Book created from a dict discards ordering

[craiga]

Since Python 3.7, dictionaries preserve order, so I'd expect the following code to pass:

book = pyexcel.Book(
    {
        "Stuff": [["some", "data", "I", "can't", "parse"]],
        "Records": [["record_id"], ["a"], ["b"], ["c"]],
        "Quarters": [
            ["id", "quarter"],
            ["one", "2021Q1"],
            ["two", "2021Q2"],
        ],
    }
)
assert list(book.sheet_names()) == ["Stuff", "Records", "Quarters"]

Unfortunately it does not.

E       AssertionError: assert ['Quarters', ...rds', 'Stuff'] == ['Stuff', 'Re...', 'Quarters']
E         At index 0 diff: 'Quarters' != 'Stuff'
E         Full diff:
E         - ['Stuff', 'Records', 'Quarters']
E         + ['Quarters', 'Records', 'Stuff']

I'd expect sheet_names to return the sheet names in the same order as the source dict.

[craiga]

Digging into this, it looks like load_from_sheet is explicitly checking for OrderedDict and re-ordering the keys if the dictionary isn't a subclass of OrderedDict.

I can work around this problem by casting my dictionary to an OrderedDict, but ideally I shouldn't have to, right?