Shell Programming and Scripting

I am having a requirement like
File A contains All the records
eg.,

Code:

ONE
TWO
THREE
FOUR
FIVE
SIX
SEVEN
EIGHT
NINE
TEN

File B contains invalid records of File A
eg.,

Code:

TWO
FOUR
FIVE
SEVEN
NINE

Is there any sed or awk command to be used in shell script to compare files A and B and filtered out only good records in A.

Basically I need to delete all the lines in A which are present in B.
My final A file should be like

Code:

ONE
THREE
SIX
EIGHT
TEN

Can Anyone please help.

Ofcourse using grep you can do this in one line...
Try man grep
If it is not clear look for
-v and -f options
Still not clear, then ask.

Using sed. Create sed rulelines from B = every value is "delete line from a"

Code:

sed "s/\(.*\)/\/^&$\/d/" b.txt > x.tmp
sed -f x.tmp a.txt

Great...Thanks a lot ..its working..if I face any problem..i wl come back again

---------- Post updated at 05:23 AM ---------- Previous update was at 05:01 AM ----------

hi..the command is working with my test data..
but my actual input & invalid file contains white spaces also. So I am getting the error as

sed: 0602-404 Function /^B <the whole line with spaces> $/d cannot be parsed

Please help

Did you check the grep option, I suggested??

Code:

grep -v -f B A

If data include special chars (reserved char) like *$()!&/ ... then you have to escape those chars like here is ex. $ *
\1 = 1st ( ) group
& = whole input
in this case they are same

Code:

sed -e "s/\$/\\\$/g" \
  -e "s/\*/\\\*/g" b.txt | \
sed "s/^\(.*\)$/\/\1\/d/"  > x.tmp
# or "s/.*/\/&\/d/"
sed -f x.tmp a.txt

grep -v -f b.txt a.txt
is nice, but same problem. Ex. if linedata include '.*', result is ...

Hi kshji..
Can you please be more clear on the above code..
Thanks in advance