Home / ubuntu / Questions / 693443
Accepted
user308393
Asked: 2015-11-04 06:01:09 +0800 CST

grep command to find 'words' that begin and end with * (e.g., *way*, *then*)

  • 772

My original thinking was that this would work

grep -w '\*.*\*' /path/file-name

This produces some good hits but it doesn't respect the word option. I get whole lines. For example, like this

[bold]*way* to the store to buy some groceries and *then*[bold]

and I want to get this

[bold]*way*[bold] to the store to buy some groceries and [bold]*then*[bold]

Another one I get is this, which I thought wouldn't appear because it has spaces (which should disqualify it because I want only words or stuff separated by spaces)

* * *
grep

1 Answers

  1. Best Answer

    You need to make the match non-greedy: in a basic (BRE) or extended (ERE) regular expression you can enforce that by using an exclusive character set [^*] in place of the inclusive .* i.e.

    grep '\*[^*]*\*'

    or

    grep '[*][^*]*[*]'

    If you a prefer perl-compatible regex, that has an explicit non-greedy modifier ?

    grep -P '\*.*?\*'
    • 2