UNIX for Dummies Questions & Answers

Hey, trying to learn more tricks with regular expressions, and got this one.

Best I could come up with for this problem is a "two-stepper" :

Code:

egrep -i "[a-zA-Z0-9]@[a-zA-Z0-9]" test.txt | egrep -v "@[a-zA-Z0-9]*.com"

Which does bring back the proper results. I'm interested, however, if this can be done with a single reg exp ?

Test file is:

Code:

quick brown fox@gmail.com jumped over
quick brown fox@il.com jumped over
lazy dog@over the hill
lazy dog@mail the hill
jack and @jill
ignore this line

Expected return is:

Code:

lazy dog@over the hill
lazy dog@mail the hill

Basically:
@xxx.com is an email, ignore it.
<space>@xxx isn't a pattern I want, ignore it.
no @ in line - don't want it, ignore it.

<string>@<string> is what I'm looking for - but not ending in .com.
(I just look for single character before and after, that way I know it has to be upper/lower case, or number only. No need to look for _, or other special characters.

In short: I'm searching for Oracle DB Links, but don't want email addresses, or SQL script calls.

Code:

egrep '[a-zA-Z0-9]+@[a-zA-Z0-9]+ '

FWIW -

egrep -v does an exclusionary search and using alternation try:

Code:

egrep -v  '( @[a-zA-Z0-9]|\.com )'  somefile

Just an opinion - sometimes really long one-liners are harder to debug. We see 250 character "one-liners" here frequently. They are hard to work with.

Understood .. I'm just curious to use this as a learning exercise.

hmm, that's cute, reverse logic. Clever.
Ok, it still pulls back a line with no @'s in it.
Trying this, however, still doesn't work ... hmmm

Code:

egrep -v  '([^@]| @[a-zA-Z0-9]|\.com )' test.txt