topic Re: Common lines between two files in Operating System - HP-UX

Common lines between two files

Christian Aguilar Varga — Tue, 15 Apr 2003 19:24:46 GMT

Hi,
I have two files with some common lines between them.
How Can I extract these common lines and put them into another file?

Are There some shell commands or scripts?

Thank You!!!!

Re: Common lines between two files

Zafar A. Mohammed_1 — Tue, 15 Apr 2003 19:26:39 GMT

use grep or egrep command.

Thanks
Zafar

Re: Common lines between two files

Wilfred Chau_1 — Tue, 15 Apr 2003 19:30:21 GMT

# sdiff -o output -l teset1 teset2

when you are prompted with %,
enter l.

Then your matching lines will be in the output file.

Re: Common lines between two files

James R. Ferguson — Tue, 15 Apr 2003 19:30:32 GMT

Hi:

See the man pages for 'comm' and 'diff'.

Regards!

...JRF...

Re: Common lines between two files

Dario_1 — Tue, 15 Apr 2003 19:33:00 GMT

Hi!

See the man pages for diff

Regards,

DR

Re: Common lines between two files

Systeemingenieurs Infoc — Wed, 16 Apr 2003 07:50:56 GMT

You can and should user the underestimated join command, it's not only very performant, but also very powerful.

file1="file1"
file2="file2"

sort1="${file1}.sort"
sort2="${file2}.sort"

sort ${file1} > ${sort1}
sort ${file2} > ${sort2}

join ${sort1} ${sort2}

rm -f ${sort1} ${sort2}

hecou

Re: Common lines between two files

H.Merijn Brand (procura — Wed, 16 Apr 2003 08:18:46 GMT

It all very much depends on how these `common' lines appear in the files.

If both files are sorted, 'comm' is *the* way to go:

# comm -12 file1 file2 >common
# comm -13 file1 file2 >onlyin2
# comm -23 file1 file2 >onlyin1

Then of course you have the question "are all lines unique?", and do you want unique lines squeezed? If lines are distributed unsorted and non-uniquely, and you want to see all lines:

# fgrep -f file1 file2

which has the disadvantage that if file1 contains short lines that might be occuring as part of files2's lines, you get too much. With GNU grep you can do better:

# fgrep -x -f file1 file2

And to squeeze the output, filter the dups:

# fgrep -x -f file1 file2 | uniq

Or if the sequence is not important:

# fgrep -x -f file1 file2 | sort -u

Or alternatively

# perl -e'$f2=pop;while(<>){$x{$_}++};@ARGV=($f2);while(<>){exists$x{$_}and print' file1 file2

Enjoy, have FUN! H.Merijn

Re: Common lines between two files

vasundhara — Wed, 16 Apr 2003 09:46:08 GMT

Hi,

you can use sdiff for this.

-o output
Use the next argument, output, as the name of a third file that is created as a user-controlled merging of file1 and file2. Identical lines of file1 and file2 are copied to output. Sets of differences, as produced by diff(1), are printed;

where a set of differences share a common gutter character. After printing each set of differences, sdiff prompts the user with a % and waits for user-typed commands.

If you give q, it will quit and the common lines will be present in file "output".

Regards
VJ