Back (Current repo: scraps)

group_by_column_regex.awk (764B)

#!/bin/awk

function push(arr, value) {
    arr[length(arr)+1] = value;
}

# Helper to print the array in semicolon-delimited format
function parray(arr) {
    for (i = 0; i < length(arr); i++) {
        printf arr[i] ";";
    }
    printf "\n";
}
BEGIN{
    FS=",";
    OFS=",";
    regex="";
    if (prefix == ""){prefix = 2}
    split("",a);
}
{
    #e.g. we want to group together rows that begin with the same two numbers, like 50,501,5010, etc
    #because the file contains rows as a hierarchy
    if(length($1)==prefix) {
        if(length(a) > 0) {
            parray(a);
        }
        split("",a);
        regex="^"$1
        push(a,$1"@"$2);
    } else {
        if($1 ~ regex) {
            push(a,$1"@"$2);
        }
    }
}
END{
    parray(a);
}