Add dump_content function to avoid unclear item_content transformation process

[MHHukiewitz]

In order to avoid mistakes when creating the item_content of a message, the extended_json_encoder has been added from the aleph-sdk-python package to this central piece of the aleph architecture.

Based on #89

[hoh]

Can you explain the issue in more details ? I do not understand what the problem is from this short description.

[MHHukiewitz]

Well, I figured that some functions in connection with messages are too important, to have them spread out around the different repos. IMHO we want:

• The whole item_hash generation process to be contained in this repo as reference for other language implementations (like TypeScript).
• Standardized serialization, like here in dump_content, on which also the item_hash generation process relies.

This way we can also write tests for the validators, to make sure that what transformations we implemented here, are working fine with the current message spec.

[hoh]

This branch is mixing unrelated stuff, can you move black and forbid to other branches ?

[hoh]

Is the code coverage increasing or at least identical after these changes ? I don't see changes in the tests.

[MHHukiewitz]

Is the code coverage increasing or at least identical after these changes ? I don't see changes in the tests.

93% -> 92%, due to the decreasing coverage on the utils.py file from 100% -> 72%. Will increase coverage