CSE 391

Shell commands

Streams, Redirection

​

Slides created by Josh Ervin and Hunter Schafer. �Based off slides made by Marty Stepp, Jessica Miller, Ruth Anderson, Brett Wortzman, and Zorah Fung

​

AGENDA

• Logistics
• Useful shell commands (wc, more, less, grep)
• Standard in, Standard out
• Input/output redirection
• Pipes

Logistics

• All the important content for the week is in the videos and these slides

FILE EXAMINATION

SEARCHING AND SORTING

You’re working on a project and have been leaving comments with the text “TODO” near all the things you still need to finish. Your project is stored in the project directory in your current directory.

Write a command to find and print all of the lines that have a TODO on them in the project folder.

Hint: Recall for HW1 we taught you how to use the “wildcard” to select all files as part of a path.

$ grep “TODO” project/*

JAVA AND THE COMMAND LINE

What is the command to compile all Java files in the current directory?

​

Hint: How can you use wildcards to find all Java files?

$ javac *.java

STANDARD STREAMS

• Every unix process has three streams, which are abstract locations that tell a program where to read input from and where to write output to.
• There are three standard streams:
• stdin (Standard Input)
• stdout (Standard Output)
• stderr (Standard Error)
• You’ve likely already seen this before when writing a Java program, the System.out in System.out.println is referring to stdout
• By default, all of these default to the console (they print to the terminal and read from user input into the terminal). However, this can be easily changed.

STANDARD STREAMS

Process

stdin

stdout

stderr

A program such as ls, cd, or grep

These default to the console (terminal)

STDIN VS PARAMETERS

• One of the most important distinctions in this class is the difference between stdin and a command’s parameters.
• A parameter is an argument you give on the command line, like so
• $ ls dir1
• dir1 is a parameter, it does not come from standard input
• Standard input comes from the user, either from a file or from the console
• $ grep “a”
• Once you type this command, it accepts input from your keyboard until you close the stream using Ctrl + D

OUTPUT REDIRECTION

command > filename

• Execute command and redirect its standard output to the given filename
• If the file does not exist, create the given file.
• If the file does exist, it will overwrite the given file (BE CAREFUL!!)
• To append to a file instead of overwrite it, use >> instead of >
• Examples:
• Output contents of current directory to files.txt: ls -l > files.txt
• Append output of wc -l veggies.txt to files.txt: wc -l veggies.txt >> files.txt

INPUT REDIRECTION

command < filename

• Execute command and read its standard input from the contents of filename instead of from the console.
• If a program usually accepts from user input, such as a console Scanner in Java, it will instead read from the file.
• Notice that this affects user input, not parameters.

Write a command to store all of the lines in fruits.txt that contain the letter a into a file called a.txt

​

$ grep “a” fruits.txt > a.txt

STDIN VS PARAMETERS: JAVA

// Read and print input from stdin

public static void main(String[] args) {

Scanner console = new Scanner(System.in);

while (console.hasNext()) {

System.out.println(console.next());

}

}

// Read and print the parameters

public static void main(String[] args) {

for (int i = 0; i < args.length; i++) {

System.out.println(args[i]);

}

}

STDERR REDIRECTION

command 2> filename

• Execute command and redirect its standard error to the given filename

command 2>&1

• Execute command and redirect standard error to standard output

command > filename 2>&1

• Execute command, redirect standard error to standard output, and redirect standard output to filename

​

PIPES

command1 | command2

• Execute command1 and send its standard output as standard input to command2.
• This is essentially shorthand for the following sequence of commands:

command1 > filename

command2 < filename

rm filename

• This is one of the most powerful aspects of unix - being able to chain together simple commands to achieve complex behavior!

Suppose we have a file berries.txt where each line is the name of a different berry. Write a command that outputs how many berries have names that contain both the letter a and the letter e.

​

​

$ grep “a” berries.txt | grep “e” | wc -l

COMBINING COMMANDS

command1 ; command2

• Execute command1, then execute command2.

​

command1 && command2

• Execute command1, and if it succeeds, then execute command2.

​

LOGS

• A common exercise in daily software development and operations is looking at log files - basically a status report of what is going on inside the program.
• We can look at the logs for all the CSE course websites by reading the file: /cse/web/courses/logs/common_log
• For example, to actively watch the log file and only look for access to our own course website, we could use the following

$ tail -f /cse/web/courses/logs/common_log | grep “391”

​