Shell Programming and Scripting

opq,098
rst,765
xyz,321

xyz,122,913,opq,876
rst,956,921,xyz,012
456,890,903,rst,467

xyz,122,913,098,876
rst,956,921,321,012
456,890,903,765,467

awk -F, 'FNR==NR{p[$1]=$2; next}
$4 in p { $4=p[$4] }
1' OFS=, pattern_file data_file > output_file

#!/bin/bash

lines=0
while read f1
do
    lines=$(( $lines + 1 ))
    kee2=$(echo $f1 | cut -d, -f4)
    val2=$(grep -e "^$kee2" trydata2)
    if [[ $val2 ]]; then
        repl=${val2##*,}
        sed -i "$lines s/\b$kee2/$repl/g" trydata1
    fi  
done < trydata1
cat trydata1

# output
# ------
# xyz,122,913,098,876
# rst,956,921,321,012
# 456,890,903,765,467

000,5xyz1,000,xyz,000

#!/bin/bash

# Load both data files into memory
< trydata1 mapfile data1
< trydata2 mapfile data2
for f1 in ${data1[*]};
do
    lines=$(( $lines + 1 ))
    kee2=$(echo $f1 | cut -d, -f4)
    for val2 in ${data2[*]};
    do
        if [[ $val2 =~ ^$kee2 ]]; then
            repl=${val2##*,}
            sed -i "$lines s/\b$kee2/$repl/g" trydata1
        fi
    done
done
cat trydata1

# Using this data set with no column issues...
# ----------------------
# xyz,122,913,opq,876
# rst,956,921,xyz,012
# 456,890,903,rst,467
# 000,5xyz1,000,xyz,000
# 4567,5rst1,opq,xyz,000
# rst,5opq1,rst,02opq,000
# 000,5xyz1,xyz,opq,000

# real    0m0.028s
# user    0m0.008s
# sys    0m0.009s

# On a machine running @ 2GHz
# If my math is right...
# 0.028 seconds to process 7 lines
# time to process 30,000 lines ~ 2 minutes

while IFS="," read -a x
do
 lookup=`awk -F"," '$1==s {print $2}' s="${x[3]}" pattern_file`
 [ -n "$lookup" ] && x[3]=$lookup
 echo "${x[@]}"
done < data_file | tr ' ' ',' > output_file