Hi,
I am trying to parse a file for email addresses. So here's the line I'm using with grep:

grep -E "^([a-zA-Z0-9_-])+([\.a-zA-Z0-9_-])*@([a-zA-Z0-9_-])+(\.[a-zA-Z0-9_-])+" Export

When I run it, I get a response however, the first letter of every email address found is truncated, and it is also outputting info like this:

... Deferred: Connection timed out with blah.blah.blah.

So lets say the email address in the file is:

user1@linuxquestions.org

But when grepped the output is:

ser1@linuxquestions.org

and the space is infront of it in the results also, it's like the results are displayed with a space replacing the first character of all matching results.

Regex are still like magic to me, and I have the hardest time figuring them out. I thought I was getting better, but I guess not. Please if anybody can help that would be great! Thanks in Advance!

Mike.

That would strike me as odd, your expression works
here... what's your grep's version number?

Cheers,
Tink

I thought it was wierd too, I have used this expression for many email comparison things. I'm using grep 2.4.2, RH 7.2

Thanks for the reply.
Mike.

grep (GNU grep) 2.5

Copyright 1988, 1992-1999, 2000, 2001 Free Software Foundation, Inc.

Slack 8.1 ... odd ...

Sorry, I don't know what to say :/

Cheers,
Tink

Try using a perl regexp.

you mean write a script in perl to do the job or use the format of a perl regexp?

Thanks for the help!

I tend to use perl full stop for most things - I doubt if using a perl like regexp in grep would make much difference. Like Tinkster - your grep command works for me. I can only assume there is somthing odd with either the file you are reading from (malformed characters or bad end of lines) or there is somthing wrong with another part of the script you are running this command from.

How are you calling the grep command? From the command line or from within a script? It may help if you post the script (if there is one) and the content of the file you are trying to grep.

I was running grep from the command line. I can't really post the contents of the file cause it is a huge file 34meg, it is just an inbox file from netscape, however there are characters in the file, after some of the email address that look like this: ^M I believe that is the windows line break character, but I can't see how that would influence it. The lines aren't that complicated, but I have tried this on another file and had it work also, I wonder if I uploaded the file in Binary...hmmm...

Thanks for the help.

Figured it out,
I search and replaced those ^M with \n and it fixed it. Now does anybody know how to have grep return only the exact matches and not the full lines where they match? For example the line is:

blah@linuxquestions.org ... Deffered ...

Grep returns:

blah@linuxquestions.org ... Deffered ...

But I only want:

blah@linuxquestions.org

TIA!
Mike.

I think you will need to use awk.

I don't know of a way to do it with grep.

AFAIK, Nutscrape uses std mbox format, which means you should be able to do "cat <nutscrape_mbox> | formail -czx From: -s" to get the "from" address for instance.
If you have got to get the 3rd field from a header and know fields are space separated, you could do "cat <nutscrape_mbox> | formail -czx From: -s" | while read i; do i=( ${i} ); printf "%s${i[2]}\n"; done". Arrays count from zero up.

Cheers,
Tink