Search Results

I'm not at home on solaris; does this help:

sed '1 {x; d; n; }; G; s/\
/ /;' file2 file1

That's not how it works. It prints the line incl. a <new line> char, and then reads and prints the text, regardless of the $ regex.
How about
sed '1 {x; d; n}; G; s/\n/ / ' file2 file1
This is a...

awk '$4 > a[$1] {a[$1]=$4; b[$1]=$0}; END {for (i in b) print b[i]}' infile

Why are you programming shell scripts inside expect?!

Use ssh as intended and you will have no need to use expect to automate the shell.

Have you tried enclosing the entire send data in single quotes?

Where is /home/file.txt (do you want the local or the remote) and what is in it?


Robin

In it it simplest form using anchors ( ^ and $ ):
sed 's/^bird/bunny/; s/dog$/kitty/' file
or depending of input variation, a little better, also using spaces to mark the other side of the word...

Try:
egrep -o '[[:digit:]]\.[[:digit:]]|[[:digit:]][[:digit:]]\.[[:digit:]]'
To grep for literal dots.

This can be simplified to the equivalent:
grep -Eo '[[:digit:]]{1,2}\.[[:digit:]]'

But...

sort -k2 -k4nr file

I get:

$ sort -k2 -k4nr myFile
xxxx apples xxxx 100
xxxx apples xxxx 90
xxxx oranges xxxx 70
xxxx oranges xxxx 80
xxxx pears xxxx 100
xxxx pears xxxx 90

is that what you're NOT after?

Ok
but what follows after:
#! /usr/bin/sh
script /home/test.log
cat somefile
exit
Is not a session but a script...
script /home/test.log script get executed and is waiting for session...

Hi,

Essentially, this is correct behaviour. What's happening is that the 'script' command is spawning its own sub-shell underneath your current shell, and that's why you get a shell prompt and...

Sorry, yes - that was a bit short sighted; I could not test it as I was away from my computer.
For exactly the given input, try

sed -n '/Name/{n;x;}; /Phone/{n;H;x; s/\n/ /; p}' file
John Davis...

The literal string aaa-bbb-ccc-a file that is in file2 does not appear anywhere in file1. Why do you think there should be any output???

If file2 contained:
aaa-bbb-ccc-a
instead of:...

a bit on a verbose side, but..
awk -f jim.awk myFile where jim.awk is:

{f4A[$4]; pA[$0]; a[$1 OFS $2 OFS $3]}
END {
for (i in a) {
printf("%s", i)
for (j in f4A)
...

Hello jimmyf,

Yes, even I am getting the same result as you got(tested it in BASH) bird1 first, so could you please try following and let me know if this helps you.

sort -k1,1 Input_file
...

From man/info sort:
*** WARNING *** The locale specified by the environment affects sort
order. Set LC_ALL=C to get the traditional sort order that uses native
byte values.

Maybe if you showed us what you tried, we could help you in your way, as I believe there are many solutions... and what was grep for? (here...)

Saying "does not work" without showing us the diagnostics you get, without telling us what operating system you're using, and without telling us what shell you're using gives us VERY little...

can you explain this..

How do you add a 2nd option?

---------- Post updated at 10:50 AM ---------- Previous update was at 10:46 AM ----------

$ ./t.sh
Is this pre or post : pre
This is pre...

Try awk '$1 != L {printf "%s%s", DL, $0; DL=RS; L = $1; next} {printf "%s%s%s%s%s", $2, OFS, $3, OFS, $4} END {print ""}' FS="," OFS="," file

awk 'F[$1]++ {$1=OFS}1' test.file

You need to specify the -f option to make file2 the pattern file:fgrep -vf file2 file1
bird 5
fish 7

Closing the output files would help:awk '{print > $NF; close($NF)}' file

Thisawk '{print > $NF}' filewould work had you given your sample a bit more carefully; as is, the last file name is just four.