0

I am trying to print only e-mail addresses from a file, file format is as follows:

{"12":"34","email":"[email protected]","firstname":"john","lastname":"smith","text":"0","sig":"js","..":"2017-08-23"}

How can I just extract using awk -F? So I can use awk -F'"' '{print $8}' but the 8th column may not always be the e-mail address.

Is there a way to use awk -F WITH a regular expression of '@'? Or print all texts in the delimiter in separate lines then pipe it to | grep '@'

Thanks

2
  • If awk isn't a strict requirement, you could write a small script using language which includes a json parser (such as python).
    – sebasth
    Aug 23, 2017 at 11:22
  • Better use jq, which will permit you to get the value of attribute "email" Aug 23, 2017 at 11:41

3 Answers 3

1

jq is a proper tool for parsing/manipulating json data:

jq '.email' jsonfile

The output:

"[email protected]"
8
  • Add - r to be rid of the quotes.
    – Kusalananda
    Aug 23, 2017 at 12:10
  • @Kusalananda, you shouldn't decide instead of OP. If the OP will mention about that requirement - it'll be added Aug 23, 2017 at 12:19
  • Sorry, my intention was to mean "if you want to be rid of the quotes (in the output)", not a critique on the answer as such. jq is definitely the correct solution here.
    – Kusalananda
    Aug 23, 2017 at 12:22
  • jq is only the correct answer to the question: How can I print a specific field. It is not the answer to this question, which specifically asks how it is done using awk. Aug 23, 2017 at 12:24
  • @garethTheRed, even if the question is "How can I just extract using awk -F from JSON". - using awk for json is a bad practice, fragile solution and a bad advice for the future Aug 23, 2017 at 12:28
0

This is an awk way.

awk -v RS=\" '/@/' infile.txt 

The above is correct when you have quoted fields, else you should have RS with multiple RecordSeperators. like RS='[[:blank:]]*:' per your sample in your comment.

2
  • I get 12:34 email:[email protected] firstname:john lastname:smith text:0 sig:js ..:2017-08-23 with this :-( Aug 23, 2017 at 11:52
  • so where is the quotes there you have in your shared sample in body of your question? if no quotes change RS to RS='[[:blank:]]*:' to get it to work, also I don't suggest to use awk where there is correct jq tool for that Aug 23, 2017 at 11:57
0

The reg-ex may need tweaking to be fully compliant with RFC 5322, but as a starter, try:

awk '{match($0,"[a-zA-z0-9._%+-]+@[a-zA-z0-9._%+-]",a)}END{print a[0]}'

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .