CODEWITHSUNDEEP
awk
Grey
Grey

Reputation: 49

How would I print the next line after a matching pattern and then also stop when the matching pattern occurs once again using awk?

I have a SQL file in the following format:

-- 1. Create a view 'c_summary' summarizing campaign contributions,
-- with four attributes: cand_name, contbr_name, amount, and zip. 

create view c_summary as
  select candidate.name as cand_name, contributor.name as contbr_name, amount, zip
    from candidate inner join contributor inner join contribution
      on candidate.cand_id = contribution.cand_id and
         contributor.contbr_id = contribution.contbr_id;

-- 2. For each of the occupations "STUDENT", "TEACHER", and "LAWYER",
-- show the occupation, and average size (in dollars) of contribution
-- from contributors with that occupation.

select occupation, round(avg(amount))
  from contributor natural join contribution
  where occupation in ("STUDENT", "TEACHER", "LAWYER")
  group by occupation;

I am trying to print out the lines that contain the query only when a question number is provided as a number to the PROB variable, and then to stop once another question is reached.

For instance, I want the output to be like this for the second question:

$ awk -f get_query.awk -v PROB=2 queries.sql
select occupation, round(avg(amount))
 from contributor natural join contribution
 where occupation in ("STUDENT", "TEACHER", "LAWYER")
 group by occupation;

Or, the output to be like this for the 1st question:

$ awk -f get_query.awk -v PROB=1 queries.sql
create view c_summary as
  select candidate.name as cand_name, contributor.name as contbr_name, amount, zip
    from candidate inner join contributor inner join contribution
      on candidate.cand_id = contribution.cand_id and
         contributor.contbr_id = contribution.contbr_id;

I tried the following, but the output I am getting is not what I want to get:

/^--/{
    RS="-- "
    getline
}
$0 ~ PROB{
    print $0
}

How would I modify my awk script to get my desired output?

Upvotes: 2

Views: 90

Answers (2)

RavinderSingh13
RavinderSingh13

Reputation: 133528

EDIT: Adding code for removing the control M characters too now in my previous code as OP has control M characters in OP's Input_file.

awk -v PROB=1 '{gsub(/\r/,"")} /^-/{next} !/^ +/&&!/^$/{count++}  count==PROB && NF' Input_file

Could you please try following and let me know if this helps you.

awk -v PROB=2 '/^-/{next} !/^ +/&&!/^$/{count++}  count==PROB && NF'  Input_file

In case you give PROB=1 then following will be the output.

awk -v PROB=1 '/^-/{next} !/^ +/&&!/^$/{count++}  count==PROB && NF' Input_file

Explanation: 

awk -v PROB=2 '                  ##Creating a variable named PROB whose value is 2 here.
/^-/             {  next     }   ##Checking if a line starts from dash then using next it will skip all further statements.
!/^ +/ && !/^$/  {  count++  }   ##Checking if a line not starts from space and is NOT a NULL line then increase the value of variable named count to 1.
count==PROB && NF                ##Checking condition here if variable count value is equal to variable PROB and line is NOT an empty line then print it.
' Input_file                     ##Mentioning the Input_file name here.

Upvotes: 1

Ed Morton
Ed Morton

Reputation: 203615

$ awk -v RS= -v p=2 '(NR/2)==p' file
select occupation, round(avg(amount))
  from contributor natural join contribution
  where occupation in ("STUDENT", "TEACHER", "LAWYER")
  group by occupation;

$ awk -v RS= -v p=1 '(NR/2)==p' file
create view c_summary as
  select candidate.name as cand_name, contributor.name as contbr_name, amount, zip
    from candidate inner join contributor inner join contribution
      on candidate.cand_id = contribution.cand_id and
         contributor.contbr_id = contribution.contbr_id;

Upvotes: 0

Related Questions