From 192c2030d3fd846852eb295ac5e4817b5070aa93 Mon Sep 17 00:00:00 2001 From: Walter Date: Wed, 11 Mar 2015 00:35:42 +0100 Subject: [PATCH] Dockerization + Fix for Tesseract 3.03 --- Dockerfile | 10 ++++++++++ build.sh | 2 ++ .../src/main/java/com/sismics/tess4j/Tesseract.java | 2 +- docs-docker-data/Dockerfile | 5 +++++ docs-docker-data/build.sh | 3 +++ docs-docker-data/run.sh | 3 +++ docs.xml | 10 ++++++++++ run-service.sh | 7 +++++++ 8 files changed, 41 insertions(+), 1 deletion(-) create mode 100644 Dockerfile create mode 100644 build.sh create mode 100644 docs-docker-data/Dockerfile create mode 100644 docs-docker-data/build.sh create mode 100644 docs-docker-data/run.sh create mode 100644 docs.xml create mode 100644 run-service.sh diff --git a/Dockerfile b/Dockerfile new file mode 100644 index 00000000..9ea8d2ad --- /dev/null +++ b/Dockerfile @@ -0,0 +1,10 @@ +FROM sismics/debian-java7-jetty9 +MAINTAINER benjamin.gam@gmail.com + +RUN apt-get -y -q install tesseract-ocr tesseract-ocr-fra + +ENV TESSDATA_PREFIX /usr/share/tesseract-ocr +ENV LC_NUMERIC C + +ADD docs-web/target/docs-web-*.war /opt/jetty/webapps/docs.war +ADD docs.xml /opt/jetty/webapps/docs.xml diff --git a/build.sh b/build.sh new file mode 100644 index 00000000..cd55f467 --- /dev/null +++ b/build.sh @@ -0,0 +1,2 @@ +#!/bin/sh +docker build -t sismics/docs . diff --git a/docs-core/src/main/java/com/sismics/tess4j/Tesseract.java b/docs-core/src/main/java/com/sismics/tess4j/Tesseract.java index 289d500d..84719470 100644 --- a/docs-core/src/main/java/com/sismics/tess4j/Tesseract.java +++ b/docs-core/src/main/java/com/sismics/tess4j/Tesseract.java @@ -53,7 +53,7 @@ public class Tesseract { private static Tesseract instance; private final static Rectangle EMPTY_RECTANGLE = new Rectangle(); private String language = "eng"; - private String datapath = "tessdata"; + private String datapath = null; private int psm = TessAPI.TessPageSegMode.PSM_AUTO; private boolean hocr; private int pageNum; diff --git a/docs-docker-data/Dockerfile b/docs-docker-data/Dockerfile new file mode 100644 index 00000000..2ec3d08a --- /dev/null +++ b/docs-docker-data/Dockerfile @@ -0,0 +1,5 @@ +FROM busybox +VOLUME ["/var/docs"] +ADD data /var/docs/ +RUN chown -R 1001:1001 /var/docs +CMD ["/bin/sh"] diff --git a/docs-docker-data/build.sh b/docs-docker-data/build.sh new file mode 100644 index 00000000..81f63c1d --- /dev/null +++ b/docs-docker-data/build.sh @@ -0,0 +1,3 @@ +#!/bin/sh + +docker build -t sismics/docs_data . diff --git a/docs-docker-data/run.sh b/docs-docker-data/run.sh new file mode 100644 index 00000000..84ea8d95 --- /dev/null +++ b/docs-docker-data/run.sh @@ -0,0 +1,3 @@ +#!/bin/sh +docker rm -f sismics_docs_data +docker run --name sismics_docs_data sismics/docs_data diff --git a/docs.xml b/docs.xml new file mode 100644 index 00000000..61756ef1 --- /dev/null +++ b/docs.xml @@ -0,0 +1,10 @@ + + / + /webapps/docs.war + + + docs.home + /var/docs + + + diff --git a/run-service.sh b/run-service.sh new file mode 100644 index 00000000..d8d77699 --- /dev/null +++ b/run-service.sh @@ -0,0 +1,7 @@ +#!/bin/bash +docker rm -f sismics_docs +docker run \ + -d --name=sismics_docs --restart=always \ + --volumes-from=sismics_docs_data \ + -e 'VIRTUAL_HOST_SECURE=docs.sismics.com' -e 'VIRTUAL_PORT=80' \ + sismics/docs:latest