Research Data Archive REST Listings

# x-method: derived
openapi: 3.1.0
info:
  title: University of Reading Repository REST Interface
  version: '3.4'
  summary: >-
    The read-only EPrints REST interface exposed on the University of Reading's two institutional
    repository hosts, CentAUR and the Research Data Archive.
  description: >-
    Machine-readable description of the /rest/ interface served by the University of Reading on its
    two repository hosts — centaur.reading.ac.uk (CentAUR, the institutional research-output
    repository) and researchdata.reading.ac.uk (the Research Data Archive). Both hosts are the
    university's own; the software serving them is EPrints 3.4.6, open-source repository software
    originally from the University of Southampton. The engineering of the interface is EPrints';
    the deployment, the host, the data and the operation are the University of Reading's.


    This document was NOT published by the University of Reading. It was derived by API Evangelist
    from live, unauthenticated probes on 2026-09-01. Each path below was fetched and its status code
    and response recorded. Nothing is described that was not observed.


    What is open and what is not, as measured:

    - GET /rest/ returned 200 on both hosts, listing the available datasets.

    - GET /rest/eprint/ returned 200 (6.5 MB on CentAUR, 47 KB on the Research Data Archive),
      enumerating every record id with links to a per-record directory and a per-record .xml.

    - GET /rest/subject/ and GET /rest/user/ returned 200 on both hosts.

    - GET /rest/eprint/{id}.xml returned **401 Auth Required** on both hosts. The listing is public;
      the record payload behind it is not.

    Callers wanting bulk machine-readable metadata without credentials should use the OAI-PMH
    endpoint at /cgi/oai2 instead, which is fully open and is described in its own document in this
    repository.


    One further measured constraint: the repository's own search and per-record export paths
    (/cgi/search, /cgi/export/...) sit behind an Anubis bot challenge and return an interstitial
    HTML page with HTTP 200 rather than the requested representation. /rest/ and /cgi/oai2 were not
    challenged.
  contact:
    name: CentAUR repository team, University of Reading
    email: centaur@reading.ac.uk
    url: https://centaur.reading.ac.uk/information.html
servers:
  - url: https://centaur.reading.ac.uk
    description: CentAUR — Central Archive at the University of Reading (research outputs)
  - url: https://researchdata.reading.ac.uk
    description: University of Reading Research Data Archive (research datasets)
tags:
  - name: Repository
    description: Read-only EPrints REST dataset listings served by the university's repositories.
paths:
  /rest/:
    get:
      tags: [ Repository ]
      operationId: listDatasets
      summary: List the datasets exposed by the repository
      description: >-
        Returns an XHTML index of the datasets the REST interface exposes. Observed values on both
        hosts: eprint, user, subject.
      responses:
        '200':
          description: XHTML index of datasets. Observed 2026-09-01 on both hosts.
          content:
            text/html:
              schema:
                type: string
                contentMediaType: text/html
  /rest/eprint/:
    get:
      tags: [ Repository ]
      operationId: listEprints
      summary: List every record id in the repository
      description: >-
        Returns an XHTML index of every eprint id, each linking to a per-record directory
        (`{id}/`) and a per-record XML document (`{id}.xml`). Large — 6.5 MB on CentAUR at probe
        time. The index itself is unauthenticated; the per-record XML it links to is not.
      responses:
        '200':
          description: XHTML index of record ids. Observed 2026-09-01 on both hosts.
          content:
            text/html:
              schema:
                type: string
                contentMediaType: text/html
  /rest/eprint/{eprintId}.xml:
    get:
      tags: [ Repository ]
      operationId: getEprint
      summary: Retrieve one record as EPrints XML (authenticated)
      description: >-
        Returns the EPrints XML representation of a single record. Measured to require
        authentication: an unauthenticated GET returned 401 with an "EPrints REST: Auth Required"
        body on both hosts on 2026-09-01. No public credential issuance process was found; the
        open route to this metadata is the OAI-PMH endpoint.
      parameters:
        - name: eprintId
          in: path
          required: true
          description: Numeric EPrints record id, as enumerated by GET /rest/eprint/.
          schema:
            type: integer
            examples: [ 5 ]
      responses:
        '200':
          description: EPrints XML record. Not observed — the endpoint requires credentials we do not hold.
          content:
            text/xml:
              schema:
                type: string
                contentMediaType: text/xml
        '401':
          description: Auth Required. This is what an unauthenticated caller receives, and what was observed.
          content:
            text/html:
              schema:
                type: string
                contentMediaType: text/html
  /rest/subject/:
    get:
      tags: [ Repository ]
      operationId: listSubjects
      summary: List the repository's subject taxonomy
      description: Returns an XHTML index of subject terms. Observed 200 on both hosts.
      responses:
        '200':
          description: XHTML index of subjects.
          content:
            text/html:
              schema:
                type: string
                contentMediaType: text/html
  /rest/user/:
    get:
      tags: [ Repository ]
      operationId: listUsers
      summary: List the repository's user records
      description: >-
        Returns an XHTML index of user record ids. Observed 200 on both hosts. Only ids are listed;
        the per-user documents behind them require authentication in the same way record XML does.
      responses:
        '200':
          description: XHTML index of user record ids.
          content:
            text/html:
              schema:
                type: string
                contentMediaType: text/html