bash字符串比较两个csv文件

Question

I have two strings (parsed from csv) which has ~200 columns each. I need to compare them and identify which column is different. Example:

str1file1="a,b,c,d,e,f,pp,qq"
str2file2="a,b,c,d,x,f,pp,qq"

I need to get column number as 5 and corresponding values as my output. Example: 5 ef As I need to compare millions of such strings, speed is the key. Actual record -

0x0009aeef,xyz,wert,57116,192.168.17.1,45320,192.168.17.2,45320,ctty,lkipop,1408477403,1408477403,,1408477722,1408477403,1408477718,2,0,5,98,0,3055925732,0,0,0,0,15756,15732,24,0,0,0,0,0,0,0.68,23,0,1,23,15776,0.00,15270,459,1,0,0,0,0,0,0,0,0,0,5.755,1408477403,1408477718,2,0,7,98,0,112988428,0,0,0,0,15776,15742,34,0,0,0,0,0,0,8.32,33,0,1,33,15756,0.01,15555,185,0,0,0,0,0,0,0,0,0,0,3.077,-0,-0,-12,-11,-23,-36,-31,-39,22,35,19,28,,,,,1.8,2.4,2.2,2.6,1.8,2.4,2.2,2.5,37,49,45,52,36,48,44,51,15625,107,891,5.60,12528,3204,14430,1312,723,2.65,13291,2451

0x0009aeef is a primary key/column (1st col), however it's not assured that both file has identical number of entries (rows). I'm using sort wrt primary key and get required columns (~135) using cut creating temp files . Followed by 'while read' to read 1st temp file and grep to get matching lines on temp2 file. If grep fails, chances are key or values are different. Hence awk for key and values. Any better approach much appreciated. Here is present code -

sort --field-separator=',' --key=1 $csv1 | cut -d "," -f1,...134 | tr -d '\t' > file1
sort --field-separator=',' --key=1 $csv2 | cut -d "," -f1,...134 | tr -d '\t' > file2
while read line; do
      sl=`grep "$line" file2`
      if [ "$line" != "$sl" ]; then
         rec=`echo $line | awk -F, '{ print $1 }'`
         slId=`grep $rec file2 | awk -F, '{ print $1 }'`
         if [ "$rec" = "$slId" ]; then
               #validation failed, primary key found
         else
               #prim key not found
        fi
     else
        #all is well
     fi
done < file1

Answer 1

If speed is the key, I'd consider parsing the CSV files using mawk or update the post with file examples so we can offer a better solution.

Using Bash:

IFS=, read -a line <<<"$str1"
IFS=, read -a line2 <<<"$str2"
for i in ${!line[@]}; do
    if [[ ${line[i]} != ${line2[i]} ]]; then
        echo -e "${line[i]}\n${line2[i]}"
    fi
done

Output:

e
x