Shell Script error unexpected else token - bash

Question

I am using below code to read records from one file and comparing its 6 field with other file and accordingly writing the output to the files.
for FILE_NAME in `GET_FILE_LIST $1 $2 $3`
do
y=`echo $FILE_NAME|awk '{print $6}'`"X"
echo "Start --- `date`" >> $LOGF
echo $y >> $LOGF
cat $x |nawk ' BEGIN { FS=",";
while ((getline < "SUBNO.txt") > 0)
myarray[$1] = $2
if(myarray[$6]==0)
' > {${FILE_NAME}_P}
else
' > {${FILE_NAME}_S}
fi
UPD_FILE_STATUS $FILE_NAME 35
done
The issue is I am geeting following error:
./TEST.sh: line 95: syntax error near unexpected token `else'
./TEST.sh: line 95: `else'
I have following case:
In one text file I have data like this:
GGO 099E7C5S 34 533196588 45696 22
PPC 93403DSA 35 784397429 44696 56
In the second file data is like this:
22,0
24,1
26,0
What I want to do is write line from first file, compare the last field value with second file.If it is value is 0 put it in new file_1 and if it is 1 then put it in file_2.
I hope this will clear the case.
Please help.

Answer 1

There are a lot of bad practices in this script which I will not address here. The problem you have is that you seem confused about quoting rules. Consider the code:
cat $x |nawk ' BEGIN { FS=",";
while ((getline < "SUBNO.txt") > 0)
myarray[$1] = $2
if(myarray[$6]==0)
' > {${FILE_NAME}_P}
else
The single quotes delimit the script that is passed to nawk, so the above snippet is essentially equal to:
cat $x | nawk cmds > {${FILE_NAME}_P}
else
The colorization of your editor (and even in the question's display here on SO) should make this clear. You will notice that the else does not have a matching if, so the shell treats that as a syntax error.
Please note that you are redirecting to a file named {xxx_P}. It is unusual to put { and } characters in a filename.

Answer 2

This will do what you updated your question to say you wanted (split file1 based on info from file2)
awk 'NR==FNR{sfx[$1]=$2; next} {print > ("file_" sfx[$NF]+0)}' FS="," file2 FS=" " file1
You didn't say what to do with lines from file1 when the last field on the line has no entry in file2 so my script treats it the same as if there was a 0-ending entry in file2. If that's not what you want, it's an easy fix.

Answer 3

The following will extract from file2 all lines whose sixth field is present with a ",0" suffix in file1. It does this by transforming the relevant lines from file1 into a simple awk script. Repeat with the ",1" suffix to get the other set, or learn enough awk to do it all in a single pass; it's not hard.
sed /,0$/!d;s///;s/.*/$6 == "&"/p' file1 |
awk -f - file2
The sed script is somewhat obscure; it deletes any non-matching lines, then replaces the matching suffix with nothing, then wraps the remainder with a suitable awk condition.
Depending on taste and requirements, you could make the second script a sed script instead (hint: look for the w command) or the first script an awk script as well.