CHANGELOG

1.1.1
 - fix: retry fetching mimetype in background job
1.1.0
 - feat: implement CLI command `dcor-purge-unused-collections-and-circles`
 - ref: rewrite `dcor_prune_draft_datasets` with database queries
1.0.2
 - fix: only group admins and editors may add or remove datasets
 - fix: only group admins may add or remove users
 - fix: do not allow listing users details for `group_list`
1.0.1
 - fix: deny user listing for non-admins
 - enh: add auth function for "user_autocomplete"
1.0.0
 - feat: send email to maintainer for new users (#36)
 - tests: add test methods for deleting collections/groups (#26)
0.24.5
 - fix: add anonymous user check when extending group listing
0.24.4
 - reg: double check whether resource changed when setting basic metadata
0.24.3
 - fix: only update basic resource metadata in background job when necessary
 - enh: add "licenses_medical.json" for non-open instances
0.24.2
 - enh: removing draft datasets did not work when search index was outdated
0.24.1
 - enh: ignore invalid, non-JSON-encodable resource metadata
 - enh: relax maximum email length to 40 characters
 - tests: add test for displaying dataset on web interface as anon user (#21)
0.24.0
 - feat: new CLI script `dcor-prune-orphaned-s3-artifacts`
 - feat: new CLI script `dcor-prune-draft-datasets`
0.23.0
 - feat: add `send_mail` CLI plugin for sending emails with CKAN SMTP
 - enh: restrict fetching user info for unauthorized parties
 - tests: improve dataset migration tests
0.22.0
 - integration with DCOR version bump for CKAN 2.11.3
 - BREAKING: remove `DISABLE_AFTER_DATASET_CREATE_FOR_CONCURRENT_JOB_TESTS`
 - fix: correctly detect missing `licenses_grou_url`
 - fix: hide featured collections and circles on landing page for CKAN 2.11
 - ref: migrate S3 public-tagging of resources from ckanext-dcor_depot
 - enh: add condition and replicate to supplementary resource schemas
0.21.5
 - fix: extract correct package_id when updating resource
 - ci: update GitHub Actions workflow to use Docker for testing
 - tests: migrate all tests to S3 upload
0.21.4
 - fix: `get_base_metadata` failed when resource not defined yet
0.21.3
 - fix: do not modify `licenses_group_url` (#32)
0.21.2
 - fix: search and fail better when looking for licenses.json
 - enh: update list of common model organisms
 - enh: add condition and repetition to resource supplement
0.21.1
 - docs: correct title of job_set_resource_metadata_base
0.21.0
 - ref: migrate to dcor_shared.RQJob for background jobs
 - enh: introduce "is_background_job" context
0.20.6
 - ref: enqueue eTag retrieval "at front"
 - ref: introduce `IS_BACKGROUND_JOB` as workaround to prevent background
   jobs to enqueue other background jobs via `after_dataset_update`
 - ref: other code cleanup
0.20.5
 - fix: avoid excessive API calls in auth.package_update (performance)
 - ref: remove unused and broken plugin..after_resource_update
 - ref: minor reduction in API calls in plugin..after_dataset_update
0.20.4
 - fix: datetime.UTC does not exist in Python < 3.11
0.20.3
 - enh: improve spam user signup detection
 - ref: replace deprecations
 - ci: update workflows
0.20.2
 - setup: fix ckanext namespace
0.20.1
 - build: fix CI pipeline
0.20.0
 - fix: register validators instead of calling them directly (#30)
 - fix: allow to restrict content visibility with config option
   ckanext.dcor_schemas.allow_content_listing_for_anon
 - enh: declare custom configuration options in IConfigDeclaration
 - setup: bump dclab to 0.60.9
 - setup: migrate to pyproject.toml
0.19.2
 - enh: add CLI script for moving a dataset to a different circle
0.19.1
 - fix: list-group-resources CLI should return draft and deleted resources
0.19.0
 - feat: implement resource_upload_s3_urls API method
 - feat: CLI command list-group-resources for listing all resources
   in a circle or collection
 - feat: add ETag property which is populated with ETag from S3 via a job
 - fix: non-sysadmins may not specify dataset ID
 - fix: improper extraction of resource suffix for mimetype
 - enh: more strict resource and dataset id validators
 - enh: do not allow user to specify SHA256 hash
 - setup: migrate to dcor_shared 0.10.0
 - tests: migrate to helper methods from dcor_shared
 - tests: increased coverage
0.18.11
 - maintenance release
0.18.10
 - ref: use `url_validator` that works with ports in CKAN 2.10.3
0.18.9
 - fix: invalid validator check for context keyword
0.18.8
 - fix: list-zombie-users still relied on activities
0.18.7
 - enh: let admin users validate invalid DC resources
0.18.6
 - fix: create default resource views since CKAN 2.10.1 does not
 - fix: deprecation in dclab.dfn
 - fix: url validator not working
0.18.5
 - enh: add `url_with_port_validator` validator
 - enh: add CLI `list-circles` and `list-collections`
 - ref: move sha256sum to dcor_shared
0.18.4
 - enh: update the resource schema with 's3_available' and 's3_url'
 - ref: make sha256sum available as separate method
0.18.3
 - enh: more protection against script kiddies
0.18.2
 - tests: don't use the new `after_dataset_update` during concurrent job tests
0.18.1
 - fix: call `after_resource_create` via custom `after_dataset_update` for
   all plugins (because this is not done by CKAN)
 - tests: fix cli tests
0.18.0
 - ref: migration to CKAN 2.10
0.17.2
 - enh: make bot registration detection more verbose
 - ref: minor regex optimization
 - ref: change "Package ID" to "Dataset ID" in dataset view
0.17.1
 - enh: next escalation step against spam registration (#14)
0.17.0
 - fix: disallow bulk menu for making datasets private (#23)
 - enh: don't show package editing options to users who don't
   have access (#22)
 - ref: removed templates related to data upload via the web
   interface which has been disabled in 0.16.0 (commit bd5461e)
0.16.4
 - fix: typo in run_jobs_dcor_schemas CLI cause OSError
0.16.3
 - fix: handle errors during run_jobs_dcor_schemas CLI operations
0.16.2
 - fix: displaying dataset failed when user not logged-in
0.16.1
 - tests: add authfunction test for collaborators (#10)
 - enh: only allow datasets with at least one valid .rtdc file (#5)
 - tests: use create_with_upload fixture for resource creation
0.16.0
 - fix: disable browser-based uploads (#20)
 - ref: use the original auth function for package_create to
   make things more transparent (even though now the error
   messages are the default ones again)
0.15.13
 - enh: do not allow signing up from gmail addresses (#14)
0.15.12
 - fix: add redis connection
0.15.11
 - fix: prevent enqueueing of same background job twice
0.15.10
 - ref: decomplicate after_create (b/c package_revise now kind of works)
 - tests: remove dcor_depot dependency
0.15.9
 - ref: Implement workaround to support package_revise (see #19)
0.15.8
 - fix: regression where resource supplement data could not be
   written to resource metadata
0.15.7
 - fix: seal auth functions for usage with package_revise
0.15.6
 - fix: prevent removal or addition of resources to active datasets
   via package_revise
0.15.5
 - setup: bump rq to actual supported version
0.15.4
 - enh: add --modified-days argument to CLI run-jobs-dcor-schemas
0.15.3
 - setup: bump rq to >=1.8.0, because we need multiple job dependencies
0.15.2
 - fix: run all background jobs sequentially to avoid race conditions
   with package_revise
 - ref: change background job IDs to "{package_id}_{position}_{task}}"
0.15.1
 - enh: allow resources with .poly and .sof suffix
0.15.0
 - feat: add new "run-jobs-dcor-schemas" CLI command that runs all
   background jobs on all dataset resources
 - enh: increase timeout values for jobs (things might be slower
   if `/data` is a network share)
 - setup: bump dcor_shared from 0.2.9 to 0.2.10
0.14.7
 - setup: bump dclab to 0.35.1
0.14.6
 - fix: remove redunant entry for "chip: comment" in the "experiment"
   section of the supplementary resource metadata schema
0.14.5
 - tests: increase coverage for empty uploads
 - build: bump dcor_shared from 0.2.8 to 0.2.9 (handle small files)
0.14.4
 - fix: allow uploading .ini files to DCOR
0.14.3
 - fix: raise tk.Invalid when the state of a package is set to "active"
   and it does not have any resources
0.14.2
 - enh: add individual resource supplement section for chip information
0.14.1
 - fix: ckanext.dcor_schemas.allow_public_datasets was not parsed
   correctly (string == True instead of boolean)
 - ci: fix codecov
0.14.0
 - feat: add new configuration option to disable public datasets
   "ckanext.dcor_schemas.allow_public_datasets"
0.13.0
 - feat: allow to optionally set a custom licenses_group_url
   in ckan.ini without breaking things
 - fix: setting custom json_resource_schema_dir did not work
 - setup: setup.py test is deprecated
0.12.4
 - tests: implement testing pipeline using vagrant and GH Actions
0.12.3
 - fix: regression where some dataset edit pages could not be accessed
 - fix: "private" key was not parsed correctly in auth.py when updating
   a dataset
 - tests: increase coverage
 - setup: bump dcor_shared from 0.2.0 to 0.2.8
0.12.2
 - fix: regression where a new dataset could not be created via the
   web interface
 - tests: increase coverage
0.12.1
 - fix: browsing to "/dataset" resulted in 500 Internal Error
0.12.0
 - enh: allow users to keep a draft dataset and remove the option
   to upload additional datasets to active datasets within 3h
 - fix: users could create resources in datasets of other users
 - ref: made it more clear what is auth and what is validate
 - tests: increase coverage
0.11.2
 - build: use install_requires instead of requirements.txt
0.11.1
 - ci: migrate to GH Actions
0.11.0
 - enh: add CLI "list-zombie-users" to list users with no datasets
   and no activity for a certain amount of time
0.10.1
 - enh: update supplementary resource schema
0.10.0
 - enh: implement supplementary resource json files (#6)
 - enh: allow to get supported resource suffixes via API
 - fix: TemplateSyntaxError in bulk_process.html
 - fix: use package_revise to avoid concurrency issues with
   background jobs that set resource metadata
 - ref: remove custom form fileds in favor of
   supplementary resource data
0.9.5
 - enh: only create relevant jobs for DC mimetype data
0.9.4
 - fix: permissions issue when patching resources in jobs
0.9.3
 - enh: add a ``job_id`` for all background jobs and put them
   in different queues
 - ref: reworked dataset slug validation (dashes instead of
   underscores)
0.9.2
 - fix: manifest did not include .yml
0.9.1
 - maintenance release
0.9.0
 - initial open-sourced version