Shell Programming and Scripting

HI,

I need one help to generate the file based on few condition,
I need to print only those line which has N and delete the line which as all Y's. ANd I need to compare only the fields which are marked as Y and N .
And one more thing is in the below file 1 and file 2 and file2 abc, dbc can come in any of the fields,

for exp , file1

Code:

abc|Y|Y|Y|
bdc|N|Y|Y|
dfg|Y|N|Y|
hjk|Y|Y|Y

output of file1

Code:

bdc|N|Y|Y|
dfg|Y|N|Y|

file2

Code:

Y|Y|abc|Y|
N|Y|bdc|Y|
Y|N|dfg|Y|
Y|Y|hjk|Y|

output of file2

Code:

N|Y|bdc|Y|
Y|N|dfg|Y|

But I need to compare only Y and N fields.

Can you please help me out.

Thanks in Advance
Rashmi

try this:

Code:

 
egrep '\|N$|^N\||\|N\|' filename > newfilename

thanks a lot, it works ...I had one more doubt, wht if the field delimiter is not fixed i.e what if I had to process multiple files without knowing the filed delimiter .
Can I pass the field delimiter during the runtime.
in the below example, my file1 has pipe delimiter and file2 has tab delimter.

file1
abc|Y|N|
cdb|Y|Y

and file2
abc Y N
cdb Y Y

can try this:

assign a shell variable as below

Code:

 
delimit="\|"  ##considering it is a pipe here, as | is a reg exp metacharacter we have to escape that.

then use egrep as below

Code:

 
egrep "${delimit}"'N$|^N'"${delimit}"'|'"${delimit}"'N'"${delimit}" filename > newfilename

Hi,

Try this

Code:

sed "/Y|Y|Y\|Y|Y|.*|Y/d" file*

And if you want to modify the file content just add -i option to sed.

Code:

grep N file*

thanks this works egrep '\|N$|^N\||\|N\|' file. Is it possible to make it read from the 2nd line, as my 1st line of the file is a header.

Thanks