README

Scope

This library facilitates the retrieval of structured metadata based on a collection of given DOIs. It resolves DOIs using content negotiation taking into account different standards used by registration agencies. It provides methods to analyse this metadata and extract ORCIDs which can then be resolved, too.

Use Cases

The use cases range from very simple to more complex ones. Initially, this library was designed to resolve DOIs to structured metadata to obtain ORCIDs for a given publication. Then more functionality was added to extract DOIs from ORCID profiles to continue the process. This means that given some DOIs as a starting point, this library can be used like a crawler following the connection between DOIs and ORCIDs. From a few DOIs, the co-author network can be constructed by combining DOI and ORCID metadata, using DOIs and ORCIDs as identifiers.

Structure

The library consists of three modules:

1. doi_ra_handler: Given a collection of DOIs, groups them by registration agency based on the DOI prefix.
2. pid_resolver: Given a collection of DOIs for a known registration agency, resolves them to structured metadata. The serialisation format and data model depends on the registration agency. Resolved DOI metadata will be cached in the corresponding directory.
3. pid_analyzer: Given DOI metadata, provides methods to analyse this data and build a general structure called PublicationInfo representing basic information such as title and author information including ORCID for a given DOI.

Caching

All resolved DOIs and ORCIDs are cached. For each registration agency (RA), a separate cache directory is used. Cache directories are created in the root of the project this lib is used in.

Licensing

This library is licensed under the terms defined in LICENSE. Software dependencies are explicitly mentioned in the dependencies document.

Usage

The library offers two CLI scripts that can be used as follows:

• Create a virtual environment, see https://docs.python.org/3/tutorial/venv.html#creating-virtual-environments
• Install the library pip install -e <path/to/local/repo> --config-settings editable_mode=compat (from locally checked out repo, since this lib has not been published yet).

Resolve DOIS

• Create a JSON file containing one or several DOIs, e.g., a file dois.json with the contents ["10.1007/978-3-031-47243-5_6"]. Note that DOIs are without base path https://doi.org/.
• Use the script as follows: pid_resolver_resolve -i 2 -d dois.json (resolve DOIs from JSON file and perform two iterations).
• Run pid_resolver_resolve for usage instructions.

The process will start with the given DOIs and perform as many iterations as configured. An iteration consists of resolving the given DOIs as well as resolving the linked ORCID profiles. The results of the analysis will be written to results.json (working directory). The cache directories will be created in the working directory.
The DOIs extracted from the ORCID profiles will be resolved in the next iteration.

Infer missing ORCIDs

• Run the resolving process as described above with a set of DOIs.
• The structure in results.json may still contain authors without ORCIDs as the information may not be present in the DOI metadata or the corresponding ORCID profile does not mention the publication. Still, ORCIDs may be inferred for an author from a different publication if several publications share common co-authors identified by an ORCID.
• Run pid_resolver_infer to infer missing ORCIDs. The results will be written to updated.json.