From 8d74e9eb55b1e530d3ac245ec6f10de6e4ce2a54 Mon Sep 17 00:00:00 2001
From: Francesco Frassinelli <francesco.frassinelli@nina.no>
Date: Tue, 19 Sep 2023 21:28:35 +0200
Subject: [PATCH] Build openrefine from sources

---
 .github/workflows/docker.yaml | 11 +++++++----
 Dockerfile                    | 28 +++++++++++++++++++++-------
 docker-compose.yml            |  4 +++-
 3 files changed, 31 insertions(+), 12 deletions(-)

diff --git a/.github/workflows/docker.yaml b/.github/workflows/docker.yaml
index 87f0bf3..d39cfb5 100644
--- a/.github/workflows/docker.yaml
+++ b/.github/workflows/docker.yaml
@@ -17,8 +17,11 @@ jobs:
       fail-fast: false
       matrix:
         include:
-          - build_args: version=3.7.5
-            tag: 3.7.5
+          - version: "master"
+          - version: "4.0"
+          - version: "3.7"
+          - version: "3.7.5"
+          - version: "3.6.2"
     permissions:
       contents: read
       packages: write
@@ -43,7 +46,7 @@ jobs:
         uses: docker/metadata-action@v4
         with:
           images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
-          tags: ${{ matrix.tag }}
+          tags: ${{ matrix.version }}
       -
         name: Build and push container
         uses: docker/build-push-action@v3
@@ -51,6 +54,6 @@ jobs:
           cache-from: type=gha
           cache-to: type=gha,mode=max
           push: ${{ github.event_name != 'pull_request' }}
-          build-args: ${{ matrix.build_args }}
+          build-args: VERSION=${{ matrix.version }}
           tags: ${{ steps.meta.outputs.tags }}
           labels: ${{ steps.meta.outputs.labels }}
diff --git a/Dockerfile b/Dockerfile
index 39743eb..4800db7 100644
--- a/Dockerfile
+++ b/Dockerfile
@@ -1,14 +1,28 @@
-FROM quay.io/centos/centos:stream9-minimal as base
-ARG VERSION=3.7.5
+FROM registry.opensuse.org/opensuse/git:latest AS sources
+ARG VERSION=master
+WORKDIR /opt/openrefine
+RUN git clone https://github.com/OpenRefine/OpenRefine.git --depth 1 --branch $VERSION .
 
-RUN microdnf install --assumeyes --setopt=install_weak_deps=0 java-17-openjdk-headless procps which gettext tar gzip
+FROM registry.opensuse.org/opensuse/bci/openjdk-devel:latest AS backend
 WORKDIR /opt/openrefine
-RUN curl -L https://github.com/OpenRefine/OpenRefine/releases/download/$VERSION/openrefine-linux-$VERSION.tar.gz | \
-    tar xz --strip-components 1
-COPY entrypoint.sh refine.ini.template LICENSE.txt .
+COPY --from=sources /opt/openrefine .
+RUN --mount=type=cache,target=/root/.m2 \
+    mvn -B package -P linux -DskipTests=true -Dmaven.antrun.skip=true
 
-EXPOSE 3333/TCP
+FROM registry.opensuse.org/opensuse/bci/nodejs:latest AS frontend
+WORKDIR /opt/openrefine/main/webapp
+COPY --from=sources /opt/openrefine/main/webapp .
+RUN --mount=type=cache,target=/root/.npm \
+    npm install
 
+FROM registry.opensuse.org/opensuse/bci/openjdk:latest
+RUN zypper --non-interactive install gettext-tools
+WORKDIR /opt/openrefine
+COPY --from=backend /opt/openrefine .
+COPY --from=frontend /opt/openrefine/main/webapp/modules main/webapp/modules
+COPY entrypoint.sh refine.ini.template ./
+
+EXPOSE 3333/TCP
 ENV REFINE_MEMORY=1400M
 ENV REFINE_MIN_MEMORY=1400M
 
diff --git a/docker-compose.yml b/docker-compose.yml
index d15ddcd..65b2459 100644
--- a/docker-compose.yml
+++ b/docker-compose.yml
@@ -3,6 +3,8 @@ services:
     build:
       context: .
       dockerfile: Dockerfile
+      args:
+        VERSION: master
     environment:
       REFINE_MEMORY: 2000M
       REFINE_MIN_MEMORY: 2000M
@@ -17,7 +19,7 @@ services:
       timeout: 10s
       interval: 1m0s
       retries: 5
-    image: ghcr.io/ninanor/openrefine:main
+    image: ghcr.io/ninanor/openrefine:master
     networks:
       default: null
     ports: