From 0273fd5c071f67c2df2791374e52b6b641168b48 Mon Sep 17 00:00:00 2001
From: eknizat <26163420+eknizat@users.noreply.github.com>
Date: Thu, 6 Aug 2020 12:30:20 +0100
Subject: [PATCH] ATS-816: Fix tika apple keynote (#285)

* ATS-816: Fix tika apple keynote
The application/vnd.apple.keynote -> text/plain transformation has been found to fail after switching the version of tika in ATS-801
The previous version of tika would use the org.apache.tika.parser.pkg.PackageParser but the new version uses an empty parser producing empty target file.

* Re enable test for application/vnd.apple.keynote to text
---
 .../org/alfresco/transformer/TikaTransformationIT.java   | 2 +-
 .../src/main/resources/tika-config.xml                   | 9 +++++++++
 2 files changed, 10 insertions(+), 1 deletion(-)

diff --git a/alfresco-transform-tika/alfresco-transform-tika-boot/src/test/java/org/alfresco/transformer/TikaTransformationIT.java b/alfresco-transform-tika/alfresco-transform-tika-boot/src/test/java/org/alfresco/transformer/TikaTransformationIT.java
index 5744c701..9f61a61e 100644
--- a/alfresco-transform-tika/alfresco-transform-tika-boot/src/test/java/org/alfresco/transformer/TikaTransformationIT.java
+++ b/alfresco-transform-tika/alfresco-transform-tika-boot/src/test/java/org/alfresco/transformer/TikaTransformationIT.java
@@ -104,7 +104,7 @@ public class TikaTransformationIT
                 Stream.of(
                     Triple.of("quick.key", "html", "application/vnd.apple.keynote"),
                     // Does not work, alfresco-docker-sourceMimetype-misc can handle this target mimetype, removed from engine_config.json
-                    // Triple.of("quick.key", "txt", "TikaAuto"),
+                     Triple.of("quick.key", "txt", "application/vnd.apple.keynote"),
                     Triple.of("quick.key", "xhtml", "application/vnd.apple.keynote"),
                     Triple.of("quick.key", "xml", "application/vnd.apple.keynote")
                 ),
diff --git a/alfresco-transform-tika/alfresco-transform-tika/src/main/resources/tika-config.xml b/alfresco-transform-tika/alfresco-transform-tika/src/main/resources/tika-config.xml
index ef9f6df4..7b6fccb7 100644
--- a/alfresco-transform-tika/alfresco-transform-tika/src/main/resources/tika-config.xml
+++ b/alfresco-transform-tika/alfresco-transform-tika/src/main/resources/tika-config.xml
@@ -3,4 +3,13 @@
     <!-- This property, when set, will hide the start up warnings of tika for libraries are missing. -->
     <!-- See https://issues.apache.org/jira/browse/TIKA-2490 -->
     <service-loader initializableProblemHandler="ignore"/>
+
+    <parsers>
+        <!-- ATS-816: Use the PackageParser for application/vnd.apple.keynote.13 as that was used in tika-1.21-20190624-alfresco-patched -->
+        <parser class="org.apache.tika.parser.pkg.PackageParser">
+            <mime>application/vnd.apple.keynote.13</mime>
+        </parser>
+        <!-- Default parser needs to be included if the PackageParser parser is specified here, otherwise just the PackageParser is added-->
+        <parser class="org.apache.tika.parser.DefaultParser"/>
+    </parsers>
 </properties>