I'm working with exported ticket conversations as .JSON
In the provided conversation history, this is a correct format of a message:
"body": "Hello, I want to check if i need an item from IKEA Abu Dhabi the item cost is AED 145 you will be able to get it and what will be your charges. Thank you.",
Now I see multiple instances where the format is not right, like in the following:
"body": "(12:50:51 PM) Elena: Hello, I am sorry I didn\u2019t reply yesterday
(12:51:10 PM) Elena: I have a problem with my fridge. The door doesn\u2019t close properly, I think the problem is the seal. But maybe something else.
I need someone who will fix the door for me
(12:51:34 PM) Elena: If you have time tomorrow to send a technician that would be lovely
(12:51:39 PM) Elena: Any time today is fine
(12:51:39 PM) Fatma Mohamed: Where you located please and when do you need the service?",
As you can see, there are many messages saved in a single instance. There should be just one. Also, there are timestamps and usernames written inside the message.
Finally, the author_id is set to -1 which makes it impossible for me to identifty who sent what
The fact that we have different structures make it very difficult to parse.
Why is it like this and how could this be solved?