Determine ORCID link for authors of zenodo-records

[haesleinhuepf]

Does the Zenodo REST API allow determining ORCID links from authors?

git-bob answer

[github-actions]

^{This message was generated by git-bob (version: 0.27.2, model: github_models:gpt-4o, log), an experimental AI-based assistant. It can make mistakes and has limitations. Check its messages carefully.}

I don’t know. Tagging @haesleinhuepf for clarification on whether the Zenodo REST API supports retrieving ORCID links for authors.

[haesleinhuepf]

When accessing the zenodo REST API, I author find entries like this:

"creators": [{"name": "Haase, Robert", "affiliation": "Center for Scalable Data Analytics and Artificial Intelligence", "orcid": "0000-0001-5949-2327"}],

If contains an "orcid", which I believe is an optional parameters.

Hence, we need a python script, that goes through our resources/nfdi4bioimage.yml file and detects all records with a zenodo url. Our yml file looks like this:

resources:
- author: 
  - person A
  - person C
  author_with_orcid:
  - person A
  - person C https://orgid.org/0000-0001-5949-2327
   url:
  - https://zenodo.org/record/4071471
  - https://doi.org/10.5281/zenodo.4071471
- author: person B
   url:
  - https://zenodo.org/record/4071472
```

I mean we should complete records in the yml file. If they don't have the `author_with_orcid` entry, but any url starting with 'https://zenodo.org/', we use the zenodo REST API to read out the author names and corresponding orcids and save all authors (also those without orcid) to the `author_with_orcid` field.

git-bob comment on how to achieve this

[github-actions]

^{This message was generated by git-bob (version: 0.27.2, model: github_models:gpt-4o, log), an experimental AI-based assistant. It can make mistakes and has limitations. Check its messages carefully.}

To achieve this, you can write a Python script that iterates over the YAML file, checks for Zenodo URLs in the url field, fetches author details with ORCIDs using the Zenodo REST API, and updates the author_with_orcid field. Below is a proposal for such a script:

import yaml
import requests

# Function to fetch author details from Zenodo REST API
def fetch_authors_from_zenodo(record_url):
    zenodo_api_url = record_url.replace("https://zenodo.org/record/", "https://zenodo.org/api/records/")
    response = requests.get(zenodo_api_url)
    response.raise_for_status()
    data = response.json()
    authors = data.get("creators", [])
    author_with_orcid = []
    for author in authors:
        if "orcid" in author:
            author_with_orcid.append(f"{author['name']} https://orgid.org/{author['orcid']}")
        else:
            author_with_orcid.append(author['name'])
    return author_with_orcid

# Load YAML file
with open("resources/nfdi4bioimage.yml", "r") as file:
    data = yaml.safe_load(file)

for resource in data.get("resources", []):
    if "author_with_orcid" not in resource:
        for url in resource.get("url", []):
            if url.startswith("https://zenodo.org/record/"):
                try:
                    author_with_orcid = fetch_authors_from_zenodo(url)
                    resource["author_with_orcid"] = author_with_orcid
                    break  # Only one valid Zenodo URL is needed
                except Exception as e:
                    print(f"Error fetching authors for URL {url}: {e}")

# Save updated YAML
with open("resources/nfdi4bioimage.yml", "w") as file:
    yaml.dump(data, file, sort_keys=False)

This script reads the YAML file, accesses Zenodo's REST API for each Zenodo URL, retrieves the authors and their ORCIDs, and updates the author_with_orcid field accordingly.

[haesleinhuepf]

One minor aspect: resource.get("url", []) may return a single string or a list of strings. Ensure to convert it to a list if it is a string.

Also, instead of reading the yml file from disc and saving to disc, use this for reading it:

    # Authenticate and fetch file content
    repo = get_github_repository(repository)
    main_branch = "main"
    file_content = repo.get_contents(file_path, ref=main_branch)
    content = yaml.safe_load(file_content.decoded_content.decode())

And this for writing (creating a pr):

    if modified:
        new_branch = create_branch(repository, parent_branch=main_branch)
        new_content = yaml.dump(content, sort_keys=False, allow_unicode=True)
        write_file(repository, new_branch, file_path, new_content, "Add missing DOIs to resources")
        pr_url = send_pull_request(
            repository,
            new_branch,
            "Add missing DOIs for Zenodo records",
            f"closes #{899}\n\nThis PR adds missing DOI links to the resources YAML file."
        )
        print(pr_url)

git-bob implement this as a new scripts/add_author_orcids.py file. The file should also contain a if main block that calls the functions above.