Why Direct Invocation of egrep or fgrep is Deprecated – Grep Usage

• egrep is 100% equivalent to grep -E
• fgrep is 100% equivalent to grep -F

Historically these switches were provided in separate binaries. On some really old Unix systems you will find that you need to call the separate binaries, but on all modern systems the switches are preferred. The man page for grep has details about this.

As for what they do, -E switches grep into a special mode so that the expression is evaluated as an ERE (Extended Regular Expression) as opposed to its normal pattern matching. Details of this syntax are on the man page.

-E, --extended-regexp
Interpret PATTERN as an extended regular expression

The -F switch switches grep into a different mode where it accepts a pattern to match, but then splits that pattern up into one search string per line and does an OR search on any of the strings without doing any special pattern matching.

-F, --fixed-strings
Interpret PATTERN as a list of fixed strings, separated by newlines, any of which is to be matched.

Here are some example scenarios:

• You have a file with a list of say ten Unix usernames in plain text. You want to search the group file on your machine to see if any of the ten users listed are in any special groups:

  grep -F -f user_list.txt /etc/group
  

  The reason the -F switch helps here is that the usernames in your pattern file are interpreted as plain text strings. Dots for example would be interpreted as dots rather than wild-cards.

• You want to search using a fancy expression. For example parenthesis () can be used to indicate groups with | used as an OR operator. You could run this search using -E:

  grep -E '^no(fork|group)' /etc/group
  

  ...to return lines that start with either "nofork" or "nogroup". Without the -E switch you would have to escape the special characters involved because with normal pattern matching they would just search for that exact pattern;

  grep '^no\(fork\|group\)' /etc/group
  

grep -i 'a' is equivalent to grep '[Aa]' in an ASCII-only locale. In a Unicode locale, character equivalences and conversions can be complex, so grep may have to do extra work to determine which characters are equivalent. The relevant locale setting is LC_CTYPE, which determines how bytes are interpreted as characters.

In my experience, GNU grep can be slow when invoked in a UTF-8 locale. If you know that you're searching for ASCII characters only, invoking it in an ASCII-only locale may be faster. I expect that

time LC_ALL=C grep -iq "thats" testfile
time LC_ALL=C egrep -q "[tT][hH][aA][tT][sS]" testfile

would produce indistinguishable timings.

That being said, I can't reproduce your finding with GNU grep on Debian jessie (but you didn't specify your test file). If I set an ASCII locale (LC_ALL=C), grep -i is faster. The effects depend on the exact nature of the string, for example a string with repeated characters reduces the performance (which is to be expected).