On Thu, 2006-02-02 at 09:39 -0600, Steven J Lamb wrote: > what I am looking for is information on how to use grep in the following > way. I want to do essentially this > > grep (string1|string2|string3) filename > > I would use this to search filename for either string1, string2 or string 3 > instances > > I realize I could do the following but would like to use it as one grep > statement as I believe it should be more efficient. > > >tmp; > grep string1 filename >>tmp; > grep string2 filename >>tmp; > grep string3 filename >>tmp > > as this would be defined in a normal regular expression. grep claims to be > able to use this functionality as best I can tell ... it is documented in > the man page and it says > > Two regular expressions may be joined by the infix operator |; the > resulting regular expression matches > any string matching either subexpression. > > I have not been able to find an example of this anywhere and have tried > several forms of the syntax but no luck. my command interpreter keeps trying > to use the pipe instead of letting grep have it. > > any one have any thoughts > > thanks > With basic regular expression you need to quote the "|" so it would be something like this: grep 'string1\|string2\|string3' file The alternative would be to get in to extended regular expression with egrep or grep -E which would look like: egrep 'string1|string2|string3' file HTH, --Rob
