egrep is equivalent to grep -E 1. It interprets the pattern as an "extended regular expression". In extended mode, some backslashes to escape certain characters are removed 2:
Pattern with grep |
Pattern with egrep |
|---|---|
abc\{3\} |
abc{3} |
a\(e\)+g |
a(e)g |
It looks like a good idea to always use egrep or grep -e first, since the regular expressions I've alredy used (for example in JavaScript) are extended regular expressions and not basic regular expressions.
fgrep is equivalent to grep -F 1. It interprets the pattern as a list of fixed strings, separated by new lines, any of which is to be matched.
For example, if you have a file users.txt with a list of users (one user on each line) and want to search the group file to see if any of the users are in it, you would use:
$ grep -F -f users.txt /etc/group
Sources: