Enable all charsets in native image for quarkus-tika

galderz · Oct 18, 2021 · 5d43d77 · 5d43d77
1 parent 93f3547
commit 5d43d77
Show file tree

Hide file tree

Showing 3 changed files with 6 additions and 17 deletions.
diff --git a/docs/src/main/asciidoc/tika.adoc b/docs/src/main/asciidoc/tika.adoc
@@ -11,19 +11,6 @@ This guide explains how your Quarkus application can use https://tika.apache.org
 
 https://tika.apache.org/[Apache Tika] is a content analysis toolkit which is used to parse the documents in PDF, Open Document, Excel and many other well known binary and text formats using a simple uniform API. Both the document text and properties (metadata) are available once the document has been parsed.
 
-[NOTE]
-====
-If you are planning to run the application as a native executable and parse documents that may have been created with charsets different than the standard ones supported in Java such as `UTF-8` then you should configure Quarkus to get the native image generator include all the charsets available to the JVM:
-[source,xml]
-----
-<properties>
-    <quarkus.package.type>native</quarkus.package.type>
-    <quarkus.native.add-all-charsets>true</quarkus.native.add-all-charsets>
-<properties>
-----
-====
-
-
 == Prerequisites
 
 To complete this guide, you need:

diff --git a/extensions/tika/deployment/src/main/java/io/quarkus/tika/deployment/TikaProcessor.java b/extensions/tika/deployment/src/main/java/io/quarkus/tika/deployment/TikaProcessor.java
@@ -26,6 +26,7 @@
 import io.quarkus.deployment.annotations.ExecutionTime;
 import io.quarkus.deployment.annotations.Record;
 import io.quarkus.deployment.builditem.FeatureBuildItem;
+import io.quarkus.deployment.builditem.NativeImageEnableAllCharsetsBuildItem;
 import io.quarkus.deployment.builditem.nativeimage.NativeImageResourceBuildItem;
 import io.quarkus.deployment.builditem.nativeimage.NativeImageResourceDirectoryBuildItem;
 import io.quarkus.deployment.builditem.nativeimage.RuntimeInitializedClassBuildItem;
@@ -87,6 +88,11 @@ public void registerPdfBoxResources(BuildProducer<NativeImageResourceDirectoryBu
         resource.produce(new NativeImageResourceDirectoryBuildItem("org/apache/fontbox/unicode"));
     }
 
+    @BuildStep
+    public NativeImageEnableAllCharsetsBuildItem registerAllCharsets() {
+        return new NativeImageEnableAllCharsetsBuildItem();
+    }
+
     @BuildStep
     @Record(ExecutionTime.STATIC_INIT)
     void initializeTikaParser(BeanContainerBuildItem beanContainer, TikaRecorder recorder,

diff --git a/integration-tests/tika/pom.xml b/integration-tests/tika/pom.xml
@@ -99,10 +99,6 @@
                     <name>native</name>
                 </property>
             </activation>
-            <!-- add some custom config, the rest comes from parent -->
-            <properties>
-                <quarkus.native.add-all-charsets>true</quarkus.native.add-all-charsets>
-            </properties>
         </profile>
     </profiles>