Very Computer

when grepping will this expression math anything?

grep '*U" thefile

It dosen't seem to. It says Match zero or more of nothing and then a U.
Am I missing something?
Thanks
Jeem

You have it back to front:

    grep 'U*' thefile

That will match any line that contains zero or more U's.

Don't be surprised when you find that it matches every single
line in the file.  Every line will contain at least zero U's.

--
Andreas K?h?ri

You're mixing quotation marks - pick either single-tick, ', or
double-tick, ", for both ends of the RE.

Quote:> It dosen't seem to. It says Match zero or more of nothing and then a U.
> Am I missing something?
> Thanks
> Jeem

Pick either type of quotation mark and it'll match a sequence of zero or
more spaces followed by a U. You'd get the same effect with just

        grep 'U'.

Regards,

        Ed.

Originally posted by Ed Morton

Not sure which greps this might apply to but mine will interpret a *
as exactly that at the beginning of the regular expression and
doesn't treat it as a metacharacter.  I can't see any reference to
this in the man page.

e.g.

echo "*U" | grep "*U" # match found

echo "U" | grep "*U" # no match found

--
Posted via http://dbforums.com

[cut]

Quote:> Not sure which greps this might apply to but mine will interpret a *
> as exactly that at the beginning of the regular expression and
> doesn't treat it as a metacharacter.  I can't see any reference to
> this in the man page.

> e.g.
> echo "*U" | grep "*U" # match found
> echo "U" | grep "*U" # no match found

You're correct.  The '*' is taken literally (i.e. it has no
special meaning) when in the beginning of a regular expression,
or just after the '^' anchor as in "^*" or "*U" (and in a couple
of other cases).  Hence, "*U" will match the character '*'
followed by the character 'U'.

--
Andreas K?h?ri

<snip>

Quote:> Not sure which greps this might apply to but mine will interpret a *
> as exactly that at the beginning of the regular expression and
> doesn't treat it as a metacharacter.  I can't see any reference to
> this in the man page.

<snip>

I thought he had a single blank immediately before the "*" in his RE. It
looked that way in the font I use and his description sounded like he
was trying to describe the effect that would have. Looking at it again,
I see there was no blank so my description was wrong. Thanks for
catching it.

        Ed.

Ah, I think I know what it is you're trying to do now...  Sorry
for the confusion.  You're trying to match a word that begins
with 'U', don't you?

In that case, try one of these:

    grep '\<U' thefile
    grep '\bU' thefile

This requires GNU grep, which extends the POSIX definition of
the utility.

Does anyone know what the difference between \b and \< is?  It's
not entierly clear from the GNU grep manual I think.

--
Andreas K?h?ri

You are confusing file name patterns expanded by the shell with regular

expressions matched by grep and other text utilities.  They are not the
same.

--
Kevin Rodgers