Closed annehaley closed 1 year ago
@annehaley the latest decision should be recorded in the data frame - however the scan notes should be a combination of all decision notes starting with the oldest one ending with the current one - does that make sense ?
@kipohl Yes, I can do that. Is there any particular way you want to combine the strings?
Acronym of person: note; acronym of person: note; ...
I can't get the initials of the person reliably from the value of the creator field, but I can include the whole "creator" value instead.
Give me an example of creator name? I removed the acronym is already in the scan notes in miqa but I might have misremembered
Maybe just take the first three letters than or something like that
the value of "creator" will be a string of the creator's MIQA username. I changed it so we record the first three letters.
I just added a comment - see below
if it is a lot of work lets leave first three letters
In order for a MIQA import to correctly interpret the creator of a decision as an existing MIQA user, the value for "creator" has to be the correct MIQA username. How is that first decision written in your import file?
The second decision, I assume, was made with the MIQA UI. It saves the decision creator properly. It can show the initials of the user because it has access to the full object of user information, unlike in the import file format.
the first one was done by miqa 1 - if there is not a creator then you omit it simply yes the second decision was done by MIQA - so if you have the initials it that point it should be easy to include them in the json file as another tag under decisions - I might be wrong then again ignore
We could turn the "creator" value into a whole object (in the JSON format only), so it would include username, first_name, and last_name (and any other info we want to export about the decision creator). But this might take a bit to update the import schema again.
lets just stay with first three letters - when is it ready to be committed
It's ready now if we are sticking with the first three letters.
There is one section where I am not sure of the desired behavior: in the function
convert_json_to_check_new_sessions_df
, if the input json has a key "decisions" with multiple decision objects, how should these be recorded to the output dataframe? Currently, I record the first entry only. How is this data structure used and can it hold multiple decisions?