Upload records to the OSSR on notebook and improve __repr__ on Records class

improve repr on Records class will dump a formatted json string instead of all the answer.

Testing this in a notebook, I don't see a real improvement. Instead of

{"conceptdoi": "10.5281/zenodo.5524912", "conceptrecid": "5524912", "created": "2021-09-23T15:03:51.644210+00:00", "doi": "10.5281/zenodo.5524913", "files": [{"bucket": "165e5f9a-17bc-4792-bd04-b01cef8b9fc7", "checksum": "md5:20d7ce7abc06541bed2655f19edfb956", "key": "codemeta.json", "links": {"self": "https://zenodo.org/api/files/165e5f9a-17bc-4792-bd04-b01cef8b9fc7/codemeta.json"}, "size": 3271, "type": "json"},
...

If get

'{\n    "conceptdoi": "10.5281/zenodo.5524912",\n    "conceptrecid": "5524912",\n    "created": "2021-09-23T15:03:51.644210+00:00",\n    "doi": "10.5281/zenodo.5524913",\n    "files": [\n        {\n            "bucket": "165e5f9a-17bc-4792-bd04-b01cef8b9fc7",\n            "checksum": "md5:20d7ce7abc06541bed2655f19edfb956",\n            "key": "codemeta.json",\n            "links": {\n                "self": "https://zenodo.org/api/files/165e5f9a-17bc-4792-bd04-b01cef8b9fc7/codemeta.json"\n            },\n            "size": 3271,\n            "type": "json"\n        },\n
...

The line breaks are now displayed and I don't feel like this is more readable.

Is this a config problem?

Where are you running the Notebook ? I get this (in my local Notebook, not in the one rendered by the MR).

I don't get that at all in a local notebook

I think it's my ipython display that is broken

Two general comments:

• it might be worth splitting the notebook in two:

  • "explore the OSSR"
  • "upload record to the OSSR"

• I was actually thinking today that this representation, however exhaustive and what is the most conform to Python convention is actually really annoying and that informative this it is not really readable. What do you think about changing the __repr__ to Record(self.id), provide a .peek() method equivalent to the current print_info and have a print_data method to print the whole data (as current __repr__)?

Second point can be addressed in a different PR though.

Note: not even sure the print_data would be necessary...

Completely agree with separating the notebooks, indeed.

added 1 commit

• 29dd8257 - Separate notebooks into two. Update internal links and re-run notebooks

Compare with previous version

Notebooks updated. Second point will be done in another PR

approved this merge request

merged

mentioned in commit f0b02cc8