UNIX for Dummies Questions & Answers

Hi all,

I want to extract rows with the pattern ALPHANUMERIC/ALPHANUMNERIC in the 2nd column.
I dont wan rows with more than 1 slash or without any slash in 2nd column.

Code:

a a/b
b a/b/c
c a/b//c
d t/y
e r
f /f

I came up with the regex

Code:

grep '[a-zA-Z0-9]\/[a-zA-Z0-9]$' file
a a/b
b a/b/c
d t/y

but I don't want "a/b/c", since it has 2 slashes.

Also since I want to match only the 2nd column, (column 197 in the real data), I need to use awk to match the regex.


In that case

Code:

awk  'match($2,/[a-zA-Z0-9]\/[a-zA-Z0-9]$/){ print }' file

where I can replace $2 with $197 in the real data,

but this doesn't produce any output at all.

Please assist where I`m going wrong.

Code:

awk '$2~"^[^/]/[^/]$"' file

Since $2 is the 2nd (white-space-separated) field, you can use the ^ and $ anchors.
For your attempt, simply add the left anchor:

Code:

awk  'match($2,/^[a-zA-Z0-9]\/[a-zA-Z0-9]$/){ print }' file

With egrep (or grep -E) you have to insist on a leading space:

Code:

egrep '[[:space:]][^/]/[^/]$' file

Using your regex

Code:

awk '$2 ~ /^[a-zA-Z0-9]\/[a-zA-Z0-9]$/ {$2=$197}1' file

Try also

Code:

awk '2==split($2,T,"/")' file
a a/b
d t/y
f /f

The last line would fulfill your requirement of exactly one slash, but no alphanumeric char in front.