Help with scripting

Polarbearcub · 04-21-2014, 09:09 PM

Hi All, I need to find differences in two files.

File1
name joe
num 100
name john
num 5460
name smith
num 561
name George
num 7558

File2
name smith
num 561
name George
num 7558
name joe
num 99
name alice
num 0

I have 4000 lines in File1 and 3996 in File2
I need to find the 4 missing lines and make sure the others match up. In my example you see that joe in file1 has 100 but in file2 it's 99. I need to know the differences and the 4 missing lines only.

Thank you for any help!

Berhanie · 04-21-2014, 09:30 PM

use can use the 'diff' command to find the difference between two files. sometimes it's helpful to sort each file with the 'sort' command before comparing them.

danielbmartin · 04-21-2014, 09:46 PM

May we trust that every name line is followed by a num line?

In other words, may we change this ...

Code:

name joe
num 100
name john
num 5460
name smith
num 561
name George
num 7558

... into this?

Code:

name joe num 100
name john num 5460
name smith num 561
name George num 7558

Daniel B. Martin

Polarbearcub · 04-21-2014, 09:52 PM

Yes, we can trust this. Each name has a number on the next line.

danielbmartin · 04-21-2014, 10:29 PM

Quote:

Originally Posted by Polarbearcub

Yes, we can trust this. Each name has a number on the next line.

With InFile1 ...

Code:

name joe
num 100
name john
num 5460
name smith
num 561
name George
num 7558

... and InFile2 ...

Code:

name smith
num 561
name George
num 7558
name joe
num 99
name alice
num 0

... this code ...

Code:

echo; echo "Create Name/Num pairs and sort them."
paste -sd " \n"  $InFile1 |sort >$Work1
paste -sd " \n"  $InFile2 |sort >$Work2

echo "Name/Num pairs found in both input files ..." >$OutFile1
comm -12 $Work1 $Work2 >>$OutFile1
echo; cat $OutFile1

echo "Name/Num pairs found only in InFile1 ..." >$OutFile2
comm -23 $Work1 $Work2 >>$OutFile2
echo; cat $OutFile2

echo "Name/Num pairs found only in InFile2 ..." >$OutFile3
comm -13 $Work1 $Work2 >>$OutFile3
echo; cat $OutFile3

... produced this result ...

Code:

Name/Num pairs found in both input files ...
name George num 7558
name smith num 561

Name/Num pairs found only in InFile1 ...
name joe num 100
name john num 5460

Name/Num pairs found only in InFile2 ...
name alice num 0
name joe num 99

Daniel B. Martin

sharky · 04-22-2014, 01:56 PM

danielbmartin, How have I been using Linux for over two decades and never heard of 'paste' and 'comm'?

Thanks for posting this solution.

danielbmartin · 04-22-2014, 02:54 PM

Quote:

Originally Posted by sharky

How have I been using Linux for over two decades and never heard of 'paste' and 'comm'?

I'll offer a speculative answer.

The Linux world seems to be roughly divided into two segments:
- system administrators
- applications programmers

Are you an administrator? Admins learn and use a certain subset of Linux commands. Apps guys use a different subset. There is some overlap, of course.

My Linux programming is purely recreational and entirely apps. I'm retired and learning Linux (self-taught) for fun and to stave off old-age brain rot.

If you learned something new today, that's good. I try to learn something new every day and don't always succeed.

Daniel B. Martin

MadeInGermany · 04-29-2014, 07:11 PM

Code:

fgrep -vxf smallfile bigfile

works with unsorted files.