我想从 file1 中的第 1 列和第 2 列中找到一些与 file2 中的第 1 列匹配的字符串/单词，并替换为 file2 中的第 2 列字符串/单词

Question

I'm still learning coding using Linux platform. I have search for problems similar to mine but the once I found they were either specific or focusing only on changing the entire column 1.

Here are example of my files:

File 1

abc Gamma 3.44
bcd abc 5.77
abc Alpha 1.99
beta abc 0.88
bcd Alpha 5.66

File 2

Gamma Bacteria
Alpha Bacteria
Beta Bacteria

Output file3

abc Bacteria 3.44
bcd abc 5.77
abc Bacteria 1.99
Bacteria abc 0.88
bcd Bacteria 5.66

I have tried: awk:

$ awk 'FNR==NR{a[$1]=$2;next} {if ($1,$2 in a){$1,$2=a[$1,$2]}; print $0}' file2 file1
$ awk 'NR==FNR {a[FNR]=$0; next} /$1|$2/ {$1 $2=a[FNR]} 1' file2 file1

They gave me:

abc Gamma 3.44
abc 5.77
abc Alpha 1.99
Bacteria abc 0.88
bcd Alpha 5.66

Only changing the $1 and remove the other text strings in column 1 which are not found in file2 $2

And this one:

$ awk -F'\t' -v OFS='\t' 'FNR==1 { next }FNR == NR { file2[$1,$2] = $1 FS $2 } FNR != NR { file1[$1,$2,] = $1 FS $2} END { print "Match:"; for (k in file1) if (k in file1) print file2[k] # Or file1[k]}' file2 file1

Didn't work

Then after i tried sed:

$ sed = file2 | sed -r 'N;s/(.*)\n(.*)/\1s|\&$|\2|/' | sed -f - file1

This gave me an error and complained about sed -e not being called properly.

Then after take only the smallest $3 if $1 and $2 or $2 and $1 are similar

file 4

bcd abc 5.77
Bacteria abc 0.88
bcd Bacteria 5.66

I have tried this code:

$ awk 'NR == $1&$2 || $3 < min {line = $0; min = $3}END{print line}' file3
$ awk '/^$1/{if(h){print h RS m}min=""; h=$0; next}min=="" || $3 < min{min=$3; m=$0}END{print h RS m}' file3
$ awk -F'\t' '$3 != "NF==min"' OFS='\t' file3
$ awk -v a=NODE '{c=a*$3+(1-a)} !($1 in min) || c<min[$1]{min[$1]=c; minLine[$1]=$0} END{for(k in minLine) print minLine[k]}' file3 | column -t

All didn't work and i tried to research what what does each line means and changed it to fit my problem. But they all failed

Answer 1

This might work for you (GNU sed):

sed -E 's#(.*) (.*)#/^\1 /Is/\\S+/\2/;/^\\S+ \1 /Is/\\S+/\2/2#' file2 |
sed -Ef - file1

Generate a sed script from file2 which is run against file1 to produce the required format.