ckanext-actor-registry 0.3.0__tar.gz

ckanext_actor_registry-0.3.0/AI_ASSISTANCE.md

@@ -0,0 +1,19 @@
+# AI assistance disclosure
+
+The implementation and documentation of ckanext-actor-registry were developed by
+Björn Hagström with assistance from OpenAI's ChatGPT and Codex and, from version
+0.2.0, from Anthropic's Claude Code.
+
+AI tools assisted with research, design discussion, code drafting, refactoring,
+documentation and operational verification. Product decisions, deployment
+authorization and acceptance of the resulting work remain human decisions.
+
+The human maintainer is the author, copyright holder and accountable project
+maintainer. AI systems are not listed as authors or maintainers. Before a public
+release, the maintainer is expected to review the released diff, verify the
+documented behavior and accept responsibility for correctness, security,
+licensing and provenance.
+
+Contributors should disclose material generative-AI assistance in their pull
+request description. Disclosure does not replace review, tests or the
+contributor's responsibility for the submitted material.

ckanext_actor_registry-0.3.0/CHANGELOG.md

@@ -0,0 +1,341 @@
+# Changelog
+
+All notable changes to this project will be documented in this file. The format
+is based on Keep a Changelog, and the project follows Semantic Versioning.
+
+## [Unreleased]
+
+## [0.3.0] - 2026-10-07
+
+### Upgrading from 0.2.x
+
+- Run `ckan db upgrade -p actor_registry` (new tables, migrations 007 and 008).
+- **List `actor_registry` before `scheming_datasets` in `ckan.plugins`.** 0.2.x
+  documented the opposite order. CKAN takes a template from the plugin listed
+  first, and the new `additional_info.html` override only applies when
+  `actor_registry` comes first. In the old order the dataset page keeps showing
+  Maintainer and Author while the DCAT export leaves them out, so the page and the
+  export disagree.
+
+### Added
+
+- A shown Author is exported as `dct:creator` (DCAT-AP-SE: *producent*), so
+  the export says what the dataset page says.
+- *Cleanup* page (`/actor-registry/cleanup`): inactive publishers and contact
+  points still used by datasets, and references to records that no longer
+  exist; replace or remove them on the datasets you may edit.
+- "Migrate to Actor Registry" (`/actor-registry/migrate`, a new tab): turns the
+  contact details in CKAN's Maintainer and Author fields into shared contact
+  points. Groups by email address, shows name variants and matching contact
+  points; create, link to existing or skip, in rounds. Original fields are
+  never changed; editors migrate their own datasets. Decisions are logged in
+  a new table (migration `actor_registry_007`).
+- The migration page shows a group of more than 25 datasets compactly: all of them are included, and
+  the datasets to leave out are ticked in a scrollable list. Only those are posted, so a group of
+  thousands of datasets is not refused by the web framework's form-size limit.
+- Migration: *Changed since linking* lists linked datasets whose Maintainer or
+  Author changed afterwards, with old and new values; update the contact
+  point, link another, create a new one or keep it (migration
+  `actor_registry_008` stores the values a decision was made on).
+- Migration: *Close the migration* (sysadmins) hides Maintainer and Author in
+  forms and on dataset pages and stops exporting them, at runtime and without
+  deleting anything; it can be reopened. Asks for confirmation while datasets
+  still lack a contact point. New `legacy_contact.html` form and display
+  snippets for those fields.
+- Setting `ckanext.actor_registry.legacy_contact_fields` (default `true`):
+  when `false`, the DCAT export never uses CKAN's free-text contact fields
+  (Maintainer, Author, ckanext-dcat's `contact_*` extras). Stored values are
+  not changed.
+- `docs/CATALOGUE_SETUPS.md`: common ways catalogues relate organizations,
+  publishers and contact points, with a recommendation for each.
+- A second example schema, `contact_points_only_schema.yaml`: organization and
+  contact points without a publisher field, for catalogues where the owner
+  organization is the publisher. With tests, including that the DCAT export
+  then uses the organization as publisher.
+- README section "Choosing your fields": why and how to choose between
+  organization, publisher and contact points, how CKAN's own Author/Maintainer
+  fields map, and that the schema decides which fields a dataset has (the
+  optional legacy contact snippets can hide Maintainer and Author).
+
+### Changed
+
+- What visitors see is what is exported: the dataset page now shows an
+  inactive publisher or contact point (the export already contained it), the
+  export follows a merged publisher to its target (as the page did), and
+  Maintainer/Author are not shown when the dataset has registry contact
+  points (they were not exported). README: "What visitors see is what is
+  exported", with the deliberate exceptions.
+- The publisher is optional in `actor_registry_schema.yaml`. When it is
+  optional and empty, the form explains that the organization is given as
+  publisher in the DCAT export; the dataset page shows no publisher.
+- The "missing" prompt on a dataset page follows the schema: it asks for a
+  publisher only when the schema requires one, never for a field the schema
+  lacks, and its heading names what is missing.
+- The Edit and Delete buttons in the publisher and contact point lists are
+  spaced, kept on one line, aligned with the row and identical in both lists;
+  Delete is outlined instead of filled. Identifiers no longer wrap.
+- The publisher and contact point pages are tidier: empty fields are no longer
+  shown as "–", the dataset counts use real plural forms ("Linked to 1
+  dataset", "Linked to 4 datasets", also on the delete and merge pages), each
+  linked dataset shows its organization and when it was last updated, and the
+  contact point breadcrumb names the contact point. The pages use the full
+  width.
+
+### Fixed
+
+- Maintainer and Author that the schema does not show on the dataset page are
+  no longer exported (ckanext-dcat used them as a contact point anyway).
+- A deactivated publisher no longer locks the datasets that use it, the same
+  way as for contact points; the form shows it marked inactive, and choosing
+  an inactive or missing publisher gives a message that names it. Inactive
+  publishers are no longer offered as "Belongs to organization" for a new
+  contact point.
+- A deactivated contact point no longer locks the datasets that use it: they
+  can be saved, and the form shows it, marked inactive, instead of dropping it
+  silently. Choosing an inactive contact point anew, or one that no longer
+  exists, gives a message that names it.
+- `ckanext.actor_registry.legacy_contact_fields` set to false in the ini file
+  is read correctly after CKAN reloads its runtime settings (it can then be an
+  empty string).
+- A dataset with a broken stored `contact_point_ids` value (invalid JSON, as old
+  imports can have) no longer makes the Publishers page and the worklist fail;
+  such a value counts as no contact point, and a bare id as one.
+- The DCAT export no longer adds Maintainer or Author as a second contact point
+  when a dataset has registry contact points.
+- Contact point and publisher references stored as JSON text are resolved like
+  lists.
+- Visitors who are not logged in no longer see the administration tabs
+  (Sysadmins, Config, Trash) or the registry tabs on a publisher's or contact
+  point's public page; logged-in users see only the registry tabs, sysadmins
+  see all.
+
+## [0.2.3] - 2026-10-01
+
+### Changed
+
+- The phone help text no longer refers to DCAT-AP-SE: "Include the country code, e.g. +46
+  or +49, so the number works from other countries too." The README and the profile
+  describe the telephone export as the usual vCard form, which ckanext-dcat itself lacks.
+- The documentation speaks of DCAT-AP and national profiles based on it, with the Swedish
+  DCAT-AP-SE only as an example of a national validator.
+- A new language needs only its catalogue: the plugin offers every locale with a compiled
+  `.mo` file in `i18n/` instead of a fixed list. A contributed language may be partial
+  (untranslated strings show in English); only the maintained Swedish catalogues must be
+  complete. A test checks that every `.mo` matches its `.po`, and CONTRIBUTING.md says how
+  to contribute a language.
+
+## [0.2.2] - 2026-10-01
+
+### Fixed
+
+- Swedish translations for the 13 strings added with contact point deletion
+  (confirmation page, list filter, messages). They were missing from the
+  `.pot` and `.po` files and showed in English on Swedish pages.
+- New test: every string a fresh extraction finds must be translated in every
+  shipped catalogue.
+
+## [0.2.1] - 2026-09-22
+
+### Fixed
+
+- `examples/ckan-2.12/smoke_check.py`: the resource-creation step's POST response was
+  discarded, so a server error there could still end in "All checks passed." The
+  underlying cause was in the script itself: its hidden-field parser assumed
+  `type="hidden"` came first in the tag, which is not true of CKAN 2.12's resource
+  form, so the required (and CKAN-mandated) `id` field was silently dropped from the
+  submission, causing a real HTTP 500 (`KeyError: 'id'` in CKAN core). The parser is
+  now attribute-order independent, and the step's response is checked like every
+  other step.
+- `examples/ckan-2.12/smoke_check.py`: `CKAN_SITE_URL` is now read from the process
+  environment first, falling back to `.env`, so a one-off `CKAN_SITE_URL=... python3
+  smoke_check.py` (for a non-default port) is honoured instead of silently testing
+  the wrong server.
+
+### Added
+
+- a regression test confirming that migration `006` fails clearly, and does not drop
+  data, when a database already has active publishers with a duplicate
+  (identifier_scheme, identifier) pair -- the behaviour the migration's own docstring
+  and this changelog already documented, now exercised.
+
+## [0.2.0] - 2026-09-21
+
+The first tagged release. It follows the initial source-only snapshot (0.1.0, below) and contains
+breaking changes: translation keys changed with the UI text ("actor" is now "publisher"), migration
+`006` adds a unique index, and CKAN 2.11.0 and older are not supported.
+
+### Added
+
+- an initial automated test suite for registry behavior, dataset projection and
+  DCAT-AP 3 RDF serialization;
+- a migration rehearsal that preserves records from the unpublished contact
+  point prototype;
+- Action API end-to-end tests for Turtle, RDF/XML and JSON-LD plus optional
+  validation with the official DCAT-AP 3 SHACL shapes;
+- documented RDF and interoperability design decisions;
+- cross-registry validation preventing actors and contact points from sharing a
+  URI.
+- a GitHub Actions CI workflow running the automated test suite and the
+  dependency-free JS regression tests on every push and pull request;
+- a README section documenting how to activate the extension on an existing
+  installation that already has datasets (the `/actors` worklist and the
+  inline per-dataset fix-it prompt);
+- `setup.py` project metadata (`url`, `author`, `project_urls`) now that the
+  project has a public repository.
+- permanent actor deletion (`actor_registry_actor_delete`, sysadmin-only),
+  for actors -- active or already merged/retired -- that are no longer
+  needed. Clears the actor as publisher on any dataset that had it, and
+  decouples (does not delete) any contact points it owned;
+- merged/retired actors are now hidden by default on `/actors`, with a
+  checkbox to show them again.
+- permanent contact point deletion (`actor_registry_contactpoint_delete`,
+  sysadmin-only), mirroring actor deletion: removes the contact point
+  from any dataset's `contact_point_ids`, leaves an owning actor
+  unaffected;
+- inactive contact points are now hidden by default on `/contactpoints`,
+  with a checkbox to show them again.
+- public dataset display snippets (`actor.html`, `contactpoints.html`) showing
+  readable publisher and contact point details instead of stored ids; merged
+  actors resolve to their survivor, inactive or missing references render
+  nothing;
+- `(identifier_scheme, identifier)` uniqueness among active actors, with
+  server-side validation in forms and inline dialogs, a localised conflict
+  message naming the existing actor, and a partial unique index (migration
+  `006`; upgrading fails if duplicates already exist);
+- email, URL, phone and kind validation/normalisation in every form and dialog. Emails are
+  checked with CKAN's own `email_validator` (plus a dot in the domain, which CKAN itself does
+  not require); a value that is unchanged from what is stored is not re-validated, so records
+  saved before a rule existed can still be edited;
+- opt-in validators `actor_registry_actor_exists` and
+  `actor_registry_contactpoints_exist` for dataset schemas;
+- localised field help texts shared by the admin forms and inline dialogs,
+  a `sv_SE` catalogue, and a localised example schema
+  (`ckanext/actor_registry/examples/`);
+- `docs/INSTALLATION.md`, an install and compatibility guide for CKAN 2.11 and 2.12;
+- `docs/TRANSLATIONS.md`, a guide to the translation catalogues and how to change or add strings
+  (it replaces the internal `I18N_PLAN.md`);
+- `examples/ckan-2.12/`, a runnable Docker Compose example (CKAN 2.11 or 2.12, PostgreSQL,
+  Solr, Redis, ckanext-scheming 3.1.0 and this extension) with a smoke test;
+- supported CKAN releases: 2.11.1 and later 2.11 patches, and 2.12. CI runs the full suite on
+  2.11.1 (the oldest supported), the newest 2.11 patch and 2.12 with ckanext-scheming 3.1.0,
+  and on 2.12 against scheming `master` (allowed to fail); `Dockerfile.test` and
+  `docker-compose.test.yml` take the CKAN and scheming versions as build arguments.
+- Swedish translations for the delete-confirmation pages (they were English-only);
+- "Belongs to organization" can now be set in the inline contact point dialog
+  (previously only in the admin form), with the same help text; a link must
+  point at an existing, active actor.
+- a contact point's page lists the datasets it is linked to, and an actor's page
+  previews its first 10 (as publisher, and via its contact points) with a
+  "Show all N" link to a paginated list (20 per page, A-Z) at
+  `/actors/<id>/datasets` and `/contactpoints/<id>/datasets`. The lists are read
+  from the database, not the search index, and respect dataset visibility:
+  private datasets are shown only to sysadmins and members of the owning
+  organisation.
+
+### Changed
+
+- documentation brought up to date for 0.2.0: the README is rewritten for the "publisher"
+  wording, with a Documentation index and a note on each screenshot link (the screenshots are
+  being retaken); `docs/BACKLOG.md` is now a short English backlog; CONTRIBUTING, SECURITY,
+  TESTING, the release notes, the community announcement draft and the catalog submission
+  draft are updated;
+- validation now lives in ONE shared layer (`validation.py`) instead of in the views: the rules are
+  CKAN validators collected in schemas that CKAN's own `navl` machinery runs, giving ordinary CKAN
+  error dicts. The full forms, the inline dialogs and the merge action all use it, and an API added
+  later would too;
+- telephone URI normalization is shared by HTML helpers and RDF serialization;
+- controlled theme and publisher-type URIs are explicitly typed as
+  `skos:Concept` in RDF output;
+- `CONTRIBUTING.md`'s development priorities no longer list a full Action API;
+  see `docs/BACKLOG.md` for why that was deliberately deprioritized rather
+  than deferred.
+- the worklist of datasets that lack something, on `/actors`, previews 10 with an exact total and a
+  "Show all" link to a paginated page (`/actors/unlinked-datasets`). A *Missing:* choice picks
+  which datasets it lists: missing both a publisher and a contact point (the default), only a
+  publisher, only a contact point, or either (`?missing=both|publisher|contact_point|either`).
+  Like the other registry lists it reads from the database, not the search index;
+- the contact point list links each name to the contact point's page, like the publisher list;
+- a publisher is no longer presented as a contact point: its page, its "show all"
+  page, the delete confirmation and the merge summary list only the datasets it
+  publishes, not datasets that merely use a contact point that belongs to it;
+- the contact point field "Belongs to actor" is now "Belongs to organization"
+  (sv: "Tillhör organisation"), and the row is hidden on the public contact
+  point page when no link is set;
+- several contact points on a dataset page are visually separated.
+- on the dataset page a publisher's type, web address, email and description are
+  shown on their own lines under its name (values only, like the contact points)
+  instead of the web address in parentheses;
+- UI text now says "publisher" instead of "actor" throughout, in English and
+  Swedish ("utgivare", not "aktör"), matching the dataset form. Only visible
+  text changes; URLs, action names and code identifiers keep "actor". The
+  translation catalogue keys change with the English text, so out-of-tree
+  translations of the old strings need updating;
+
+### Fixed
+
+- `actor_registry_actor_merge` wrote the chosen field values (email, URL, identifier, ...) onto the
+  surviving publisher without validating them, and an identifier that clashed with a third
+  publisher's surfaced as a database error. The chosen values now pass the same rules as any other
+  edit, and a violation is a `ValidationError` on `fields`, with nothing changed;
+- an input longer than its database column (a name or identifier over 255 characters, an email
+  over 320, a phone number over 100) crashed with a database error, i.e. an HTTP 500. Every
+  field now has a maximum length and shows a translated message instead;
+- the Persistent URI and Type fields of a publisher, and the URI of a contact point, were not
+  validated on the server; they must now be absolute URIs (`http://`, `https://` or `urn:`);
+- stored web addresses and URIs are only turned into links when CKAN's own `h.is_url` accepts
+  them (also in the JavaScript previews, which allow only http(s)); anything else, e.g. a `javascript:` value from a
+  row saved before validation existed, is shown as plain text instead of a clickable link;
+- the queries that find the datasets linked to a publisher or contact point read the
+  dataset's extras from `package.extras`, a jsonb column that only exists in CKAN 2.12;
+  on CKAN 2.11 (where extras live in the `package_extra` table) they failed with
+  `column "extras" does not exist`. Both layouts are now supported;
+- the test image no longer needs `wget` (it is gone from the CKAN 2.12 images) and no
+  longer depends on the host's file modes.
+- merging actors would have failed to hand the surviving actor the merged
+  actor's identifier once identifiers became unique; the merged actor is now
+  retired first;
+- `MANIFEST.in` now packages the translation catalogues and the example schema.
+
+### Known limitations
+
+- the dataset lists on publisher and contact point pages only include private datasets
+  that the visitor may see as a sysadmin or as a member of the dataset's organization.
+  Users who can see a private dataset only as a CKAN *dataset collaborator*
+  (`ckan.auth.allow_dataset_collaborators`, off by default) do not see it in these lists;
+  nothing is revealed to anyone who should not see it.
+
+## [0.1.0] - 2026-09-18
+
+Published as source only (the initial standalone snapshot); no release or tag was created.
+
+### Added
+
+- reusable actor and contact point registries;
+- sysadmin management pages and inline creation from Scheming forms;
+- publisher and contact point selectors with record previews;
+- per-user latest-publisher and recent-contact-point convenience choices;
+- stable dataset references resolved at display and export time;
+- optional European DCAT-AP 3 profile extension with VCARD telephone output;
+- Alembic migrations for all extension-owned tables;
+- documentation for installation, integration, security and contribution;
+- English source strings throughout templates, Python and JavaScript, with a
+  compiled Swedish translation catalog using CKAN's own gettext-based
+  translation mechanism (`ITranslation`), the same one `ckanext-skelleftea`
+  uses.
+
+### Known limitations
+
+- automated verification covers CKAN 2.11.1 and later 2.11 patches and CKAN 2.12 (see
+  `docs/INSTALLATION.md`; this line said "2.11.6 only" before CI was extended);
+- browser and multi-version compatibility tests and a formal security review
+  remain work for a later release. A full registry Action API was
+  deliberately deprioritized rather than deferred -- see
+  `docs/BACKLOG.md` and `CONTRIBUTING.md`.
+
+[Unreleased]: https://github.com/bjornhagstrom/ckanext-actor-registry/compare/v0.3.0...HEAD
+[0.3.0]: https://github.com/bjornhagstrom/ckanext-actor-registry/compare/v0.2.3...v0.3.0
+[0.2.3]: https://github.com/bjornhagstrom/ckanext-actor-registry/releases/tag/v0.2.3
+[0.2.2]: https://github.com/bjornhagstrom/ckanext-actor-registry/releases/tag/v0.2.2
+[0.2.1]: https://github.com/bjornhagstrom/ckanext-actor-registry/releases/tag/v0.2.1
+[0.2.0]: https://github.com/bjornhagstrom/ckanext-actor-registry/releases/tag/v0.2.0
+[0.1.0]: https://github.com/bjornhagstrom/ckanext-actor-registry/tree/df14240

ckanext_actor_registry-0.3.0/CODE_OF_CONDUCT.md

@@ -0,0 +1,16 @@
+# Code of conduct
+
+We want participation in this project to be safe, respectful and constructive.
+
+Be considerate, assume good intent, welcome questions and critique ideas rather
+than people. Harassment, discriminatory language, threats, deliberate
+disruption and publication of another person's private information are not
+acceptable.
+
+Project maintainers may edit or remove contributions and restrict participation
+when this code of conduct is violated. Reports should be made privately to a
+maintainer through the repository's available private contact channel. Reports
+will be handled as confidentially as practical.
+
+This code applies in the repository and in public spaces where someone is
+representing the project.

ckanext_actor_registry-0.3.0/CONTRIBUTING.md

@@ -0,0 +1,76 @@
+# Contributing
+
+Thank you for helping improve ckanext-actor-registry. Version 0.2 is an alpha,
+so bug reports, compatibility results and design feedback are especially useful.
+
+## Before opening an issue
+
+- search existing issues;
+- include the CKAN, Python, ckanext-scheming and ckanext-dcat versions used;
+- describe the expected and actual result;
+- include reproducible steps and relevant logs with credentials and personal
+  data removed.
+
+Use a private security report instead of a public issue for vulnerabilities; see
+[SECURITY.md](SECURITY.md).
+
+## Translations
+
+A new language is welcome as a pull request: add
+`ckanext/actor_registry/i18n/<locale>/LC_MESSAGES/ckanext-actor-registry.po` and its
+compiled `.mo`. No code change is needed; the plugin offers every compiled catalogue.
+A partial translation is fine: strings not yet translated show in English. The tests
+check that the `.mo` matches the `.po`; only the maintained Swedish catalogues must be
+complete.
+See [docs/TRANSLATIONS.md](docs/TRANSLATIONS.md).
+
+## Pull requests
+
+1. Discuss substantial behavior or data-model changes in an issue first.
+2. Keep each pull request focused.
+3. Add or update tests when changing behavior. [TESTING.md](TESTING.md) shows how to run the
+   suite; continuous integration runs it on CKAN 2.11.1, the newest 2.11 patch and 2.12.
+4. Update the README and changelog when user-visible behavior changes.
+5. Internationalize all new user-visible strings: wrap them in `_(...)`
+   with English source text (`from ckan.common import _` in Python,
+   `{{ _("...") }}` in templates; JavaScript strings go through a
+   `data-i18n-*` attribute rendered server-side instead, see
+   `assets/js/actor-quick-create.js` for the pattern), then add the
+   Swedish `msgstr` to
+   `ckanext/actor_registry/i18n/sv/LC_MESSAGES/ckanext-actor-registry.po`,
+   copy the same entry to `.../i18n/sv_SE/LC_MESSAGES/ckanext-actor-registry.po` (the
+   `sv_SE` catalogue is a copy of `sv`, and a test fails if they differ) and recompile with `pybabel compile -d ckanext/actor_registry/i18n -D
+   ckanext-actor-registry` (also wired into `dev-install-extras.sh`, so a
+   dev-stack rebuild picks it up too). To re-extract the full catalog
+   from scratch, run `pybabel extract -F babel.cfg -o ckanext/actor_registry/i18n/ckanext-actor-registry.pot .` from this
+   repo's root -- see `babel_ckan_tag_shims.py` first if that ever
+   silently drops a template's strings again.
+6. Confirm that migrations preserve existing data and provide a downgrade where
+   practical.
+7. State whether and how generative AI tools assisted the contribution. The
+   contributor remains responsible for reviewing the work and for its license
+   and provenance.
+
+By submitting a contribution, you agree that it may be released under the MIT
+license used by this repository and confirm that you have the right to submit
+it.
+
+## Development priorities
+
+- browser-based (end-to-end) tests;
+- documented import and export workflows;
+- screenshots and a short walkthrough for the README.
+
+The automated model, authorization, form, validation and RDF serialization tests, and
+compatibility testing on CKAN 2.11.1+ and 2.12, are in place.
+
+A full `ckan.logic.action` CRUD API for registry management (create/read/
+update/delete actors and contact points programmatically) was considered and
+deliberately deprioritized: it is uncommon among comparable CKAN registry/
+metadata extensions, and the admin GUI plus the existing merge and delete
+actions already cover this extension's actual use cases. All validation lives in
+one shared layer (`validation.py`), so an API added later would reuse it. See `docs/BACKLOG.md` for the full rationale. This is an
+intentional design decision, not an open gap -- please raise it in an issue
+first if you think it should be revisited.
+
+All contributors must follow [CODE_OF_CONDUCT.md](CODE_OF_CONDUCT.md).

ckanext_actor_registry-0.3.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 Björn Hagström
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

ckanext_actor_registry-0.3.0/MANIFEST.in

@@ -0,0 +1,9 @@
+recursive-include ckanext/actor_registry/templates *
+recursive-include ckanext/actor_registry/assets *
+recursive-include ckanext/actor_registry/migration *
+recursive-include ckanext/actor_registry/i18n *
+include ckanext/actor_registry/config_declaration.yaml
+recursive-include ckanext/actor_registry/examples *.yaml
+recursive-include ckanext/actor_registry/tests/fixtures *.yaml
+include README.md LICENSE CHANGELOG.md CONTRIBUTING.md SECURITY.md CODE_OF_CONDUCT.md AI_ASSISTANCE.md
+recursive-include docs *.md