This is a grails application for managing species lists. It supports:
- Upload of species profile data in CSV format
- UI for viewing and faceting species
- Supports any number of key/value pairs
- JSON services used by
- BIE - species pages (https://bie.ala.org.au)
- Biocache - indexing occurrences (https://biocache.ala.org.au)
- Supports single owner & multiple editors of a single list
- Lists can be tagged as:
- authoritative
- conservation
- sensitive
- Supports sensitive species lists
- Grails web application
- MySQL database
There is an Ansible playbook for this application here ala-install/ansible/species-list-standalone.yml See example inventory ala-install/ansible/inventories/vagrant/species-list-vagrant
This application only needs a running MySQL instance and the app itself. There is a docker compose file that can be used to run a local MySQL instance. To use, install docker and docker compose and run:
docker-compose -f mysql.yml up
This should setup a running MySQL instance and expose port 3306. This is for development purposes only. Once the docker instance is running, the app can be started using
grails run-app
- **Version 5.2.1 **:
- improved the performance of rematching processing
- Fixed list csv downloads has duplicate columns
- Fixed the broken species list upload from spatial issue
- Replaced deprecated functions provided by AuthService plugin
- Grails 6 upgraded
- Version 3.0:
- Grails 3, Lucene 5 or above
- Version 2.0:
- upgraded to Grails 3
- Version 1.0:
- upgraded to asset pipeline and grails 2.5.6
- fixed unit and integration tests
-
Version 5.2.1: Database change:
- rematch_Log table changed:
DROP TABLE IF EXISTS `rematch_log`; CREATE TABLE `rematch_log` ( `id` bigint NOT NULL AUTO_INCREMENT, `version` bigint NOT NULL, `end_time` datetime(6) DEFAULT NULL, `processing` varchar(255) NOT NULL, `start_time` datetime(6) NOT NULL, `by_whom` varchar(255) NOT NULL, `status` varchar(255) DEFAULT NULL, `latest_processing_time` datetime(6) NOT NULL, `history` longtext, PRIMARY KEY (`id`) ) ENGINE=InnoDB AUTO_INCREMENT=1 DEFAULT CHARSET=utf8mb4;
After the database change, change 'dataSource.dbCreate=none' into local config file to speed up the application startup.