Skip to content

data dictionary - dcat-us:describedBy #220

Open

Description

@jqnatividad

The data dictionary should be an extended JSON Schema file, using additionalProperties to capture non-standard JSON Schema elements.

dathere/qsv#2373

Since there is no widely deployed standard for Data Dictionaries, we should create our own.

However, we should still hew closely to related standards like DDI to ensure wider adoption.

Activity

  1. jqnatividad commented on May 19, 2026

    @jqnatividad
    CollaboratorAuthor

    This will wait until describegpt has been updated to save Data Dictionaries in jsonschema format.

  2. jqnatividad commented on May 19, 2026

    @jqnatividad
    CollaboratorAuthor

    Deferring this work until qsv issue #2373 lands — i.e., until qsv describegpt produces JSON Schema output natively.

    Rationale: building DP+'s data-dictionary-as-extended-JSON-Schema on top of describegpt's current output shape would mean writing (and then throwing away) a conversion layer once qsv emits JSON Schema directly. The cleaner sequence is:

    1. qsv#2373 lands → describegpt returns JSON Schema with additionalProperties for non-standard fields
    2. DP+ AI suggestions stage consumes that JSON Schema directly (jobs/stages/ai_suggestions.py — currently parses the Dictionary/Description/Tags PascalCase envelope per fix(ai-suggestions): real-qsv schema + CLI flag corrections (PR #301/#302 follow-up) #303)
    3. Persist the Data Dictionary as the extended JSON Schema, surface it via dcat-us:describedBy
    4. Wire DDI alignment where overlapping

    Keeping this issue open and unblocked-pending-upstream. Will revisit when qsv ships the schema output.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions