I try to search for lines that start with "1" using

ls -1 | grep ^1*

but it returns lines that do not start with 1. What I am missing here?

  • Just another note: If you want to use grep to limit ls output, you will run into problems if filenames contain some weird characters, such as newlines... Many other commands (somehow not ls, at least not the version that I have to check) have a option to give null separated output, which you can then safely process with grep -z Dec 31, 2012 at 6:34
  • Thanks, @GertvandenBerg! What problems can some weird characters such as newlines cause to grep? What commands give null separated output?
    – Tim
    Dec 31, 2012 at 12:32
  • 1
    find with -print0, most other GNU tool have a -0 or -z option. (sort, xargs, etc as well). If the filenames contain newlines, the it would be impossible to know if two lines are a filename containing a newline or two seperate filenames. Jan 2, 2013 at 6:48

4 Answers 4


Your regular expression doesn't mean what you think it does. It matches all lines starting (^) with one (1) repeated zero or more (*) times. All strings match that regular expression. grep '^1' does what you want.

  • 4
    or likewise this works: grep "^$some_var" Feb 29, 2020 at 21:18

Did you try the following?

 ls -1 | grep "^1"

That is, remove the *, which basically tells grep, find zero or more occurances of the ^1 expression. In other words: match the lines that start with a 1, or not.


Although this doesn't answer your question, this is a better solution to what appears to be your goal:

ls -ld 1*

You can use a shell glob to list all files that start with 1. Note that * has a different meaning in shell globbing than regular expressions.

  • Thanks! +1. How does * have different meaning in shell globbing and in regular expressions?
    – Tim
    Dec 31, 2012 at 3:33
  • 2
    In a shell glob * is a wildcard meaning "0 or more of any character". The regex meaning is explained in otokan's answer.
    – jordanm
    Dec 31, 2012 at 5:10

^1.* matches the whole line, or just as said above, ^1 got the string contained in the line.

different styles of regex uses different symbols representing chars, some options specify which style you want. and different options specify whether you want to match the whole line, or just part of it, or the whole input as one string.

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .