grep: finding lines

grep PATTERN FILE reads the file line by line and prints every line that contains the pattern. That is the whole idea. The name comes from the old ed editor command g/re/p — globally search for a regular expression and print — and the regular-expression part is where its power comes from, but you can get a long way treating the pattern as plain text. This lesson does exactly that; lessons 33 and 34 add the regex syntax.

One file, several files

With a single file, grep prints the matching lines as they are. With several files, or a wildcard like *.txt, it prefixes each line with the file name and a colon, so you know where each hit came from. -h hides that prefix and -H forces it even for a single file. The match is case-sensitive and happens anywhere in the line: grep cat matches concatenate.

~% cd /root/lab/l31
l31% cat fruits.txt
apple
banana
grape
pineapple
cherry
l31% grep apple fruits.txt
apple
pineapple
l31% grep an fruits.txt notes.txt
fruits.txt:banana
notes.txt:bananas are cheap
l31% grep /bin/false /etc/passwd
daemon:x:1:1:daemon:/usr/sbin:/bin/false
bin:x:2:2:bin:/bin:/bin/false
sys:x:3:3:sys:/dev:/bin/false
mail:x:8:8:mail:/var/spool/mail:/bin/false
www-data:x:33:33:www-data:/var/www:/bin/false
operator:x:37:37:Operator:/var:/bin/false
nobody:x:65534:65534:nobody:/home:/bin/false

grep in a pipe

Given no file, grep reads stdin, which makes it the universal filter: ls /bin | grep grep lists the commands whose name contains grep, ps | grep sh shows the shells running, history | grep ssh finds that command you typed last week. Any output that is too long to read becomes manageable once you grep it for the piece you care about.

Quote the pattern

The shell processes the command line before grep sees it. A pattern with a space would be split into two arguments, * would expand to file names, and $ would start a variable. Wrap patterns in single quotes whenever they contain anything other than letters and digits: grep 'No such file' log.txt. It costs nothing when the quotes were unnecessary and saves you from strange errors when they were.

The exit status

grep exits with 0 when it found at least one match, 1 when it found none, and 2 on an error such as a missing file. That turns it into a test: grep -q root /etc/passwd && echo 'root exists'. The -q (quiet) option prints nothing and stops at the first match, which is exactly what you want when only the yes/no answer matters. You will use this constantly in scripts.