Write a regular expression to parse following string - regex

Question

I am using sed command and I want to parse following string:
Mr. XYZ Mr. ABC, PQR
Ward-2, abc vs. MG Road, Pune,
Pune Dist.,
(Appellant) (Respondent)
Now I want to parse the above string and I want to get Appellant part separated from above example and respondent part separated.
That is I want following output:
Mr. XYZ Ward-2, abc(Appellant) that is one output and Mr. ABC, PQR MG Road, Pune, Pune Dist.,(Respondent) is another output by using sed command.
I used following regex but not getting proper output:
sed -n '/assessment year/I{ :loop; n; /Respondent/Iq; p; b loop}' abc.txt

Answer 1

sed is always the wrong tool for any job that involves looking at multiple lines. Just use awk, it's what it was invented for. Here's GNU awk for a couple of extensions:
$ cat tst.awk
BEGIN { FIELDWIDTHS="30 7 99" }
{
for (i=1;i<=NF;i++) {
gsub(/^\s*|\s*$/,"",$i)
if ($i != "") {
rec[i] = (rec[i]=="" ? "" : rec[i] " ") $i
}
}
}
/^\(/ {
print rec[1]
print rec[3]
delete rec
}
$
$ awk -f tst.awk file
Mr. XYZ Ward-2, abc (Appellant)
Mr. ABC, PQR MG Road, Pune, Pune Dist., (Respondent)

Answer 2

I achieved this with following way by using ruby:
appellant_respondent = %x(sed -n '/assessment year/I{ :loop; n; /respondent/Iq; p; b loop}' #{#file_name}).split("\n")
appellant_name_array = []
respondent_name_array = []
appellant_respondent.delete("")
appellant_respondent.each do |names|
names_array = names.split(/\s+\s+/)
appellant_name_array << names_array.first if names_array.first != ""
respondent_name_array << names_array.last if names_array.last != ""
end
#item[:appellant] = appellant_name_array.join(' ').gsub(/\s+vs\.*\s+/i, ' ').strip
#item[:respondent] = respondent_name_array.join(' ').gsub(/\s+vs\.*\s+/i, ' ').strip