implement a callback checking for malicious delete password

[ptrf]

As the current hushfile API.md mentions:

It is possible for a malicious client to construct an upload where the deletepassword
POSTed to the server is different from the one inside the encrypted metadata blob.

The idea is to proactively check whether a different delete password was included in the metadata, by calling an api-endpoint immediately after decoding the metadata blob.

Such a check would probably be fit to call in the try block here https://github.com/hushfile/hushfile-web/blob/master/hushfile-download.js#L23

It would also be nice with a notification saying "Warning: Delete password supplied is incorrect, possibly malicious file" or something like that.

[rlindsgaard]

Invalid due to new API spec. See documentation.