borsna / daget Goto Github PK

Should output relevant debug info and http-response codes

https://datadryad.org/search

example dataset:
https://doi.org/10.5061/dryad.5d23f

Related to #12
fix the workflow for publishing directly to pypi on release
https://github.com/borsna/daget/blob/main/.github/workflows/python-publish.yml

Via option or by omitting the destination folder?

Feature request from @David-Rayner-GVC is issue #4 (comment)

Cleanup script to expose public functions to get filelist as a json dict

The redirect from DOI to dataverse landing page seems a bit od.
The json-ld output from dataverse does contain the file listing for the dataset.

Common data repository software installed on multiple domains e.g:

• https://data.gov
• https://opendata.swiss/en

No information about software in http-header on landing pages but in the html header there is a tag for indication ckan:

<meta name="generator" content="ckan 2.10.1">

Two problems with suporting zenodo:

1. Zenodo does not suport json-ld via content negotiation.
2. Unfortunatly the json-ld representation does not contain data distribution links to the files

Solution:

• Add sppecial case for zenodo to get the file list in zenodos native json-format via the api e.g. https://zenodo.org/api/records/8307895

Skip files with same name and size when downloading.
Add override option.

Add option to get metadata.
For repositories delivering schema.org this json-repsentation should be stored.

Additional exports of metadata should be fetched if possible form (not all DOI:s are registerd via DataCite):
https://support.datacite.org/docs/datacite-content-resolver#supported-content-types

Example on the repository SITES:

• https://meta.fieldsites.se/objects/Mhz3oXqA-1Tzcf-C0mfbgTKe

Not sure how to get file url and name via metadata call to figshare.

Posible solution

Example landing page:

https://figshare.scilifelab.se/articles/dataset/Metagenomic_dataset_from_Swedish_urban_lakes/22270225/2

API:

https://api.figshare.com/v2/articles/22270225/versions/2

e.g. giving 10.48723/w728-p041 should internally be translated to https://doi.org/10.48723/w728-p041

https://osf.io

use only colored text and standard characters in the output (emojis is not sported in all terminals)

Resructure repo and add support for distribution and installation via pip.