manage.py dumpdata outputs YAML in unhelpful order

[sampablokuper]

The ​docs suggest providing initial YAML data in the following format:

- model: myapp.person
  pk: 1
  fields:
    first_name: John
    last_name: Lennon

but the manage.py dumpdata outputs YAML in the following form:

-   fields: {first_name: John, last_name: Lennon}
    model: myapp.person
    pk: 1

The fact that this is flow rather than block style YAML is the subject of ​#11927, but there's another problem, which is that rather than putting the model and PK first (which is most human-readable), dumpdata outputs the fields first.

[sampablokuper]

​This might be helpful for generating a patch.

[Aymeric Augustin]

Change UI/UX from NULL to False.

[Aymeric Augustin]

Change Easy pickings from NULL to False.

[Tim Graham]

The current ​PR needs a test.

[Tobias Kunze]

Since on our way to 3.0 we have dropped support for Python < 3.6, dictionaries can be assumed to be ordered, and YAML output was already in the order suggested in this ticket. I submitted a PR to guard against regressions, though.

[Mariusz Felisiak]

Fixed in 5bc3123479bd97dc9d8a36fa9a3421a71063d1da.

[Mariusz Felisiak <felisiak.mariusz@…>]

In a9b88982:

Refs #11929 -- Fixed ordering of test YAML dumps.

Thanks Nick Pope for the review.