How to grep a whole string, with spaces and dots etc

[Autosys]

Hi All ...

My Script looks like this:

********************************************************
cat sanity.cfg | while read line
do

grep $line bigfile

if [ "$?" -ne "1" ]
then

echo "Cannot find $line"

fi

done;
**********************************************************


And the sanity.cfg file will have many strings in it that look like the one below:

<remote host="1.1.1.1" port="6018"/>

I need to check if these strings exist in a bigger file called bigfile, and if they don't raise a warning. The problem is that grep seems to complain about the special characters, dots and spaces in my string and break it up ... although when I do a "echo $line" I get the correct/whole line back. I tried grep -x & grep -w etc but they don't appear to work, also grep -E. I also tried putting the strings I'm searching for in the cfg file in single quotes .... Any ideas or advice or a possible better way of doing it would be fantastic .. Thanks a lot!

S

 

[PHV]

You may try something like this:
while read line
do
grep -F "$line" bigfile || echo "Cannot find '$line'"
done < sanity.cfg

Hope This Helps, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884 or FAQ181-2886

 

[SamBones]

Use [tt]fgrep[/tt]. It does no pattern matching, so you can look for any regex characters without them being treated as special characters.

#!/bin/ksh

while read LINE
do
    fgrep "${LINE}" bigfile || print "Cannot find ${LINE}"
done < sanity.cfg

[tt]fgrep[/tt] is also usually a little faster since the pattern match is simpler.

 

[PHV]

SamBones, fgrep is the deprecated version of the grep -F suggestion i've made ...

 

[Autosys]

Thanks a lot! Your posts helped me to finish my little script! Maximum Appreciation!

S