Single item vs collection

[phsd0]

Let's say I have endpoint /todos with support of all filters that on demand mode provides this endpoint returns only preview of todo items not their full content. And also /todos/:id which returns exact todo item with all data.

In tanstack query i would use ['todos'] and ['todos', someId] keys accordingly. Lets say todo item have a lot of data attached which isn't required to show list of todo but required to present single todo item on separate page.

How to deal with it using tanstack db? I figured out how to deal with list of items from docs:

const todosCollection = createCollection(
  queryCollectionOptions({
    queryKey: ["todos"],
    queryFn: async () => {
      const response = await fetch("/api/todos")
      return response.json()
    },
    queryClient,
    getKey: (item) => item.id,
  })
)

But how I should define collection for single (expanded, full) todo item? What queryKey should I use? How to query this data?
Maybe I missing something but I didn't find anything in docs about this case

[oschade]

I would create a new collecton for the detailed todo-items and use the "on-demand"-mode, something like this:

const TodoDetailSchema = z.object({
  id: z.string(),
  title: z.string(),
  description: z.string(),
  isCompleted: z.boolean(),
})

type TodoDetail = z.infer<typeof TodoDetailSchema>

const getId = (where: LoadSubsetOptions['where'] | undefined): string => {
  const comparison = extractSimpleComparisons(where)

  let id = ''

  comparison.forEach(({ field, operator, value }) => {
    if (field[0] === 'id') {
      if (operator === 'eq') {
        id = value
      } else if (operator === 'in') {
        id = value[0]
      }
    }
  })

  return id
}

export const todoDetailCollection = createCollection(
  queryCollectionOptions({
    queryKey: (ctx) => {
      const id = getId(ctx.where)

      if (!id) {
        throw new Error('id is required')
      }

      return ['todos', 'detail', id]
    },
    queryFn: async (ctx) => {
      const id = getId(ctx.meta?.loadSubsetOptions?.where)

      if (!id) {
        throw new Error('id is required')
      }

      const response = await fetch(`/api/todos/${id}`).then((res) => res.json())
      return [response] as TodoDetail[]
    },
    schema: TodoDetailSchema,
    queryClient,
    syncMode: 'on-demand',
    getKey: (item) => item.id,
  })
)

For the sake of the inArray-comparsion operator, it would be nice to be able to create multiple query-keys at once inside the queryKey-fn.