Adding for loop to sed process

Question

Have a script

#!/bin/bash

sed -E 's/([^,]*,([^,]*),) ?(([[:alpha:]])[^ ]* +)(([^,]*),[^,]*,)[^,]*/\1\u\3\u\5\L\4\6\[email protected]/' file.csv > output.csv

Have a file.csv:

id,location_id,name,title,email,directorate
1,1, Amy Lee,Singer,,, 
2,2,brad Pitt,Actor,[email protected],Production 
3,5,steven spielberg,Producer,Screenwriter, [email protected],Production
4,8,Andy lee,Comedian,,Radio

A few problem, that I need resolve:

title value can be more than one title, example: Steve Spielberg - Producer,Screenwriter. Now the script cuts off the value after the comma, but I need to save all titles.
script concatenate first letter of first name and a last name, plus location_id and @google.com, but I need add location_id, only when have a equals emails.

In the end it should be:

id,location_id,name,title,email,directorate
1,1, Amy Lee,Singer,[email protected],, 
2,2,Brad Pitt,Actor,[email protected],Production 
3,5,Steven Spielberg,Producer,Screenwriter,[email protected],Production
4,8,Andy Lee,Comedian,[email protected],Radio

If you can have commas in the field of a CSV file, the field should be delimited with quotes. Unfortunately, regular expressions are generally not powerful enough to parse formats like that. — Barmar
– Barmar, Commented Nov 9, 2022 at 17:22
Generally not solvable imo. The floating entries could be associated to any column. Even if you check the @ in email and deduce its location you're out of luck if the email is missing. — Andre Wildberg
– Andre Wildberg, Commented Nov 9, 2022 at 17:42
@AndreWildberg maybe not with sed? Is there any other way to solve this problem from the start? — Art
– Art, Commented Nov 9, 2022 at 17:45
With a lot of assumptions (e.g. names have spaces etc) it could be solved for a special case/file. But there is no stable general approach other than a correct user input with quotes around the field. Imagine this entry ,,,,,,,,,,. Which one is the entry with comma inside a field? Only correct user input will tell you e.g. ,,,",,,,,",, — Andre Wildberg
– Andre Wildberg, Commented Nov 9, 2022 at 18:06
Making up random email addresses is a bad idea; you end up having spammers pick them up and send unwanted messages to real people if those addresses happen to be taken. Use @example.com for examples; it's guaranteed to not have this problem. — tripleee
– tripleee, Commented Nov 9, 2022 at 18:08

dan · Accepted Answer · 2022-11-10 00:38:51Z

0

Assuming that directorate doesn't contain commas:

sed -E 's/([^,]*,([0-9]+), *([[:alpha:]])[^ ]* *([[:alpha:]]*).*,)(,[^,]*$)/\1\3\4\[email protected]\5/'

This always adds the location id number. It is possible to do it only for duplicate example emails by adding a second filter. Also consider using ID (first column) instead, in case of duplicate location + name.

answered Nov 10, 2022 at 0:38

dan

5,4119 silver badges17 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Art Over a year ago

and how I can add a new filter?

Collectives™ on Stack Overflow

Adding for loop to sed process

1 Answer 1

1 Comment

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

1 Comment

Your Answer

Sign up or log in

Post as a guest

Linked

Related