Shell Programming and Scripting

Hi, I want to read lines from a file, and I'm using two methods
1 use

Code:

while read line
do
done<filename

2 use

Code:

line=`sed -n '3p' filename`

however, in both of them, I notice that the tab between fields are automatically converted to space

because I want to use awk over the lines, I hope the 'tab' can be kept
are there any ways? thanks

When evaluating, double-quote the variable line:

Code:

echo "$line"

You are looping over the file lines to use awk on a line at a time? That'd be pretty inefficient. You may replace everything in the loop with just a single awk.

no , I need to do several processings for each line
not just use awk over each line

Awk can be used to do lots of things to individual lines, to every line, or to any line matching a pattern.

If you'd tell us what "processings" you need to perform, we may be able to help you come up with a MUCH more efficient way to perform them in a single invocation of awk even if you're processing multiple files.

I have many lines like the following in a file(there are also other kinds of lines)

Code:

Host: 72.52.104.74 (tserv1.fmt2.he.net) Ports: 22/open/tcp//tcpwrapped///, 53/open/tcp//domain//PowerDNS 3.3/, 179/open/tcp//tcpwrapped///      Ignored State: closed (997)     Seq Index: 207  IP ID Seq: All zeros

and in these lines, fields which starts with "Host:", "Ports:" are separated by '\tab'. The number of fields in different lines may be different, which means that maybe in this line there are only fields starts with "Host:", "Ignored States:", while in another line the fields are as the above.

Now, I want to process these lines so that different fields are written into different files, like fields starting with "Hosts:" are written into file host.log, bla bla

I have tried this solution provided by elixir_sinari and it looks perfect for this...

Code:

$ while read line; do echo "$line" | awk -F "\t" '{ print NF }'; done<file2
4
$  awk -F "\t" '{ print NF }' file2
4
$ while read line; do echo $line | awk -F "\t" '{ print NF }'; done<file2
1

in the file, different lines have different number of fields
how can you use

Code:

$  awk -F "\t" '{ print NF }' file2

?

can for each field, you haven't process the keyword

Moderator's Comments:

Use code tags as you did use them in your previous posts in this thread here - so you already know how and when, thanks