AWK 字段分隔符新行以匹配开头

Question

I have a file with repeat attributed seperated by one common value -- .

The idea is to take what is in between and use awk to print the new line values in one line so I can sort by unique count.

File:

--
country = canada
region = ontario
isp = bell
type = isp
--
country = usa
region = newyork
isp = comcast
type = isp
--

Desired output:

country = canada, region = ontario, isp = bell, type = isp
country = usa, region = newyork, isp = comcast, type = isp

Answer 1

There are a few ways to accomplish this.

One would be to reset/print a string variable every time you hit the -- lines, and in between concatenate line contents to that variable.

awk '/^--$/ { 
  if(line) print substr(line,3);
  line=""; 
  next; 
}
/\S/ { line = line ", " $0; }'

Another would be to change the record separator (RS) to --\n , and then use some regex subs to replace the CRs with commas:

awk -v RS="--\n" '{ 
  gsub(/^[\n \t]+|[\n \t]+$/, "");
  gsub(/\n+/, ", "); 
  if($0) print $0; 
}'

Answer 2

awk -F "\n" -v RS="--" -v OFS=", " 'NF >= 5 { print $2, $3, $4, $5 }' file