Wrong format for exceptions?

I was looking into the [ABNF of SPDX](https://spdx.github.io/spdx-spec/v3.0.1/annexes/spdx-license-expressions/#) and was wondering about the exceptions that are listed in the Scancode licensedb.

The ABNF states:
```
addition-expression = license-exception-id / addition-ref
```
with
```
license-exception-id =
addition-ref = [%s"DocumentRef-"(idstring)":"]%s"AdditionRef-"(idstring)
idstring = 1*(ALPHA / DIGIT / "-" / "." )
```

This brings me to the conclusion that exceptions from Scancode's license DB should start with `AdditionRef` instead of `LicenseRef`. However, currently, they start with `LicenseRef`, e.g.:
- [`LicenseRef-scancode-apromore-exception-2.0`](https://scancode-licensedb.aboutcode.org/apromore-exception-2.0.json)
- [`LicenseRef-scancode-aptana-exception-3.0`](https://scancode-licensedb.aboutcode.org/aptana-exception-3.0.json)
- [`LicenseRef-scancode-bash-exception-gpl-2.0`](https://scancode-licensedb.aboutcode.org/bash-exception-gpl.json)

Or is it intended to have the `AdditionRef` as prefix, e.g.

`GPL-2.0-or-later WITH AdditionRef-LicenseRef-scancode-apromore-exception-2.0`

in contrast to

`GPL-2.0-or-later WITH AdditionRef-scancode-apromore-exception-2.0`?