Release notes

2.2.0 (2026-04-22)

The release candidate is now released as stable version.

Bugfixes

  • Fixed broken asset references.

Project maintenance

  • Upgraded dependencies to their latest security releases.

  • Application logs are now structured (JSON) using structlog.

2.2.0-rc.0 (2026-03-04)

GPP-publicatiebank 2.2.0-rc.0 is a maintenance release.

Upgrade procedure

No notes.

Bugfixes

  • [#109] Updated the branding (favicon + titles) with the GPP-Woo assets.

  • Fixed phone number validation for API credential contact details being too strict.

Project maintenance

  • Upgraded dependencies to their latest security releases.

  • Ignore npmjs links in the docs build, as they block bots now.

  • Upgraded deprecated Github Actions to their latest versions.

  • Upgraded dependencies to their latest versions, preparing for OTel integration.

  • [#115] Upgrade to the latest Celery version.

  • Documentation updates:

    • Updated the CVE link to the new CVE-website.

    • [#117] Fixed copy-paste leftovers from GPP-publicatiebank docs.

  • Refactored some internal tooling to use more shared-library utilities.

  • Removed optional and unused dependencies.

  • Simplified the frontend build system by using esbuild.

  • Optimized the memory usage of flower.

2.1.0 (2025-09-01)

GPP-zoeken 2.1.0 is a feature release.

Upgrade procedure

Warning

Manual intervention required.

The elastic search mappings are updated to store the publication date (gepubliceerd_op) of documents, publications and topics (which is a carbon copy of the registratiedatum as a convenience field) To make sure that this new property contains the registratiedatum of existing indexed documents either:

  • re-index them from the GPP-publicatiebank (version 2.0.0 or newer)

  • or run the migration script in the container:

    python /app/src/manage.py sync_publication_date

Features

  • [#100] Added more date fields to publication, document and topic search documents. All three documents gain the field gepubliceerd_op and the publication document additionally gains the fields datum_begin_geldigheid and datum_einde_geldigheid. You can now also query/filter on these fields in the search endpoint.

  • [#101] Chronological ordering of search results is now done on the gepubliceerd_op field

    rather than laatst_gewijzigd_datum.

2.1.0-rc.0 (2025-07-16)

Upgrade procedure

Warning

Manual intervention required.

The elastic search mappings are updated to store the identifiers (“kenmerken”) of documents and publications together. To make sure that this new property contains the identifier of existing indexed documents, either:

  • re-index them from the GPP-publicatiebank

  • or run the migration script in the container:

    python /app/src/manage.py sync_identifiers

Deprecations

  • [#31] The identifier field is now deprecated, use the identifiers array instead.

Bugfixes

  • [#81] Ensure that zip files with contents that exceed the filesize limit still get partially indexed, up to the total file size limit.

Project maintenance

  • Replaced boilerplate utilities with their equivalents from maykin-common.

  • Updated “supported versions” documentation section.

  • Upgraded some dependencies to their latest available versions.

2.0.0 (2025-07-10)

Definitive stable release.

The release candidate did not reveal any issues, so we turned it into a stable release.

Upgrade procedure

Warning

Manual intervention required.

After deploying this version, the index mappings in Elastic Search need to be updated to support topics. We don’t have an automated procedure for this yet. The easiest way to do this, is by deleting and re-creating the indices, and then re-index the data from GPP-publicatiebank (version 1.1.0 or newer).

DELETE https://my-elastic.example.com/document HTTP/1.1

DELETE https://my-elastic.example.com/publication HTTP/1.1

where document and publication are the names of the indices we manage.

Then, open an interactive shell for GPP-zoeken (with kubectl exec or docker exec), and run:

python src/manage.py initialize_mappings --wait

You should then get output confirming the indices have been re-created.

Breaking changes

  • Dropped PostgreSQL 13 support (our underlying framework doesn’t support it anymore).

  • The index mappings need to be dropped and re-created, see the upgrade procedure above.

Features

  • [#76, #43] Added “Topics” as resource type to group multiple publications together.

  • [#63] Added support for indexing ZIP (.zip and .7z) files. The archives are now extracted and the content of the extracted files is indexed and searchable.

Project maintenance

  • Switched code quality tools to Ruff.

  • Simplified documentation test tools.

  • Added upgrade-check mechanism for “hard stops”.

  • Upgraded framework version to next LTS release.

  • Addressed API schema linter error for URL-value defaults.

2.0.0-rc.0 (2025-05-19)

Upgrade procedure

Warning

Manual intervention required.

After deploying this version, the index mappings in Elastic Search need to be updated to support topics. We don’t have an automated procedure for this yet. The easiest way to do this, is by deleting and re-creating the indices, and then re-index the data from GPP-publicatiebank (version 1.1.0 or newer).

DELETE https://my-elastic.example.com/document HTTP/1.1

DELETE https://my-elastic.example.com/publication HTTP/1.1

where document and publication are the names of the indices we manage.

Then, open an interactive shell for GPP-zoeken (with kubectl exec or docker exec), and run:

python src/manage.py initialize_mappings --wait

You should then get output confirming the indices have been re-created.

Breaking changes

  • Dropped PostgreSQL 13 support (our underlying framework doesn’t support it anymore).

  • The index mappings need to be dropped and re-created, see the upgrade procedure above.

Features

  • [#76, #43] Added “Topics” as resource type to group multiple publications together.

  • [#63] Added support for indexing ZIP (.zip and .7z) files. The archives are now extracted and the content of the extracted files is indexed and searchable.

Project maintenance

  • Switched code quality tools to Ruff.

  • Simplified documentation test tools.

  • Added upgrade-check mechanism for “hard stops”.

  • Upgraded framework version to next LTS release.

1.0.0 (2025-04-16)

The release candidate is now released as stable version.

There are no changes compared to release candidate 1 - see the changelog entry below.

1.0.0-rc.0 (2025-03-26)

We proudly announce the first release candidate of GPP-zoeken!

The 1.0 version of this component is ready for production. It provides the necessary functionality to provide your organisation-specific search index of public documents.

Features

  • Admin panel for technical and functional administrators

    • Manage API clients and user accounts.

    • Configure connections to external services, like the GPP-publicatiebank and OpenID Connect provider.

  • JSON API for indexing, deleting and searching publications and documents to/from Elasticsearch, with the ability to search through the text contents of files.

  • OpenID Connect or local user account with MFA authentication options for the admin panel.

  • Extensive documentation, from API specification to (admin) user manual.