ianco / anoncreds-w3c-mapping

Python scripts for mapping between Anoncreds presentations and W3C credentials

AnonCreds to W3C Format Verifiable Credential and Presentation Converter

This repository contains prototype Python scripts that:

• Convert AnonCreds Verifiable Credentials into compliant W3C Format Verifiable Credentials and back, and
• Convert AnonCreds Presentations into compliant W3C Format Verifiable Presentations and back.

These scripts are preliminary work in a possible transition of using the W3C Verifiable Credential Data Model Standard for AnonCreds verifiable credentials and presentations. As demonstrated by the back and forth nature of the conversion and the absence of any cryptography library dependencies, nothing is being done to change the AnonCreds signatures or to alter the semantics of the AnonCreds credentials and presentations. The same AnonCreds code that is currently used in issuing, holding, proving and verifying AnonCreds continues to be used. The only change is the representation (arrangement of the JSON) within the credential and presentation.

To run the generator with your own examples:

• Place your AnonCreds credentials JSON in the credentials folder, and your AnonCreds presentations in the presentations folder.
• Run the bash script ./generateAll
• Look in the W3COutput folder to see all of the converted credentials and presentations.

To run the scripts directly, use:

• python credential_to_w3c.py <input file> [> output.json]
• python presentation_to_w3c_vp.py <input file> [> output.json]

Notes

• Attributes are encoded as name-value pairs. This bypasses the need to define a custom JSON-LD contexts for each credential schema, as without changing AnonCreds, there is not a way for the Issuer to say where the context can be found.
• The converter is not including the encoded values of the AnonCreds verifiable credential attributes in the W3C formats, using instead a flag "encoding": "auto",, and converting raw to encoded data on the fly. The encoder matches Aries RFC 0592 Indy Attachments section on Encoding Claims. We think that the code for generating the encoded values should be moved into AnonCreds and out of the hands of the Issuer.

Feedback, issues and pull requests welcome!