Export nomination data to JSON

Motivation

Currently the process of generating the nomination report is time-consuming and error-prone. The combination of this means there is only really scope to generate one form of the report in a pdf. Ideally the report would be created in multiple formats, including a machine-readable format that others can use to analyse the results and create alternative presentations.

If nomnom exported the data it has in a machine readable format, this can be supplemented with additional data by the Hugo team and fed into scripts that can generate the pdf report, as well as other formats.

The raw data out of nomnom would not be expected to be publishable. Instead, it would need to be cleaned up (e.g. nominee names suitable for publishing added) before it would be published. The Hugo admin team would be responsible for arranging how and where it would be published.

Proposed format

I have been working on the JSON schema for a JSON format that would be suitable for the final publication. I propose nomnom generates a subset of this format, that can then be supplemented with additional data by the Hugo team.

https://ajanuary.github.io/hugo-nomination-stats-schema/nominations-2025-10/

Where the nomnom export format would differ:

• Nominee name would be the internal name used for tracking deduplication, rather than the name on the published ballot.
• The finalist and removed properties on the nominees would never be provided.

Start by reviewing the linked nominations-2025-10 schema and tracing the existing nomination report-generation flow. Define the export as the schema subset described in the issue, using internal nominee names and omitting finalist and removed properties. Done means nomnom produces machine-readable nomination data that the Hugo team can supplement.