refactor musicbrainz crawling

[sertansenturk]

One-line summary (optional)

Reason

Design

1.1 when parsing url, we should get both url and type (work, recording, etc.)
1 2. Crawler part should also receive the type and parse metadata received from musicbrainz accordingly instead of duck typing.

Acceptance Criteria
1.3 unit tests, mocks and integration tests
1.4 inline documentation

Outcome
1.3 unit tests, mocks and integration tests
1.4 inline documentation

Don't forget to

• Point the ticket
• Set priority as a label
• Assign to a person
• Link dependencies, if applicable