Separate task history end point

[Sujanadh]

Updates

I tried updating task history in task-list. It was working fine. But to apply sorting and filtering operations for history list, it will be quite hectic since it is inside list of nested json. Using separate endpoint for only task history seems good idea and makes easy to apply filter.
This PR address this feature.
And yes, there are so many schemas with similar name, I will restructure those and use only required schemas.

[Sujanadh]

Response

[
  {
    "task_id": 605,
    "action_text": "Status changed from LOCKED_FOR_MAPPING to MAPPED by: Sujan Adhikari",
    "action_date": "2023-12-20T07:44:48.339376",
    "status": "MAPPED",
    "username": "Sujan Adhikari",
    "profile_img": "https://www.openstreetmap.org/rails/active_storage/representations/redirect/eyJfcmFpbHMiOnsiZGF0YSI6MzI4NTE3MTksInB1ciI6ImJsb2JfaWQifX0=--6b73d47fac28854f227f3d0ecd56f2f6554c8709/eyJfcmFpbHMiOnsiZGF0YSI6eyJmb3JtYXQiOiJwbmciLCJyZXNpemVfdG9fbGltaXQiOlsxMDAsMTAwXX0sInB1ciI6InZhcmlhdGlvbiJ9fQ==--473f0190a9e741bb16565db85fe650d7b0a9ee69/sujan.webp"
  },
  {
    "task_id": 605,
    "action_text": "Status changed from READY to LOCKED_FOR_MAPPING by: Sujan Adhikari",
    "action_date": "2023-12-20T07:44:23.825255",
    "status": "LOCKED_FOR_MAPPING",
    "username": "Sujan Adhikari",
    "profile_img": "https://www.openstreetmap.org/rails/active_storage/representations/redirect/eyJfcmFpbHMiOnsiZGF0YSI6MzI4NTE3MTksInB1ciI6ImJsb2JfaWQifX0=--6b73d47fac28854f227f3d0ecd56f2f6554c8709/eyJfcmFpbHMiOnsiZGF0YSI6eyJmb3JtYXQiOiJwbmciLCJyZXNpemVfdG9fbGltaXQiOlsxMDAsMTAwXX0sInB1ciI6InZhcmlhdGlvbiJ9fQ==--473f0190a9e741bb16565db85fe650d7b0a9ee69/sujan.webp"
  }
]

[spwoodcock]

Thanks for the example response, very useful 😄

What immediately stands out is that the profile image URL can be quite long, but is replicated in each item.

Not a big deal if the history is only a few items. But if there are 100s of entries, then it's a bit inefficient to send all that extra data in the json payload. I wonder if there is a way to optimise.

[spwoodcock]

See my comment on usage of @computed_fields.

I'm not sure what this endpoint is needed for, but I will merge it & it can be updated as needed.

[spwoodcock]

I moved all the logic to the pydantic model.

Note we can use @computed_field decorators to generated computed values.
We can also use this to generate values based on database relationships.
To do this we need the model_config from_attributes=True:

    model_config = ConfigDict(
        from_attributes=True,
    )

[Sujanadh]

See my comment on usage of @computed_fields.

I'm not sure what this endpoint is needed for, but I will merge it & it can be updated as needed.

it was for the purpose of applying filters, ordering on the task history.

[Sujanadh]

i couldn't get back to it as i got busy on other things 😃

[spwoodcock]

No worries! The code is merged & we can built on it at another point if needed 👍