Deduplication

```
sqlite> select count(distinct isbn) from books;
1525
sqlite> select count(*) from books;
7031
```

It would be great to have some support for removing duplicates based on ISBN or file hash; my library is served from a NAS that has lots of links.