0

I need to some strings in a file in the following fashion

Old file:

Real/Test1
Real/Test1
Real/Test2
Real/Test3
Real/Test3
Real/Test4

New File:

Real/Test1  a1 b1 c1 d1
Real/Test1  a1 b1 c1 d1
Real/Test2  a2 b2 c2 d2
Real/Test3  a3 b3 c3 d3
Real/Test3  a3 b3 c3 d3
Real/Test4  a4 b4 c4 d4

I have an intermediate file which has the old string in the column 1 and then the new string, something like this.

Test1 a1 b1 c1 d1
Test2 a2 b2 c2 d2
Test3 a3 b3 c3 d3
Test4 a4 b4 c4 d4

Could anyone please help with this?

With my very primitive knowledge, I tried following:

(while read n1 n2 do set n1 n2 sed -i "s/$n1/$n1 $n2/g" old > final done)

where "old" and "intermediate" inputs are the content mentioned above.

Thanks a lot !

Improve this question
3
  • 4
    SE is not a free coding service. I'm certain you tried something already, add that to your question. Commented May 9, 2017 at 21:30
  • Agree, let's see what you've tried already, see if we can collectively get you on the right track if it's not working quite right. Commented May 9, 2017 at 21:50
  • Thanks Grochmal and Steve for asking for the trial I am making. My scripting ability is very primitive and therefore I was hesitant putting my attempts here..So what I did is following: (while read n1 n2 do set n1 n2 sed -i "s/$n1/$n1 $n2/g" old > final done)<intermediate where old is the old file as I wrote in my question and intermediate is the intermediate file. Commented May 9, 2017 at 21:57

3 Answers 3

Reset to default
2

Since your files appear to be sorted in order of the join field, you could use the join command fairly easily e.g.

join old <(sed 's;^;Real/;' intermediate)

or (if your shell doesn't support process substitution)

sed 's;^;Real/;' intermediate | join old -

ex.

$ sed 's;^;Real/;' intermediate | join old -
Real/Test1 a1 b1 c1 d1
Real/Test1 a1 b1 c1 d1
Real/Test2 a2 b2 c2 d2
Real/Test3 a3 b3 c3 d3
Real/Test3 a3 b3 c3 d3
Real/Test4 a4 b4 c4 d4
Improve this answer
1
  • Clever and naughty solution! Commented May 9, 2017 at 23:16
1

Try something like this with your gnu awk :

awk -F"[/ ]" 'NR==FNR {a[$1]=$2OFS$3OFS$4;next}$2 in a {print $0,a[$2]}'  intermediatefile oldfile >newfile
Improve this answer
6
  • Thanks George, but this doesn't take care of repeated lines in the old file. Commented May 9, 2017 at 22:38
  • @ashu Yeah, right. Try the revised code. Mind that file read order has been reversed now. Commented May 9, 2017 at 22:48
  • Sorry, this is printing the old file. Commented May 9, 2017 at 22:52
  • @ashu I had a typo. Try again please. Sorry i am on the road and i can not test my proposal... Commented May 9, 2017 at 22:57
  • Thanks so much George, still one problem, spaces between columns 2-4 are lost, they come together. Commented May 9, 2017 at 23:00
0 
perl -lne '
   @ARGV and $h{$1}=s/(\S+)//r,next;
   s|/(\S+)\K|$h{$1}|;print;
' intermediate.file old.file

Results

Real/Test1 a1 b1 c1 d1
Real/Test1 a1 b1 c1 d1
Real/Test2 a2 b2 c2 d2
Real/Test3 a3 b3 c3 d3
Real/Test3 a3 b3 c3 d3
Real/Test4 a4 b4 c4 d4

Explanation

  • Using the intermediate file (@ARGV is > 0) we populate the hash using the first field as the key and the remaining fields as the corresponding value.
  • When we process the old file (@ARGV = 0), we look at the string after the slash and use that to pullup the hash value and put it back in the current line.
Improve this answer

You must log in to answer this question.

Start asking to get answers

Find the answer to your question by asking.

Ask question

Explore related questions

See similar questions with these tags.