month—on the fourth, for example—you would see the following:
$ date +%d 4 $ date +%e 4
The D option outputs the common numerical date format (month/day/year) used in the United States:
$ date +%D 09/28/91
Other options—c, x, and X—output the date in the format for whichever country was specified when SVR4 was installed. If you designated a country other than the United States during installation, try these options on your own; their output
differs for different countries.
The options H and I output the hour in numeric form: H in 24 hour or military form, and I in 12 hour form.
$ date +%H 13 $ date +%I 1
The j option is rather interesting. It outputs the so-called "Julian" date—the day as one of the 365 days of the year (or 366 in a leap year). The following example shows the j option returning the date for September 28:
$ date +%j 271
This option is useful when calculating the elapsed time between two dates.
The U and W options both output the week as one of the 52 weeks of the year (or 53 in a leap year). They differ in that U begins the week with Sunday and W begins the week with Monday. When executed on September 29, 1991, date produces the following:
$ date +%U 39 $ date +%W 39
The m option outputs the month as one of the 12 months of the year:
$ date +%m 09
The M option gives the minutes value in the range of 00 to 59, and the S option shows the seconds value in the range of 00 to 61 (to allow for a "leap second" or two):
$ date +%M 48 $ date +%S 41
The R option combines the H and M options, and the T option combines H, M, and S:
$ date +%R 13:48 $ date +%T 13:48:51
The p option outputs either AM or PM, and r combines I, M, S, and p:
$ date +%p AM $ date +%r 1:48:25 AM
The w option shows the day of the week as a number between 0 and 6, with 0 representing Sunday:
$ date +%w 6
The y option shows the year as a number between 00 and 99, and the Y option shows the year as a four-digit value:
$ date +%y 91 $ date +%Y 1991
The Z option outputs the abbreviated time zone name for your computer. In the following example, the computer is located in the Eastern time zone, but during Daylight Savings Time:
$ date +%Z EDT
You can combine two or more options along with text strings to produce more descriptive outputs, such as the following:
$ date "+Today is %A, the %e of %B, %Y" Today is Saturday, the 28 of September, 1991
Displaying a Monthly or Yearly Calendar with cal
The cal command is very simple but quite handy when you need to see the calendar of any month (or all 12) in any given year. Used with no arguments, cal simply prints a calendar of the current month:
$ cal February 1994 S M Tu W Th F S 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 $
If you specify a year, as in cal 1993, you get a calendar for all 12 months, but if you specify a month and a year, as in cal 4 1992, you get just that month.
CAUTION: If you specify a year as a two-digit number, cal gives you the calendar for that year. In other words, cal 93 produces a calendar for the year 93, not the year 1993.
Historical Note: In 1752 the calendar was adjusted to account for the "discovery" of leap year by eliminating 11 days (September 3 through September 13). Type cal 9 1752 to see the strangest
month of all!
Getting Information About Users
To get information about users (including yourself), you can use several commands. The who command reports on users who are presently logged in, and finger reports on anyone who has an account on the computer. The id command reports information about
the user who invokes it.
The who Command
The who command normally reports certain information about logged-in users. By using its options, you can specify that it report information about the processes initiated by init, and that it report reboots, changes to the system clock, and logoffs.
(See chapters 18 and 35 for more information on processes and init.) If you invoke who with no options or arguments, you get the following output:
$ who juucp tty00 Sep 28 11:13 pjh slan05 Sep 28 12:08
The output shows that two users are currently logged in: the user juucp, who logged in at 11:13, and the user pjh, who logged in at 12:08. Notice that juucp is logged in on a tty line (actually, juucp is a neighboring site that is called in over a
modem) and that pjh logged in over a network (STARLAN, which is shortened to slan in who's output).
NOTE: The term tty is short for teletypewriter, the first kind of terminal to be connected to a UNIX computer system. Even though terminals have advanced significantly, the terminology has not changed. A
"neighboring site" is a computer that is located nearby.
The -u option adds the "time since the last activity" (also called the idle time) and the process ID number for each logged-in user. A "process ID number" or PID is an interger number assigned by UNIX to uniquely identify a given
process (usually, a process is a program that is running). PIDs are needed in UNIX because they allow—yea, encourage—simultaneous running of multiple processes. (See Part IV, "Process Control," for more information.)
$ who -u juucp tty00 Sep 28 11:13 . 5890 pjh slan05 Sep 28 12:08 . 7354
NOTE: If the user has been active within the last minute, who displays a dot (.) for the idle time.
The -T option reports a plus sign (+) if you are able to send messages to the user's terminal, or a minus sign (-) if it is not:
$ who -T juucp + tty00 Sep 28 11:13 pjh + slan05 Sep 28 12:08
The -q (quick) option simply shows login IDs and a count of the logged-in users:
$ who -q juucp pjh # users=2
There's a special case for who: who am i. This case is useful if you are logged in from several terminals under different accounts, and you forget which one you are currently using:
$ who am i pjh slan05 Sep 28 12:08
The finger Command
You can use the finger command with or without arguments. Without arguments, finger's output looks a little like who's:
$ finger Login Name TTY Idle When Where pjh Pete Holsberg pts000 6d Mon 13:03 ajh Alan Holsberg sxt/002 Sat 15:00
This output lists all currently logged users—with a heading line—plus the user's full name and the name of the remote computer (in the "Where" column) if the person is logging in over a network.
This command is more useful if you give it a person's login ID as an argument, as in the following example:
$ finger lam Login name: lam In real life: Pak Lam dp168 Directory: /home/stu/lam Shell: /usr/bin/ksh On since Feb 23 19:07:31 on pts016 from pc2 2 days 21 hours Idle Time No unread mail No Plan.
Here, finger displays personal information for the user lam: his real name, home directory, which shell he uses, and when he last logged in and from which computer (pc2). The last line indicates that he does not have a text file called .plan in his home
directory. The finger command displays the contents of .plan and .project if they exist. As you can see, users can reveal as much or as little of themselves as they choose.
If your computer is on a network and you know another person's e-mail address, you can display information about that person by issuing a command such as the following:
$ finger holsberg@pilot.njin.net
The finger command provides many options that can suppress one or more fields of the normal, long output. However, because finger responds so quickly, simply disregarding that extra information when it appears onscreen is easier than learning the
appropriate option.
The id Command
The id command reports four things: the user ID number, login name, group ID number, and group name of the person who invokes it. If the real and effective IDs (see Chapter 44, "System Security") are not the same, id prints both sets of
values.
NOTE: id is often used in shell scripts to restart the use of a shell by a certain user or group of users.
When a system administrator creates an account for a user, that user is given a login ID and also placed in a group. This is important because UNIX provides access to files according to whether the user is the owner of the file and whether the user belongs to the group that has been granted access to the file.
In the following example, the user ID is 102, the login name is pjh, and the user belongs to the root group, which is group number 0:
$ id uid=102(pjh) gid=0(root)
Switching Accounts with su
If you have more than one account on a system, you need not log out of one and then log into the second to use the second account. Instead, you can simply use the su (switch user) command. The system administrator uses su frequently to check a user's
complaint, because su enables the administrator to become that user and run the programs that were causing problems.
The usual su command syntax is
su - userID
The minus sign tells su to cause the current user to take on the identity, userID. To do this, su executes the user's shell—as specified in the shell field of /etc/passwd—and invokes /etc/profile and that user's .profile file. The su
command then displays the user's normal environment. If the user omits the minus sign, su invokes the user's shell, but the environment is the same as before. Of course, before you can switch to another user, you must know that user's password—unless
you are the system administrator, in which case your privileges enable you to assume identities without using passwords.
NOTE: /etc/profile and .profile are files that control the working environment of each user.
Suppose, the system administrator wishes to become user lam:
#su - lam $
To reassume identity as sysadm, he merely logs out of lam's account.
Learnig More About Commands with man
UNIX supplies information about its commands in two forms: reference manuals (printed documentation called man pages) and online files. To use the man pages, you simply find the appropriate reference manual and look up the command. However, reference
manuals seem to have lives of their own and are never to be found where you last put them. That's when online man pages—and the man command—become useful.
The simplest form of the man command takes a single argument: the name of the command in which you are interested. For example, here's the man page for man:
NAME cat - concatenate and print files SYNOPSIS cat [-u] [-s] [-v [[-t] [-e]] file . . . DESCRIPTION cat reads each file in sequence and writes it on the standard output. Thus cat file prints the contents of file on your terminal, and cat file1 file2 >file3 concatenates file1 and file2, and writes the results in file3. If no input file is given, or if the argument - is encountered, cat reads from the standard input. cat processes supplementary code set characters according to the locale specified in the LC_CTYPE environment variable [see LANG on environ(5)]. The following options apply to cat: -u The output is not buffered. (The default is buffered output.) -s cat is silent about non-existent files. -v Causes non-printing characters (with the exception of tabs, new-lines, and form-feeds) to be printed visibly. ASCII control characters (octal 000 - 037) are printed as ^n, where n is the corresponding ASCII character in the range octal 100 - 137 (@, A, B, C, . . ., X, Y, Z, [, \, ], ^, and _); the DEL character (octal 0177) is printed ^?. Other non-printable characters are printed as M-x, where x is the ASCII character specified by the low-order seven bits. All supplementary code set characters are considered to be printable. The following options may be used with the -v option: -t Causes tabs to be printed as ^I's and formfeeds to be printed as ^L's. -e Causes a $ character to be printed at the end of each line (prior to the new-line). The -t and -e options are ignored if the -v option is not specified. FILES /usr/lib/locale/locale/LC_MESSAGES/uxcore.abi language-specific message file [See LANG on environ(5).] SEE ALSO cp(1), pg(1), pr(1) The NAME section gives the name of the command and a brief description of what it does. The SYNOPSIS section shows the command's syntax by listing all possible options and arguments. Arguments to cat are the names of one or more files. The fact that file is written in italic type means that file is merely a placeholder for the name of an actual file. The cat command has three options and one of them has two "sub-options". The brackets around options indicate that they are optional. For example, the command could be any one of the following: cat file cat -u file cat -s file cat -u -s file (or cat -s -u file) cat -us file (or cat -su file) cat -v file cat -v -t file (or cat -vt file) cat -v -e file (or cat -ve file) cat -v -t -e file (or cat -vte file) cat -u -v -t file (or cat -uvt file) and so on. With three options and two suboptions, there are many possible combinations. Note that the order of writing the options doesn't matter and options can be combined so that only one dash (-) need be written. While these are generally true, there are some Unix commands that require a particular order, some that do not permit combinations and even a few that require a specific order and deny any combinations. The DESCRIPTION section is the heart of the manual page. It describes each option and argument, and frequently contains one or more examples. However, its explanations are frequently sprinked with referenced to other commands, so that in general it is difficult to learn about a command from its man page. The FILES section lists other files that are used or related to the command. The SEE ALSO section lists related commands. The numbers in parentheses refer to the section of the manual in which the command will be found. Among the most important are: Section 1 - User Commands Section 1C - Basic Networking Commands Section 1M - Administration Commands Section 2 - System Calls Section 3 - BSD Routines Section 3C - C Library Functions Section 3M - Math Library Functions Section 3S - Standard I/O Functions Section 4 - File Formats Section 5 - Miscellaneous Section 6 - (not included) Section 7 - Special Files Section 8 - System Maintenance Procedures
As a user or adminstrator, you will be most interested in section 1 (which contains 1, 1C and 1M). Programmers will be interested in 2, 3 and possibly 4.
Some man pages have the following additional sections:
The NOTES section is self-explanatory.
The DIAGNOSTICS sections contains explanations of error messages that the command may cause to be displayed. Not enough commands have a DIAGNOSTICS section!
The WARNINGS section describes anything that may limit the use of a command.
The BUGS section tells of known problems that have not been fixed in the current version of the command.
Finding Information About Disk Utilization with du and df
Occasionally, a user needs to know how much disk space his or her files occupy—perhaps the system administrator has sent e-mail requesting that the user is approaching some sort of limit. Frequently, the system administrator needs the same
information for all users or on all parts of one or more disks, and frequently needs information on the overall utilization in all file systems. UNIX provides two commands that report that information: du and df.
Summarizing Disk Usage with du
The du command reports the number of disk blocks used for each directory and subdirectory, and the files found there. It has an option to display the number of blocks of just the "top level" directory.
NOTE: A block is the smallest piece of a disk that can be allocated to a file. Frequently, the size of a block is 512 bytes (but the size depends on things that the system administrator did when setting up
the disks). A file that is 1 byte long and a file that is 511 bytes long each occupies one block. A 513 byte file and a 1,023 byte file each occupies two blocks. That is, every 512 bytes of a file occupy a disk block, plus one block for the last piece that
less than 512 bytes.
Here are some sample executions. First, issue the command to display what you have in the directory hierarchy, from the current directory on down:
$ ls -lR total 259 drwxr-xr-x 2 pjh other 48 Oct 17 1990 unixbin -rw-r—r— 1 pjh root 130137 May 10 1993 vpix.img ./unixbin: total 16 -rw-rw-rw- 1 pjh other 7576 Apr 8 1993 rununix.exe
The current directory has one regular file (vpix.img) and one directory file (unixbin), and the subdirectory has one regular file (rununix.exe).
Now find out the total disk usage:
$ du -s 274 .
That's 274 blocks, or somewhere between 13,312 and 13,823 bytes.
Use the following command to find out how many blocks are in the subdirectory:
$ du 17 ./unixbin 274 .
Then find out how many blocks each file has:
$ du -a 16 ./unixbin/rununix.exe 17 ./unixbin 256 ./vpix.img 274 .
NOTE: The total for a directory or directory hierarchy includes the blocks that the directory files themselves occupy.
Reporting Blocks and Files with df
The df command helps the system administrator maintain the file systems on the disks. The command displays information about mounted and unmounted resources, types of file systems, and numbers of inodes. It also deals with special devices rather than
easily identifiable directory paths. Most of these terms have little meaning to the ordinary user, but they are discussed in Chapter 35. However, if a user wants to see what's going on with the disk's file systems, the system permits it.
Multiple options are available with df. In this section, you'll see some of these options demonstrated and explained. But first, here is a typical result that you get when you invoke df with no options:
$ df / (/dev/root ): 381964 blocks 58845 files /proc (/proc ): 0 blocks 128 files /dev/fd (/dev/fd ): 0 blocks 0 files /stand (/dev/dsk/c0t0d0sa): 12643 blocks 194 files /var (/dev/dsk/c0t0d0sb): 961502 blocks 64499 files /tmp (/dev/dsk/c0t0d0sd): 191308 blocks 24560 files /usr (/dev/dsk/c0t1d0s3): 772900 blocks 53842 files /home (/dev/dsk/c0t1d0s4): 923256 blocks 63303 files /.NetWare (/.NetWare ): 0 blocks 499 files
The first column is the path name of the file system. This system's special device name appears in the second column. The third column reports the number of disk blocks available on each disk partition. Finally, the last column displays the number of
files that can be added to each partition. When setting up a new disk, the system administrator partitions it into file systems and allocates several blocks and inodes (or files) to each file system according to how that partition will be used. For
details, see Chapter 35, "File System Administration."
The -b option shows the free space in kilobytes rather than blocks:
$ df -b Filesystem avail /dev/root 190982 /proc 0 /dev/fd 0 /dev/dsk/c0t0d0sa 6321 /dev/dsk/c0t0d0sb 480751 /dev/dsk/c0t0d0sd 95654 /dev/dsk/c0t1d0s3 386450 /dev/dsk/c0t1d0s4 461628 /.NetWare 0
The -k option shows some details about allocated space—how much has been used and how much is available—in kilobytes:
$ df -k filesystem kbytes used avail capacity mounted on /dev/root 316416 125434 190982 40% / /proc 0 0 0 0% /proc /dev/fd 0 0 0 0% /dev/fd /dev/dsk/c0t0d0sa 10240 3918 6321 38% /stand /dev/dsk/c0t0d0sb 509952 29201 480751 6% /var /dev/dsk/c0t0d0sd 102368 6714 95654 7% /tmp /dev/dsk/c0t1d0s3 512000 125550 386450 25% /usr /dev/dsk/c0t1d0s4 512976 51348 461628 10% /home /.NetWare 0 0 0 0% /.NetWare
If you are curious about file systems, the -n option tells you the name of the file system type used for each partition. (Again, see Chapter 35 for details.)
$ df -n / : vxfs /proc : proc /dev/fd : fdfs /stand : bfs /var : vxfs /tmp : vxfs /usr : vxfs /home : vxfs /.NetWare : nucam
Finally, the -t option shows used blocks and files, and the total allocation for each file system:
$ df -t / (/dev/root ): 381964 blocks 58845 files total: 632832 blocks 65520 files /proc (/proc ): 0 blocks 128 files total: 0 blocks 202 files /dev/fd (/dev/fd ): 0 blocks 0 files total: 0 blocks 50 files /stand (/dev/dsk/c0t0d0sa): 12643 blocks 194 files total: 20480 blocks 200 files /var (/dev/dsk/c0t0d0sb): 961502 blocks 64499 files total: 1019904 blocks 65536 files /tmp (/dev/dsk/c0t0d0sd): 191308 blocks 24560 files total: 204736 blocks 24576 files /usr (/dev/dsk/c0t1d0s3): 772900 blocks 53842 files total: 1024000 blocks 65536 files /home (/dev/dsk/c0t1d0s4): 923256 blocks 63303 files total: 1025952 blocks 65536 files /.NetWare (/.NetWare ): 0 blocks 499 files total: 0 blocks 500 files
Summary
UNIX provides several small programs that you can use to quickly obtain information that you frequently need. These programs include the following:
- bc, dc—calculator programs
- date and cal—to display time and date
- finger, who and id—provide user information
- man and apropos—explains commands
- su—"switch user"
- du and df—display information about disk contents
These commands are rather specialized commands that, in general, retrieve information about different parts of the system: users, files, directories, clock, and so on. The basis of information storage in UNIX is the file, so Chapter 6, "Popular File Tools," explores the somewhat more specialized commands for working with files—particularly text files.
