Extract Regex Pattern For Each Line - Leave Blank Line If No Pattern Exists - awk

Question

I am working with the following input:
"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"phone":"549-287-5287","city":"Marshall","gender":"female","email":"mortina.curabia#gmail.com"
I need to be able to extract both the phone number and email of each line into separate files. However, both values don't always appear in the same field - they will always be prefaced with "phone": or "email":, but they may be in the first, second, third or even twentieth field.
I have tried chopping together solutions in SED and AWK to remove everything up until "phone" and then every after the next , but this doesn't not work as desired. It also means that, if "phone" and/or "email do not exist, the line is not changed at all.
I need a solution that will give me an output with the phone value of each line in one file, and the email value in another. HOWEVER, if no phone or email value exists, a blank line in the output needs to be in place.
Any ideas?

Answer 1

This might work for you (GNU sed):
sed -Ene 'h;/.*"phone":([^,]*).*/!z;s//\1/;w phoneFile' -e 'g;/.*"email":([^,]*).*/!z;s//\1/;w emailFile' file
Make a copy of line.
If the line does not contain a phone number empty the line, otherwise remove everything but the phone number.
Write the result to the phone number file.
Replace the current pattern space by the copy of the original line.
Repeat as above for an email address.
N.B. My first attempt used s/.*// instead of z to empty the line which worked but should not have. If the line contained no phone/email, the substitution should have reset default regexp and the second substitution should have objected that it did not contain a back reference. However the second substitution worked in either case.

Answer 2

After fixing your file to be valid json and adding an extra line missing the phone attribute so we can test more of your requirements:
$ cat file
{"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"phone":"549-287-5287","city":"Marshall","gender":"female","email":"mortina.curabia#gmail.com"}
{"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"city":"Marshall","gender":"female","email":"foo.bar#gmail.com"}
you can do whatever you like with the data:
$ jq -r '.email // ""' file
mortina.curabia#gmail.com
foo.bar#gmail.com
$
$ jq -r '.phone // ""' file
549-287-5287
$
As long as it doesn't contain embedded newlines you can used sed 's/.*/{&}/' file to convert the input in your question to valid json as in my answer:
$ cat file
"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"phone":"549-287-5287","city":"Marshall","gender":"female","email":"mortina.curabia#gmail.com"
"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"city":"Marshall","gender":"female","email":"foo.bar#gmail.com"
$ sed 's/.*/{&}/' file
{"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"phone":"549-287-5287","city":"Marshall","gender":"female","email":"mortina.curabia#gmail.com"}
{"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"city":"Marshall","gender":"female","email":"foo.bar#gmail.com"}
$ sed 's/.*/{&}/' file | jq -r '.email // ""'
mortina.curabia#gmail.com
foo.bar#gmail.com
but I'm betting you started out with valid json and removed the {} by mistake along the way so you probably just need to not do that.

Answer 3

Using grep
Try:
grep -o '"phone":"[0-9-]*"' < Input > phone.txt
grep -o '"email":"[^"]*"' <Input > email.txt
Demo:
$echo '"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"phone":"549-287-5287","city":"Marshall","gender":"female","email":"mortina.curabia#gmail.com"' | grep -o '"phone":"[0-9-]*"'
"phone":"549-287-5287"
$echo '"visit_date":{"$date":"2017-11-28T04:43:00.000Z"},"phone":"549-287-5287","city":"Marshall","gender":"female","email":"mortina.curabia#gmail.com"' | grep -o '"email":"[^"]*"'
"email":"mortina.curabia#gmail.com"
$