comm command help with unicode chars in file

Shell Programming and Scripting
1

Hi,

I have a Master file (file.txt) with good and bad records( records with unicode characters). I ahve a file with only bad records (bad.txt)

I want the records in file.txt which are not present in bad.txt ie only the good records.

I tried

 comm -23 file.txt bad.txt

It is giving all the records in file.txt.
Any help??

2

you need paste the sample of master file and bad records file

3

Take this as example.. a1 is master file and a2 is bad record file.My output should be
9
8
6
4
2

$ cat a1
9
8
7
6
5
4
2
3
$ cat a2
1
3
5
7
$ comm -23 a1 a2
9
8
7
6
5
4
2
3
4

Hi.

Best wishes ... cheers, drl

5

Is there any way i can do it without sorting the file.The file is big.

6

find the different.

$ grep -v -f a2 a1
9
8
6
4
2

find the same records.

$ grep -f a2 a1
7
5
3
7

Getting the following error with grep -v -f

$ grep -v -f a2 a1
grep: illegal option -- f
Usage: grep -hblcnsviw pattern file . . .
8 
grep -v a2 a1

this shud get u going..

u can use ur master and bad sector file and get the good records

lets say
master file = m, bad record=b
cat m > result
cat b >> result
uniq -u result

this will only give u good results.

-u Displays only lines not duplicated (uniq lines).
9

Assuming each record is unique and that the output order does not matter:

cat file.txt bad.txt | sort |uniq -u

Or sort each file before the "comm".

cat file.txt | sort >file.sor
cat bad.txt  | sort >bad.sor
comm -23 file.sor bad.sor
10

There, now you know the real words.

11

@corona :Thats what dazdseg told rite.. Did you add anything to it...

12

thanks for letting me know.

13

Please don't use 'l33t sp33k' - this isn't twitter. Plain English. If English is a problem for you we have translation services. That is what corona is telling you.

14

Hi, dazdseg.

However:

cheers, drl

15

I will remember that as well since i am new to the forum.