When you enter commands line by line (pressing Return
after each command), you are telling the system to complete the
command (or program) before executing the next command. Executing:
will complete each command before going on to the next. You
can place all of the commands on the same line by using the ";"
separator. For example,
is equivalent to entering each command on a separate line.
This process is called sequential processing.
New programs or commands cannot be started until the preceding program
or command has completed.
If parameters are required by the program, they are entered
as usual. The semicolon is placed after the last parameter.
While a program is running as a sequential process, there
is no response to keyboard activity until after the program has
completed (other than the keyboard buffer delay).
Programs already in progress when a program with sequential
processing is executed continue to run as usual. While a program
is running as a sequential process, you have the option of waiting
for the program to finish.
Nonsequential (Background) Processing |
 |
Programs can also be run nonsequentially,
in which case, each program runs without waiting for the previous
program to complete. This type of execution is more commonly called
"running in the background." Follow the program
name with &
to specify background processing.
program1 & program2 & program3 & |
This example runs program1,
program2, and
program3, in
the background, and returns a prompt to the user immediately.
Note that programs that write to the terminal or require input
are poor choices for background execution, since the output will
be intermixed on the screen, or the input may not be read by the
correct program.
Redirecting Input and Output |
|
Every program has at least three data paths
associated with it: standard input, standard output, and standard
error output. Programs use these data paths to interact with you.
By default, standard input (stdin)
is your keyboard. The default destination for both standard output
(stdout) and
standard error (stderr)
is your screen.
Redirecting input and output is a convenient way of selecting
what files or devices a program uses. The output of a program that
is normally displayed on the screen can be sent to a printer or
to a file. Redirection does not affect the functioning of the program
because the destination of output from the program is changed at
the system level. The program is unaware
of the change.
I/O redirection enables you to change a specific data path
of a program while leaving its other data paths unchanged. For example,
stdout can be
stored in a file instead of written to your screen.
How to Redirect Input and Output
I/O redirection symbols are entered on the shell command line,
or from a shell program. The program begins executing with the data
paths specified by the redirection symbols. To specify I/O redirection
for a program, each file name is preceded by a redirection symbol,
as in:
programA < filename
programB > filename |
Spaces between the redirection symbols and the file names
are optional. The symbol identifies the name that follows it as
a file for input or output. The redirection symbols are listed in
Table 3-1 “Redirection Symbols”.
Table 3-1 Redirection Symbols
Symbol | Function | Example |
|---|
< | Read standard input
from an existing file. | program1 < input.data |
> | Write standard output
to a file. | program2 > output.data |
>> | Append standard output
to an existing file. | sample.prog >> output.data |
|
| |
|
 | NOTE: Using >
destroys any previous contents of the file specified to receive
the output. If a file's contents must be preserved, use >>.Be careful not to use the same file for standard input
and standard output. When input and output operations access the
same file, the results are unpredictable. |
|
| |
|
If a file you specify with a redirection symbol is not in
the current directory, you should use a path name to identify it.
The following actions are taken when the system does not locate
files named with the redirection symbols:
If a file specified for input with
the < symbol
is not located, an error message is displayed.
If a file specified for output with the >
or >> symbol
is not located, it is created and used for program output.
The following examples show how the data paths of programs,
commands, or utilities can be modified with the redirection symbols.
Runs the program CHItest
using the file data1 as input.
Adds the current time and date to the end of the file syslog.
Pipes |
|
Two or more programs or commands can be connected so the output
of one program is used as the input of another program. The data
path that joins the programs is called a pipe.
Pipes allow you to redirect program input and output without the
use of temporary files.
When programs are connected with pipes, the shell coordinates
the input and output between the programs. The pipes only transfer
data in one direction, from the standard output of one program to
the standard input of another program.
How to Connect Programs With Pipes
The vertical bar (|)
is the "pipe" symbol. Parameters for the program
are listed after the program name, but before the |
symbol. Spacing between the program names and vertical bars is optional.
The syntax used for connecting programs with pipes is as follows:
program-a | program-b | program-c
where program is a command or executable
program. Pipes operate on or transform data by separate programs
in stages. For example, program-a might
require input that you type from the keyboard. program-a
could collect this data and then direct it to stdout.
This output would be passed through the first pipe to become the
input to program-b. program-b
might check that data for validity and process it in some way, perhaps
sort it. The processed data would then go to stdout
and be passed through the second pipe to become the input to program-c.
program-c might format that input into
a report.
Here are some examples.
To print the number of files in the current directory, type:
To print a listing of each file in the directory, and paginate
it for convenient screen viewing, type:
To send the contents of file
to pr, which
formats the data and then passes it to lp
for printing on the line printer, type:
Redirection in Pipes |
|
The redirection symbols can be used for programs connected
with pipes. However, only the data paths not connected with pipes
can be changed. If you specify a change to a data path being used
with a pipe, then an error occurs. The following changes are permitted:
The standard input of the first program
using a pipe can be redirected with the <
symbol.
The standard output of the last program using a
pipe can be redirected by using the >
symbol or appended to an existing file with the >>
symbol.
The following commands show how programs can be connected
with pipes and how additional changes can be made to data paths
with redirection symbols.
The first example takes the standard output from test_prog1
and uses it as standard input to /usr/output_prog.
test_prog1 | /usr/output_prog
The next example runs four programs connected with pipes and
puts the output of the fourth program in store_file.
get_it | check_it | process_it | format_it > store_file
Pipe Example |
|
The following pipe uses several of the symbols we just discussed.
Try to figure out what will happen before you read the description
below. (The backslash (\) at the
end of the first line concatenates the two lines into one.)
sort pdir; ( ( pr pdir | lpr )& \
(sort local)& ); cat local >>pdir |
This pipeline will run three sets of commands sequentially.
The first command is to sort the pdir
file. When it is completed, the second command set is executed.
The parentheses separate the commands so the shell knows which command
to associate with a symbol (for more on command grouping, see “Command Grouping”). Therefore, the
two commands (pr
and sort) are
run nonsequentially. So, at the same time, the pdir
file is formatted and sent to the printer, and the local
file is sorted. Finally, the cat
command is run which appends the local
file to the pdir file.
File Name Generation |
|
A helpful way to reduce typing is to use patterns to match
file names. If you are in a directory with a file "programming"
you can see a listing with either:
or you can use a pattern to match:
where "*"
will match any character or string of characters. If you have another
file beginning with "p",
it too will be listed. Table 3-2 “File Generation Symbols”
shows the file generation symbols you can use:
Table 3-2 File Generation Symbols
Symbol | Description |
|---|
* | Matches any string of characters including
the null string. |
? | Matches any single character. |
[ ...
] | Matches any one of the characters enclosed
in the brackets. A pair of characters separated by a minus will
match any character between the pair (lexically). |
will match a file name that begins with any character a
through z (lower
case), followed by any single character, followed by the string
"cubit",
followed by any number of characters, and that ends in ".c"
or ".a".
Printable version
