`get_snapshot_value` only works when at least one assertion has been made

[qdewaghe]

For example, this test fails.

from inline_snapshot import get_snapshot_value, snapshot


def test_no_assert():
    assert get_snapshot_value(snapshot([])) == []

    def test_no_assert():
>       assert get_snapshot_value(snapshot([])) == []
E       assert Ellipsis == []
E        +  where Ellipsis = get_snapshot_value(Ellipsis)
E        +    where Ellipsis = snapshot([])

Calling _visible_value() in in unwrap does not seem quite right to me.

[qdewaghe]

I have made a PR for this but I suspect there might be subtleties when the snapshot gets updated but does not fail the test.

def test():
    s = snapshot(0)
    s == 1  # no assert
    assert get_snapshot_value(s) == 1

But of course after the snapshot is updated that would be fine 🤷

[15r10nk]

picking the correct value is tricky here.

I think the following should work, because the semantic would otherwise be confusing.

s=snapshot("some_value")

assert "x" == s
assert "x" == get_snapshot_value(s)

but it would fail when you return always the _old_value and run --inline-snapshot=fix.

So it should be maybe the new_value with a fallback to the old_value?