extract columns from multiple .csv files and merge them into one - bash

Question

I have three files from which I want to extract some columns and paste them in a new file. The files don't necessarily have the same number of lines. They are sorted on the values in their first column.
File 1 has the following structure:
col1;col2;col3;col4
SAMPLE-1;1;1;1
SAMPLE-2;1;1;1
SAMPLE-3;1;1;1
SAMPLE-4;1;1;1
This file is seperated by ";" instead of ","
File 2 has the following structure:
col5,col6,col7,col8
SAMPLE-1_OTHER_INFO,2,2,2
SAMPLE-2_OTHER_INFO,2,2,2
SAMPLE-3_OTHER_INFO,2,2,2
File 3 has the following structure:
col9,col10,col11,col12
SAMPLE-1_OTHER_INFO,3,3,3
SAMPLE-2_OTHER_INFO,3,3,3
SAMPLE-3_OTHER_INFO,3,3,3
The output file (summary.csv) should look like this:
col1,col2,col4,col6,col7,col10,col12
SAMPLE-1,1,1,2,2,3,3
SAMPLE-2,1,1,2,2,3,3
SAMPLE-3,1,1,2,2,3,3
SAMPLE-4,1,1,,,,
Basically the first columns of all three files contain the sample identifier. 'col1' of file1 should be the first column of the output file. The identifiers in col1 should then be matched with those in col5 and col9 of file2 and file3. The '_OTHER_INFO' part should not be taken into account when doing the comparison.
If there is a match, the info the col6, col7, col10 and col12 values of file 2 and 3 should be added.
If there is no match, the line should still be in the output file, but the last four columns should be empty (like in this case 'SAMPLE-4')
I was planning to perform this action with awk or the 'cut/paste' command. However I don't know how I should look for a match between the values in col1, col5 and col9.

Answer 1

try following and let me know if this helps you.
awk 'BEGIN{
FS=";"
}
FNR==1{
f++
}
f==1 && FNR>1{
a[$1]=$2","$4;
next
}
f>1 && FNR==1 {
FS=","
}
f==2 && FNR>1{
sub(/_.*/,"",$1);
b[$1]=$2","$3;
next
}
f==3 && FNR>1{
sub(/_.*/,"",$1);
c[$1]=$2","$4;
next
}
END{
print "col1,col2,col4,col6,col7,col10,col12";
for(i in a){
printf("%s,%s,%s,%s\n",i,a[i],b[i]?b[i]:",",c[i]?c[i]:",")
}
}
' file1 file2 file3
Will try to add explanation too in sometime.
EDIT1: adding a one-liner form of solution too.
awk 'BEGIN{FS=";"}FNR==1{f++} f==1 && FNR>1{;a[$1]=$2","$4;next} f>1 && FNR==1{FS=","} f==2&&FNR>1{sub(/_.*/,"",$1);b[$1]=$2","$3;next} f==3&&FNR>1{sub(/_.*/,"",$1);c[$1]=$2","$4;next} END{print "col1,col2,col4,col6,col7,col10,col12";for(i in a){printf("%s,%s,%s,%s\n",i,a[i],b[i]?b[i]:",",c[i]?c[i]:",")}}' file1 file2 file3

Answer 2

sort + sed trick (for sorted input files):
join -t, -j1 -a1 -o1.1,1.2,1.4,2.2,2.3 <(tr ';' ',' < file1) <(sed 's/_[^,]*//g' file2)
| join -t, - -a1 -o1.1,1.2,1.3,1.4,1.5,2.2,2.4 <(sed 's/_[^,]*//g' file3)
The output:
SAMPLE-1,1,1,2,2,3,3
SAMPLE-2,1,1,2,2,3,3
SAMPLE-3,1,1,2,2,3,3
SAMPLE-4,1,1,,,,