Bash - 16 - egrep

grep stands for “global/regular expression/print.”
It is a powerful tool for searching text and is commonly used in UNIX text editors.

grep program searches one or more files by line to match against a specified regular expression. Each line is treated as the string and grep searches for any substring of the line that matches the regex. If found, by default grep outputs the matching lines.

grep only applies the POSIX basic regular expression set and extended regular expressions set egrep is used.
egrep can do everything grep can do, there is no need to use anything other than egrep.

egrep [options] regex filename(s)

regex is a string that can but does not have to include metacharacters.
Files can be listed with space or wildcards.

[!important] The regex is placed in single quotes ' ' to avoid metacharacters being interpreted as wildcards and doing filename expansion (globbing) on wildcards * ? [].
| will sets up pipes instead of performing OR action. Items Found in " " are interpreted, they are not treated literally like in ' '

To search for a specific phrase using quotes makes it easier to search for phrases or single words:

$ egrep "is not" haiku.txt

$ egrep "not" haiku.txt

egrep 2022 *.txt  
# all lines that contain 2022

egrep ^a *.txt
# all lines that start with an a

egrep '[Ss]mith' *.txt
# Find lines having smith or Smith

egrep 'Duke|Zappa' *.txt
# Find lines containing Duke or Zappa

egrep '[0-9]+' *.txt
# Find lines that contain atleast one digit

egrep '[A-Za-z]+[[:punct:]]' *.txt
# Find lines that contain letters followed by any punctuation mark anywhere in the line. 

egrep '^[0-9]*[^0-9]+$' *.txt
# Finds lines that if they have digits are found at the beginning of the line.

egrep '^[A-Z][a-z]+ [A-Z][a-z]+$' *.txt
# Find all lines that contain exactly two words, both capitalized

egrep '[a-z]+ [a-z]+ [a-z]+' *.txt
# Find lines that contain at least 3 words, lowercased

Options

• Recursive Search (-r): Searches through all files in the directory and subdirectories.

$ egrep -r "Yesterday"

• Count (-c): Gives the count of the matches for each file instead of a list.

$ egrep -c '[0-9]{1,3}' /etc/resolve.conf

• Word Boundary (-w): Limits matches to whole words only.

$ grep -w "The" haiku.txt

• Line Numbers (-n): Numbers the results with the line numbers.

$ grep -n "it" haiku.txt

• Case Insensitivity (-i): Makes the search case insensitive.

$ grep -n -w -i "the" haiku.txt

• Invert Match (-v): Inverts the search to get all lines that do not match.

$ grep -v -n -w "the" haiku.txt

Regex can be placed in a file and -f option can be used to reference the file.

-e and -f ???

The outputs of ls are piped to egrep, Outputs are all regular files whose owner permissions are rwx. ^ starts the regex, since $ is not given, doesn’t match whole string.

ls -l /etc | egrep '^-rwx'

???

ls -l /etc | egrep '.{13}[^1]'