Super User is a question and answer site for computer enthusiasts and power users. It's 100% free, no registration required.

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I use that notation at my log files:


somethings are necessary and variable length characters and numbers. My special log file includes line that has characters explained above. I.e.

Line 1: DEBUG [AS-kmksdf] Error occured!
Line 2: something somethingg..

I want to get Line 1. My log files grows dynamically and I show it at my screen with:

tail -f log.txt

How can I show just that special lines?

I know that I will use grep and my regex will be sth. like


how to combine them with that command. This doesn't work:

tail -f log.txt | grep [[A-Z]*-[A-Z]*]

I use cygwin with windows7 and can run some linux commands on my cmd.exe.

share|improve this question
Look at grep. – soandos Jun 20 '12 at 11:21
Your edit made this a completely different question, and it is still not precisely described what you actually want. No developer with experience is willing to work towards such "requirements"... – DevSolar Jun 20 '12 at 11:56
up vote 2 down vote accepted

There are a couple of problems with your regular expression:

  • First of all, I'd advice you to always use the -P switch, which enables Perl Compatible Regular Expressions.

    By default, grep uses Basic regular expression, which aren't used widely and suffer from inconsistent espacing.

  • As is [A-Z], brackets have special meaning in regular expressions. To use a literal bracket in PCREs, escape it with a backslash: \[ or \]

  • Regular expressions are case sensitive, unless you set the -i switch.

  • * means any number of occurrences, including (0). You might want to use + instead (any positive number of occurrences).

  • Always surround your regular expression with double quotes, or the bash will expand characters.

As a result, the following command will work:

tail -f log.txt | grep -Pi "\[[A-Z]+-[A-Z]+\]"
share|improve this answer
I have edited my question. I realized another issue for me. – kamaci Jun 20 '12 at 11:47
@kamaci: Your edit requires a completely different approach, so it's an entirely different question. Instead of editing it, it's preferable to ask a new question. – Dennis Jun 20 '12 at 11:49
Oops my grep command has no -P option. Maybe because of cygwin. – kamaci Jun 20 '12 at 11:50
That's probably a bug. You could either install pcregrep or use @crizot's approach. – Dennis Jun 20 '12 at 11:59
I have rollbacked my question and new question is:… – kamaci Jun 20 '12 at 12:11
tail -f log.txt |grep -E '\[[a-zA-Z]*-[a-zA-Z]*\]'

You can use Regexpal to test your regex expressions. Also you should use -E on grep. It is not mandatory, and not necessary in this case, but allow you to use extended regex expressions.

share|improve this answer
Recursive? Why? – DevSolar Jun 20 '12 at 11:25
Sory. I edit my answer, it is -E. It is not mandatory but I use -E as a rule because usualy I work with extended regex expressions. In this case it is not necessary. – criziot Jun 20 '12 at 11:30
that's nice. How about getting all the lines started with that special line until a line starts with number,number and : as like 14: ? – kamaci Jun 20 '12 at 11:38
You're right Dennis. I fixed my answer. Thanks – criziot Jun 20 '12 at 11:45
@Kamaci I don't think what you want is possible with just regex. You will need something like Bison. – criziot Jun 20 '12 at 11:47

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

Not the answer you're looking for? Browse other questions tagged or ask your own question.