- sharectl command line tool
- Profiling middleware for local development
- Janitor tasks to find and process unprocessed data
- Timestamp field to RawData
- Mendeley Harvester!
- Started to use deprecation warning
- Timeouts for harvests
- The concept of "Bots"
- A lot of dead code
- A GPL licenced library
- Upgraded to Celery 4.0
- Deleted works now return 403s from the API
- Deleted works are now excluded from the API
- Corrected to date fields used to audit the Elasticsearch index
- Strongly defined the Harvester interface
- Harvests are now scheduled in a more friendly manner
- Updated the configurations for many OAI sources
- HarvestLogs no longer get stuck in progress
- Text parsing transformer utilties
- Elasticsearch Janitor task to keep Postgres and ES in sync
- Concurrently added indexes
- Admin updates to allow quicker fixing of broken data
- More test coverage
- Elasticsearch's scroll API explicitly disabled
- Upgraded to Django 1.11
- Elasticsearch now pulls last_modified from itself rather than Postgres
- API pagination no longer times out on large collections
- Timestamps are now included in the ATOM feed
- OAI endpoint
- Sources
- OpenBU
- Updated documentation
- Sources
- A table for managing SHARE data sources
- Replaces the apps in the providers folder
- SourceConfigs
- A table for managing different methods of acquire data from given source
- Replaces nested apps/app labels
- HarvestLogs
- First class support for managing harvesting/back harvesting
- Source Unique Identifiers
- First class representation of what was RawData.provider_doc_id
- The Django admin now supports starting harvesters over long periods of time
- Support for the MODs OAI PHM prefix
- Provider Django applications have been removed
- Source specific fields have been removed from ShareUser
- Harvesters have been relocated to share/harvesters/
- Various renaming/vocabulary changes
- RawData -> RawDatum
- Favicon -> Icon
- Provider -> Source
- Provider App -> SourceConfig
- Normalizer -> Transformer
- Updates to the getting started guide
- Squashed migrations to speed up local development
- Harvesters are now expected to return utf-8 strings
- Sources are no longer tied to the ShareUser model
- Title now has an "exact" multi-field in elasticsearch
- A robot that archives old succeeded celery jobs
- New Harvesters
- Scholarly Commons @ JMU
- Compensate for potential race conditions with the push API
- New Harvesters
- Research Registry Harvester
- SSOAR
- Status API endpoint
- Updated set_specs for University of Kansas
- ClinicalTrials.gov now output registrations
- Source icons are now stored in the database
- Removed "Notify" from the page title in the browsable API
- Support for OSF Registries
- New Harvesters
- University of Utah
- Updated the API
- Improved Elasticsearch mappings
- Updated NIH and NSFAwards
- Affiliations are now gathered
- Non-Unique URLs are no longer collected
- Lots of under the hood changes to make dev's lives easier
- New Harvesters
- es.csic
- edu.purdue.epubs
- Site status banners
- Retraction harvesting
- A little bit of documentation
- OAuth login failure pages look nice now
- Cascade deletes are now implemented as database cascades
- New Harvesters
- edu.cornell
- edu.richmond
- edu.scholarworks_montana
- edu.ucf
- edu.umd
- edu.utahstate
- org.seafdec
- Relations between creative works
- Updated harvesters
- Figshare v2 API
- PeerJ XML API
- Pubmed PMC prefix
- Datacite 4.0
- BePress Taxonomy for subjects
- Travis now uses postgres 9.5
- Comprehensive test suite for normalization and disambiguation
- Updated data model
- More expressive relations between people/organizations and works
- Type hierarchies
- Creative works: Publication, Preprint, DataSet, Patent, Thesis, Software, etc.
- Agents: Person, Organization, Institution, Consortium
- More aggressive and intelligent data parsing
- Stricter validation of incoming data
- Prune duplicate objects from submitted changesets
- Various bug fixes
- Formalized disambiguation methods
- App bootstrap time improved by 4x
- Better elasticsearch mappings
- URI may now be searched/matched directly
- Prettier table names
- Backport of the V1 push API
- New and improved source registration form
- JSON schema endpoint
- New sources
- College of William and Mary
- University of Wisconsin