Geonetwork Code Sprint: Openlayers and ElasticSearch migration

The last Geonetwork code sprint took place on the premises of Camptocamp in France at Le Bourget du Lac from December 2 to 6, 2019. It had the dual objective of migrating the OpenLayers and ElasticSearch components to their most recent version.

The organization of this code sprint was made possible by the support of several organizations that we warmly thank here: EEA, Ifremer, Camptocamp, titellus, GeoCat BV, OSGeo, OSGeo-UK.

Openlayers

The Openlayers library was shipped via the third-party Ngeo library providing helpers for integrating Openlayers with AngularJS. Ngeo has migrated its architecture to a more modern operating mode based on ES6 modules, to follow the architecture proposed by Openlayers during its release 5.0. The front architecture of Geonetwork being still based on Closure Compiler and wro4j, it was impossible to update ngeo, and therefore Openlayers. We therefore included Openlayers in direct dependency and rewrote the Ngeo components in the native code of Geonetwork, to continue to provide the same services. By isolating the integration of Openlayers, we were able to update the library to its latest version, 6.1.1. Geonetwork is therefore ready to take advantage of the latest advances in the Openlayers library and its new WebGL engine, which offers rich and dynamic data visualization possibilities.

ElasticSearch

This code sprint is part of the current evolutions towards the future version 4 of Geonetwork. Building on ElasticSeach, Geonetwork 4 will offer users more efficient search (rough search, suggestions) with high performance gains. Geonetwork 4 can also be deployed in a cluster in a high availability architecture. More technical information is available here: https://github.com/geonetwork/core-geonetwork/wiki/Openlayers-migration-codesprint-December-2019