/GATEteamware

Docker build for GATE teamware server for collaborative extraction of entites, relations, correfference from unstructured text. WORK IN PROGRESS

NOT YET WORKING - not finished

Help gratefully received.

Gate General Architecture Text Extraction - unofficial

GATE Teamware provides a collaborative platform for text processing to extract entities, relations and correference. 1 GATE Teamware is hosted by sourceforge

Todo

  • runninr ant install on container start seems to exit the container. An error on install quits the container. fix BASEDIR issue on start
  • lockdown and secure
  • migrate setup perams to docker-compose.yml

Possible issues to investigate

  1. openjdk-6-jre vs openjdk-8-jre
  • seems to build with openjdk-8
  1. ANT_VERSION 1.9.7 vs the specified 1.8.4 (which didn't seem to build)

General disclaimer / plea

This is my first attempt at a more complex docker build, so be nice! Comments & suggestions welcomed.

Build process

Prerequisites

  • mysql
  • Java 6 JDK
  • ant version 1.8.x
  • PERL
  • wget
  • Image from mysql image

Customise Install

  • build and run the db first, it needs to be running for the app to finnish building. e.g. it needs to access the db.
  • I'm using a lan apt-cache-ng server, so you need to remove the RUN echo 'Acquire::http { Proxy "http://192.168.0.210:3142"; };' >> /etc/apt/apt.conf.d/01proxy line
  • Download gate from sourceforge and place the trunk folder in the same folder as the docker file. It gets loaded and copied to the image when it's built. I did this to stop me having to thrash my bw and the sourceforge servers every re-build!
  • change default values to suite env in antinstall-config.xml
  • mirror values in build.properties

Licence

GATE Teamware is available under the GNU Affero General Public Licence 3.0 so as I understand you are free to use the source, but must not sublicence without permission 2.

I'm trying to avoid the Oracle behemoth by using openjdk.

As for this docker file - do what you like. Let me know how you get on.

This was build by..

Peter Coghill (@petecog) of Aleph Insights