feat: add tika

apps/tika/Dockerfile

@@ -0,0 +1,64 @@
+ARG VERSION
+FROM ghcr.io/deedee-ops/ubuntu:22.04 as builder
+
+ARG VERSION
+ENV NEAREST_TIKA_SERVER_URL="https://dlcdn.apache.org/tika/${VERSION}/tika-server-standard-${VERSION}.jar" \
+    ARCHIVE_TIKA_SERVER_URL="https://archive.apache.org/dist/tika/${VERSION}/tika-server-standard-${VERSION}.jar" \
+    BACKUP_TIKA_SERVER_URL="https://downloads.apache.org/tika/${VERSION}/tika-server-standard-${VERSION}.jar" \
+    DEFAULT_TIKA_SERVER_ASC_URL="https://downloads.apache.org/tika/${VERSION}/tika-server-standard-${VERSION}.jar.asc" \
+    ARCHIVE_TIKA_SERVER_ASC_URL="https://archive.apache.org/dist/tika/${VERSION}/tika-server-standard-${VERSION}.jar.asc"
+
+SHELL ["/bin/bash", "-o", "pipefail", "-c"]
+#hadolint ignore=DL3008
+RUN set -x && apt-get update \
+ && apt-get -y install gnupg2 --no-install-recommends \
+ && curl -sL https://downloads.apache.org/tika/KEYS | gpg --import \
+ && curl -sL "${NEAREST_TIKA_SERVER_URL}" -o "/tika-server-standard-${VERSION}.jar" || echo "$?" || ls /tika* || rm "/tika-server-standard-${VERSION}.jar" \
+ && sh -c "[ -f /tika-server-standard-${VERSION}.jar ]" || curl -sL "${ARCHIVE_TIKA_SERVER_URL}" -o "/tika-server-standard-${VERSION}.jar" || ls /tika* ||  rm "/tika-server-standard-${VERSION}.jar" \
+ && sh -c "[ -f /tika-server-standard-${VERSION}.jar ]" || curl -sL "${BACKUP_TIKA_SERVER_URL}" -o "/tika-server-standard-${VERSION}.jar" || ls /tika* || rm "/tika-server-standard-${VERSION}.jar" \
+ && sh -c "[ -f /tika-server-standard-${VERSION}.jar ]" || exit 1 \
+ && curl -sL "${DEFAULT_TIKA_SERVER_ASC_URL}" -o "/tika-server-standard-${VERSION}.jar.asc" || rm "/tika-server-standard-${VERSION}.jar.asc" \
+ && sh -c "[ -f /tika-server-standard-${VERSION}.jar.asc ]" || curl -sL "${ARCHIVE_TIKA_SERVER_ASC_URL}" -o "/tika-server-standard-${VERSION}.jar.asc" || rm "/tika-server-standard-${VERSION}.jar.asc" \
+ && sh -c "[ -f /tika-server-standard-${VERSION}.jar.asc ]" || exit 1 \
+ && gpg --verify "/tika-server-standard-${VERSION}.jar.asc" "/tika-server-standard-${VERSION}.jar"
+
+FROM ghcr.io/deedee-ops/ubuntu:22.04
+
+ARG JRE='openjdk-17-jre-headless'
+ARG VERSION
+ENV VERSION="${VERSION}"
+
+SHELL ["/bin/bash", "-o", "pipefail", "-c"]
+#hadolint ignore=DL3008
+RUN set -eux \
+ && apt-get update \
+ && apt-get install --yes --no-install-recommends gnupg2 software-properties-common \
+ && add-apt-repository -y ppa:alex-p/tesseract-ocr5 \
+ && apt-get update \
+ && apt-get install --yes --no-install-recommends $JRE \
+    gdal-bin \
+    tesseract-ocr \
+    tesseract-ocr-eng \
+    tesseract-ocr-pol \
+ && echo ttf-mscorefonts-installer msttcorefonts/accepted-mscorefonts-eula select true | debconf-set-selections \
+ && apt-get install --yes --no-install-recommends \
+    xfonts-utils \
+    fonts-freefont-ttf \
+    fonts-liberation \
+    ttf-mscorefonts-installer \
+    cabextract \
+ && apt-get clean -y \
+ && rm -rf /var/lib/apt/lists/* /var/tmp/* /usr/share/man /usr/local/man /usr/local/share/man \
+ && mkdir -p /config/tmp /config/extras \
+ && chown -R abc:abc /config \
+ && chmod -R 777 /config
+
+COPY --from=builder /tika-server-standard-${VERSION}.jar /app/tika-server-standard-${VERSION}.jar
+USER 65000:65000
+
+VOLUME ["/tmp"]
+EXPOSE 9998
+ENTRYPOINT ["/usr/bin/tini", "--", "/usr/local/bin/base-entrypoint.sh"]
+CMD [ "/bin/sh", "-c", "java -cp \"/app/tika-server-standard-${VERSION}.jar:/config/extras/*\" org.apache.tika.server.core.TikaServerCli -h 0.0.0.0 $0 $@"]
+
+LABEL org.opencontainers.image.source="https://github.com/apache/tika/"

apps/tika/ci/goss.yaml

@@ -0,0 +1,32 @@
+---
+# https://github.com/goss-org/goss/blob/master/docs/manual.md#group
+group:
+  abc:
+    exists: true
+    gid: 65000
+# https://github.com/aelsabbahy/goss/blob/master/docs/manual.md#file
+file:
+  /usr/share/tesseract-ocr/5/tessdata/eng.traineddata:
+    exists: true
+  /usr/share/tesseract-ocr/5/tessdata/pol.traineddata:
+    exists: true
+# https://github.com/aelsabbahy/goss/blob/master/docs/manual.md#http
+http:
+  http://localhost:9998/parsers:
+    status: 200
+    timeout: 1000
+# https://github.com/goss-org/goss/blob/master/docs/manual.md#mount
+mount:
+  /tmp:
+    exists: true
+# https://github.com/aelsabbahy/goss/blob/master/docs/manual.md#port
+port:
+  # https://github.com/aelsabbahy/goss/issues/149
+  tcp:9998:
+    listening: true
+# https://github.com/goss-org/goss/blob/master/docs/manual.md#user
+user:
+  abc:
+    exists: true
+    uid: 65000
+    gid: 65000

apps/tika/metadata.json

@@ -0,0 +1,18 @@
+{
+  "app": "tika",
+  "base": false,
+  "testMuteCmd": false,
+  "channels": [
+    {
+      "name": "stable",
+      "renovate::dataSource": "github-tags",
+      "renovate::depName": "apache/tika",
+      "version": "2.9.1",
+      "platforms": [
+        "linux/amd64",
+        "linux/arm64"
+      ]
+    }
+  ]
+}
+