how to remove some strings in emeditor(Regex ) - regex

Question

i want to remove some string and save other part of string that i need from a file with emeditor ..
file line like :
{"message":"{\"_\":\"user\",\"pFlags\":{\"contact\":true},\"user_flags\":2143,\"id\":702212125,\"access_hash\":\"914250561826\",\"first_name\":\"david\",\"last_name\":\"jones\",\"username\":\"david_d192\",\"phone\":\"051863329875\",\"status\":{\"_\":\"userStatusRecently\"}}","phone":"051863329875","version":"3","type":"unknown","token":"1556189892619764206","p_id":702212125,"username":"david_d192","type":"redis","user_flags":2143,"host":"win",from":"contacts"}
{"index": {"_type": "_doc", "_id": "36GG54F"}}
{"message":"{\"_\":\"user\",\"pFlags\":{\"contact\":true},\"user_flags\":2143,\"id\":702212125,\"access_hash\":\"914250561826\",\"first_name\":\"david\",\"last_name\":\"jones\",\"username\":\"david_d192\",\"phone\":\"051863329875\",\"status\":{\"_\":\"userStatusRecently\"}}","phone":"051863329875","version":"3","type":"unknown","token":"1556189892619764206","p_id":702212125,"username":"david_d192","type":"redis","user_flags":2143,"host":"win",from":"contacts"}
{"index": {"_type": "_doc", "_id": "36GG54F"}}
{"message":"{\"_\":\"user\",\"pFlags\":{\"contact\":true},\"user_flags\":2143,\"id\":702212125,\"access_hash\":\"914250561826\",\"first_name\":\"david\",\"last_name\":\"jones\",\"phone\":\"051863329875\",\"status\":{\"_\":\"userStatusRecently\"}}","phone":"051863329875","version":"3","type":"unknown","token":"1556189892619764206","p_id":702212125,"type":"redis","user_flags":2143,"host":"win",from":"contacts"}
{"index": {"_type": "_doc", "_id": "36GG54F"}}
i want to save id, first_name , last_name , phone , username(if exist) in every line =>
id:702212125 first_name:david last_name:jones phone:[phone] username:david_d192,
id:702212125 first_name:david last_name:jones phone:[phone] username:david_d192,
id:702212125 first_name:david last_name:jones phone:[phone],
how i can do this ?
thanks

Answer 1

JSON parsing is the optimal way to do this (https://linuxconfig.org/how-to-parse-data-from-json-into-python). But you can make life harder and use regex (here presented in PCRE (PHP) flavor):
Get all id's:
(?<=id\":\s\")(\w+)(?=\")
See example:
https://regex101.com/r/g5vfEd/1
Get all first names:
(?<=first_name\\\":\\\")(\w)+(?=\\)
See example:
https://regex101.com/r/g5vfEd/2
Get all last names:
(?<=last_name\\\":\\\")(\w)+(?=\\)
See example:
https://regex101.com/r/g5vfEd/3
Get all phone numbers:
(?<=phone\\\":\\\")(\w)+(?=\\)
See example:
https://regex101.com/r/g5vfEd/4
Get all user names if they exist:
(?<=username\\\":\\\")(\w)+(?=\\)
See example:
https://regex101.com/r/g5vfEd/5
complete pattern to match everything:
id\\?\":\s?\"?(\w+),?[\\\"].*first_name\\\":\\"(\w+).*last_name\\\":\\\"(\w+).*phone\":\"(\d+).*(?=username)?\":\"(\w+).*
Returns 3 matches, each with the following 5 groups (here match 1 is shown):
Group 1. 85-94 702212125
Group 2. 145-150 david
Group 3. 169-174 jones
Group 4. 285-297 051863329875
Group 5. 454-462 contacts
See link: https://regex101.com/r/g5vfEd/6

Answer 2

As you've tagged regex and Emeditor you can try this.
Emeditor version 19.1 onwards supports regex named groups like this:
(?<id>expression)
and named backreference by using this form:
\k<id>
So steps:
Find and Replace (Ctrl-H). Tick "Match Case" and select "Regular Expressions".
Find:
\\"id\\"[\\":]*(?<id>[^\\":,]*).*?\\"first_name\\"[\\":]*(?<first_name>[^\\":,]*).*?\\"last_name\\"[\\":]*(?<last_name>[^\\":,]*).*?\\"phone\\"[\\":]*(?<phone>[^\\":,]*)(.*?"username"[\\":]*(?<username>[^\\":,]*))?
Replace with:
id:\k<id>\tfirst_name:\k<first_name>\tlast_name:\k<last_name>\tphone:\k<phone>\tusername:\k<username>
Click the down Arrow next to the Extract button and select "To New Document"
Click the Extract button to output to a new tab delimited file.