How to concatenate rows of 2 files !

noony123 · 04-03-2011, 10:31 AM

Hi all.

How can i concatenate rows of 2 files.

File1:
10.1.1.
10.1.2.
10.1.3.

File2:
.1
.2
.3

How want file3 to be like
10.1.1.1
10.1.2.2
10.1.3.3

Any help pls ?

colucix · 04-03-2011, 10:35 AM

Check the paste command. You have to specify a null delimiter using the -d option, since the default one is TAB. Also you have to remove the extra dot from the end of the first file or from the beginning of the second one. Suggestion using process substitution:

Code:

paste -d '' file_one <(cut -c2 file_two)

bibiki · 04-04-2011, 11:15 AM

only: 'paste file1 file2' (as long as the files are on your current directory) would do, too.

grail · 04-04-2011, 11:58 AM

@bibiki - and what about the extra period?

colucix · 04-04-2011, 12:25 PM

The paste command uses a tab as default separator. The -d option is mandatory in this case.

Code:

$ paste file1 file2 | od -c
0000000   1   0   .   1   .   1   .  \t   .   1  \n   1   0   .   1   .
0000020   2   .  \t   .   2  \n   1   0   .   1   .   3   .  \t   .   3
0000040  \n
0000041

Telengard · 04-04-2011, 12:35 PM

Quote:

Originally Posted by grail

@bibiki - and what about the extra period?

The following is one possible solution. It is by far not the most efficient, nor the most graceful possible solution.

Code:

foo$ cat file1
10.1.1.
10.1.2.
10.1.3.
foo$ cat file2
.1
.2
.3
foo$ paste -d '' file1 file2
10.1.1..1
10.1.2..2
10.1.3..3
foo$ paste -d '' file1 file2 | awk 'BEGIN {FS="."; OFS="."}; {print $1, $2, $3, $5}'
10.1.1.1
10.1.2.2
10.1.3.3
foo$

Enlightening comments welcomed

Trickie · 04-04-2011, 01:03 PM

I'm no Awk expert but the following works:-

awk -F "." '{OFS="."} {$4=$3}1' file1 > file3

Richard

colucix · 04-04-2011, 03:31 PM

Quote:

Originally Posted by Telengard

The following is one possible solution. It is by far not the most efficient, nor the most graceful possible solution.

foo$ paste -d '' file1 file2 | awk 'BEGIN {FS="."; OFS="."}; {print $1, $2, $3, $5}'

Enlightening comments welcomed

That's nice. But you might exploit all the potential of awk. For example:

Code:

awk '{getline var < "file2"; print $0 substr(var,2)}' file1

No need for the paste command, here.

grail · 04-04-2011, 08:51 PM

@Trickie - you are only referencing one file ... so how is your code going to work?

colucix has the right idea (as usual

). Another alternative:

Code:

awk -F. '{var = $0; getline < "file2"; print var $NF}' file1

Telengard · 04-04-2011, 11:56 PM

Quote:

Originally Posted by colucix

you might exploit all the potential of awk.

My awk-fu is admittedly still quite weak. Must get stronger.

Quote:

awk '{getline var < "file2"; print $0 substr(var,2)}' file1

That's how input redirection works? Sweet!

Another alternative solution, inspired by my efforts to decode colucix's first post in this thread. This one doesn't use process substitution.

Code:

foo$ cut -c2 file2 | paste -d '' file1 -
10.1.1.1
10.1.2.2
10.1.3.3
foo$

Here's one using only Bash internal functions (meaning no programs from outside Bash itself). I couldn't think of any way to do it in a single loop, so I leave that as an exercise for the reader.

Code:

foo$ t="$IFS"; IFS=.; i=0; while read a b c; do l[++i]="$a.$b.$c"; done < file1; IFS="$t"; i=0; while read d; do echo "${l[++i]}$d"; done < file2
10.1.1.1
10.1.2.2
10.1.3.3
foo$

grail · 04-05-2011, 02:12 AM

Quote:

Originally Posted by Telengard

Here's one using only Bash internal functions

Your on

Code:

#!/bin/bash

exec 3<file1 4<file2

while read -ru3 LINE && read -ru4 END
do
    echo $LINE${END#.}
done

exec 3<&- 4<&-

Trickie · 04-05-2011, 03:41 AM

@grail - my code does work; check it out. The rationale is that, from the example, it is obvious that "file 2" echoes the final column of "file 1", therefore, it is unnecessary to call "file 2" as the output from "file 2" is already embedded in "file 1".

Richard

grail · 04-05-2011, 07:44 AM

@Trickie - ok ... I will concede that with your rationale this specific example works, but based on the fact that the OP says he has 2 files to be joined, it seemed obvious that this was a simple
example to help illustrate the point. Assuming your version, you could simply do:

Code:

awk -F. '$0=$0$3' file1

Trickie · 04-05-2011, 10:24 AM

@grail - err, no you can't. There are two errors: it doesn't save into a third file, which "noony123" asked for and secondly, your code produces no "dot" as a field separator between the third and fourth fields, again, as "noony123" requested.

Richard