117 Emodule mapping script

[alFrie]

Pull request for issue #117 .

[alFrie]

The mapping script from Ilias has been now replaced with a mapping script like the one used for SimpleOntology. It is not finished yet!
Here we don't append the name of the specimen (as stated in the metadata-yaml) to the individuals, but instead we want to create an unique ID within the script and append it. One possible way of doing this, which got mentioned during the Hackathon, is this one: https://pypi.org/project/python-datauri/

Edit: Or this one might be even better https://docs.python.org/3/library/uuid.html

[alFrie]

Changes at the emodule-metadata-extraction-script:
I have changed the names of the keys for the metadata dictionary to fit the placeholders, so that the mapping script can work.
Some information has been hardcoded: the specimen age, the lab location and the transducer/compression column.

There is still more metadata extracted than needed for mapping (f.e. the units and information about the software), but I decided to not delete that code for now, in case we want to add placeholders for them in the ontology. (I could also comment them out, for now they will be in the generated yaml-file). I marked these useless keys with a comment.

The transducer-value ([1,2,3]) needs to be fixed, it seems like a list can't be passed easily to the mapping script. On that point, also the names of the placeholders for transducer column and compression column inside the ontology might need to be changed, since they don't fit the corresponding individuals and the mapping script can't work with that.

[joergfunger]

Could you check the tests, they are not working.

[alFrie]

Could you check the tests, they are not working.

@soudehMasoudian The mapping test is outdated, because it depends on Ilias old mapping script. Could you please look into this?

[alFrie]

I noticed that my script fails to append the unique ID after a lot of underscores, like here. I will fix this now.

Edit: Done.

[alFrie]

I think this script is almost finshed now. Tests are said to fail because the tests don't fit the script but a very old version of it. See #134 for this.