testing for regular expressions

[jescat]

Is there a way in unix scripting to test variable using regular expressions similarly to awk or sed

For example:

awk:

$1 ~ /^[A-z].*$/ print "found regular expression"

Can you incorporate regular expression metacharaters in unix scripts

 

[MikeLacey]

Hi there,

Yes, you can put regular expression metacharacters in scripts and spotting *some* re's will be fairly easy - but spotting *ALL* re's will be quite a task.

Tell us a bit more about what you're trying to achieve. Mike
michael.j.lacey@ntlworld.com
Email welcome if you're in a hurry or something -- but post in tek-tips as well please, and I will post my reply here as well.

 

[jescat]

I am not trying to accomplish anything in particular. It would just be nice to have the ability to test on re's like awk, sed or Perl. I figured there was but just didn't know the syntax. What were you thougthts?

Thanks

 

[MikeLacey]

Oh I see, well - you can use the standard command 'grep' to use regular expressions in a UNIX shell script.

You can use grep to process files or its STDIN - so you can echo or print variables to grep's STDIN to check those. Mike
michael.j.lacey@ntlworld.com
Email welcome if you're in a hurry or something -- but post in tek-tips as well please, and I will post my reply here as well.

 

[aigles]

You can also use the expr command :


[ $(expr "$VAR" : "^[A-z].*$&quot -ne 0 ] && echo "found regular expression"



[COLOR=/blue]expr[/color] expr1 : expr2
The matching operator : compares the first argument
with the second argument which must be a regular
expression. expr supports the Basic Regular
Expression syntax (see regexp(5)), except that all
patterns are ``anchored'' (i.e., begin with ^) and,
therefore, ^ is not a special character, in that
context. Normally, the matching operator returns
the number of characters matched (0 on failure).
Alternatively, the \(...\) pattern symbols can be
used to return a portion of the first argument.


See the the man page ... Jean Pierre.