Documentation for metadata extraction

Docling lists "Metadata extraction, including title, authors, references & language" as a feature, however I don't see anything in the documentation about how one gets these values out.

I've tried:
```
docling --to json https://assets-eu.researchsquare.com/files/rs-66715/v1/2c370dd5-7279-4a77-89dc-94297f0a7843.pdf\?c\=1637243468
```

and
```
docling --to json https://arxiv.org/pdf/1904.09664
```

But it doesn't seem to find any metadata. Am I using it correctly?