How can I grep columns that match in the same row?

Question

I want to get the end date for an installed subscription, and currently using

subscription-manager list --installed | grep Ends | awk '{print $2 }')

The complexity comes when there are multiple subscriptions. How can I get just the end date and name that matches a subscription where the name has "Server" in it?

+-------------------------------------------+
    Installed Product Status
+-------------------------------------------+
Product Name:   Red Hat Enterprise Linux Server
Product ID:     69
Version:        7.7
Arch:           x86_64
Status:         Subscribed
Status Details:
Starts:         27/11/19
Ends:           27/11/20

Product Name:   Red Hat Enterprise Linux High Availability for x86_64
Product ID:     83
Version:        7.7
Arch:           x86_64
Status:         Subscribed
Status Details:
Starts:         27/11/19
Ends:           27/11/20

Answer 1

This is basically a shorter version of AdminBee's approach:

$ subscription-manager list --installed | 
    awk '{
            if(sub(/^Product Name: */,"")){
                a=0; 
                n=$0;
                if(/<Server>/){ a++ }
            } 
          }/Ends/ && a{print n,$2}'  
Red Hat Enterprise Linux Server 27/11/20

The script will set a to 0 if the current line begins with Product Name:, since we want to reset it every time and then, if the line also has the string Server, it will set it to 1. Then, if the current line starts with Ends:, and a is not 0, it will print the second field.

Note that this will also match for a Product Name: that contains notAServer or anything else that contains Server. To limit to word matches only, you can use this instead (assuming you have gawk, GNU awk):

subscription-manager list --installed | 
    gawk '{if(sub(/^Product Name: */,"")){a=0; n=$0;if(/\<Server\>/){a++}} }/Ends/ && a{print n,$2}'

Answer 2

Please note that it is rarely necessary to combine awk, sed and grep for a single task. In your case, the following should work:

subscription-manager list --installed | awk '$2=="Name:" && $0~/Server/ {relevant_section=1;print;} \
$1=="Ends:" && (relevant_section==1) {relevant_section=0;print;}'

This will set a flag relevant_section to 1 if a line that starts with "Product Name:" (or rather, where the second space-separated token is Name:) contains the string Server. For convenience, this line is also printed.

When a line starting with Ends: is encountered while the flag is 1, it will print that line and reset the flag to 0.

If you really only want the end date printed, you can shorten it as:

subscription-manager list --installed | awk '$2=="Name:" && $0~/Server/ {relevant_section=1;} \
$1=="Ends:" && (relevant_section==1) {relevant_section=0;print $2;}'

Answer 3

Using Miller

$ tail -n +4 subscriptions | mlr --xtab --ips : --ops ': ' \
    filter '${Product Name} =~ "Server"' then cut -f 'Product Name,Ends'
Product Name:    Red Hat Enterprise Linux Server
Ends:            27/11/20