Skip to main content

For researchers

API & harvesting

The published corpus is openly available for programmatic access. No key, no registration, no rate tier — only published, editorially reviewed sources are exposed. Every record carries its attribution, and that attribution propagates into citations, the API, and OAI-PMH alike.

Read API

JSON over HTTPS. Responses are paginated with limit and offset, and carry a total. Follow any path below to see a live response.

GET/api/public/sources
q, source_type, language, sort, limit, offset
List published sources, newest first.
GET/api/public/sources/{slug}One source, with full provenance, rights and attribution.
GET/api/public/facetsCounts by source type and language, for building filters.
GET/api/public/export
format=jsonl (default) or csv
The entire published catalogue as one dated file. Free, no key. Descriptive metadata only.
GET/api/search
q, source_type, language, limit, offset
Keyword search across titles, descriptions and attribution.

OAI-PMH

Standard metadata harvesting (OAI-PMH 2.0, Dublin Core) for libraries and aggregators. Base URL: https://api.julyarchive.org/api/oai. Withdrawn records are disseminated as deleted-status tombstones, so a harvester that already holds a record learns when it is taken down.

GET?verb=IdentifyRepository name, protocol version, earliest datestamp.
GET?verb=ListMetadataFormatsSupported formats. Dublin Core (oai_dc).
GET?verb=ListRecords
metadataPrefix=oai_dc, from, until, resumptionToken
Full records, 100 per page, with a resumptionToken.
GET?verb=ListIdentifiers
metadataPrefix=oai_dc, from, until
Headers only — cheaper for incremental harvesting.
GET?verb=GetRecord
identifier, metadataPrefix=oai_dc
A single record by OAI identifier.

Persistent identifiers

Records carry an ARK identifier and resolve at /ark:/<naan>/<id>, so a citation keeps working even if a URL changes. Each source page also offers a “Cite this” widget in Chicago, MLA and APA.

Terms

The archive holds preservation copies; rights in the underlying material remain with the original publishers and creators, and each record states its own rights. Reuse with due credit — reproduce the attribution the record carries. Working at volume, mirroring the corpus, or building on the archive's own metadata, translations or timeline? See institutional access & licensing.

Ask