Challenging Script

[dave_nithis]

Hi all,

I have two files namely
a.txt
b.txt
a.txt contains 100 records and b.txt contains 9 records.These 9 records in b.txt is also present in a.txt.So,I need a script that compares these two files and outputs the unique 91 records alone in another file.

Thanks to all in advance!!!!

Regards
dave.

[masijade]

Cool. What do you have so far, and what problem are you having with it.

[dave_nithis]

I am trying to do this.But i dont know how to do the comparison part.

[dave_nithis]

Also am trying this in microsoft Access.

[masijade]

Also am trying this in microsoft Access.

What do you mean by this? This forum is for Linux/Unix Shell Scripting, not Windows Batching (unless things have changed).

As far as doing the comparisons, an easy, but not necessarily performant way is a simple grep command.

[eggi]

a's got 100
b's got 9
you need the 91 uniques - I'll put 'em in c.txt

while read line
do
   grep $line b.txt >/dev/null 2>&1               
   if [ $? -ne 0 ]
   then
       echo $line >>c.txt
   fi
done <a.txt

Sorry - just needed to drop a post somewhere - it's been a while :)

, Mike

[masijade]

a's got 100
b's got 9
you need the 91 uniques - I'll put 'em in c.txt

while read line
do
   grep $line b.txt >/dev/null 2>&1               
   if [ $? -ne 0 ]
   then
       echo $line >>c.txt
   fi
done <a.txt

Sorry - just needed to drop a post somewhere - it's been a while :)

, Mike

Its actually easier, if with a few more file operations.

cp b.txt /tmp/b1.txt
while read line
do
   grep -v $line /tmp/b1.txt >/tmp/b2.txt 2>/dev/null
   mv -f b2.txt b1.txt
done <a.txt
mv b1.txt c.txt

[dave_nithis]

Masijade,

I finished the file comparison in MS Access.But I am trying to do the same file comparison using Shell script...Thats why posted you a thread..

I didn't understand this line in your script :
grep -v $line /tmp/b1.txt >/tmp/b2.txt 2>/dev/null

Can you please explain?

Thanks in advance
Regards
dave.....

[dave_nithis]

a's got 100
b's got 9
you need the 91 uniques - I'll put 'em in c.txt

while read line
do
   grep $line b.txt >/dev/null 2>&1               
   if [ $? -ne 0 ]
   then
       echo $line >>c.txt
   fi
done <a.txt

Sorry - just needed to drop a post somewhere - it's been a while :)

, Mike

Mike,

why are you using ">/dev/null 2>&1" in your script?
What condition you are checking in your "if" loop?

Regards
Dave

[masijade]

Masijade,

I finished the file comparison in MS Access.But I am trying to do the same file comparison using Shell script...Thats why posted you a thread..

I didn't understand this line in your script :
grep -v $line /tmp/b1.txt >/tmp/b2.txt 2>/dev/null

Can you please explain?

Thanks in advance
Regards
dave.....

grep -v means to match everything except the next text (you should probably place $line in quotes) from the the first file listed (b1.txt) and output it to the second file listed ( > b2.txt) and ignore any error messages (2>/dev/null).

The next line then moves b2.txt to b1.txt so you can repeat the procedure with the next line. Once you have gone through all lines, b2.txt contains all lines from b that were not in a.

[eggi]

Hey Dave,

I'm using ">/dev/null 2>&1" after grep in my script so no output will go to the screen from the grep command.

In the if loop, I'm checking the value of $? "errno - or the return code" - grep returns 0 if it's a success. The check needs to come right after the grep command is executed, because the $? variable gets populated with the return code (or error number) of whatever command was run last.

Hope that helps explain :)

, Mike

Mike,

why are you using ">/dev/null 2>&1" in your script?
What condition you are checking in your "if" loop?

Regards
Dave

[laksh.ss]

Just used one command in script

diff a.txt b.txt | uniq > c.txt

We get the unique different records in c.txt