svn commit: r1064232 - in /tika/trunk: tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java

[ni...@apache.org]

Author: nick
Date: Thu Jan 27 17:54:37 2011
New Revision: 1064232

URL: http://svn.apache.org/viewvc?rev=1064232&view=rev
Log:
Fix up the iwork mime types with the patch from TIKA-588, and also add a unit test for the detection using the non-container detector (we already had container aware detector iwork tests)

Modified:
    tika/trunk/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml
    tika/trunk/tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java

Modified: tika/trunk/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml
URL: http://svn.apache.org/viewvc/tika/trunk/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml?rev=1064232&r1=1064231&r2=1064232&view=diff
==============================================================================
--- tika/trunk/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml (original)
+++ tika/trunk/tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml Thu Jan 27 17:54:37 2011
@@ -562,20 +562,23 @@
 
   <mime-type type="application/vnd.apple.iwork">
     <sub-class-of type="application/zip"/>
-    <glob pattern="*.key"/>
-    <glob pattern="*.pages"/>
-    <glob pattern="*.numbers"/>
   </mime-type>
-
   <mime-type type="application/vnd.apple.keynote">
     <root-XML localName="presentation" namespaceURI="http://developer.apple.com/namespaces/keynote2" />
+    <sub-class-of type="application/vnd.apple.iwork" />
+    <glob pattern="*.key"/>
   </mime-type>
   <mime-type type="application/vnd.apple.pages">
     <root-XML localName="document" namespaceURI="http://developer.apple.com/namespaces/sl" />
+    <sub-class-of type="application/vnd.apple.iwork" />
+    <glob pattern="*.pages"/>
   </mime-type>
   <mime-type type="application/vnd.apple.numbers">
     <root-XML localName="document" namespaceURI="http://developer.apple.com/namespaces/ls" />
+    <sub-class-of type="application/vnd.apple.iwork" />
+    <glob pattern="*.numbers"/>
   </mime-type>
+
   <mime-type type="application/vnd.arastra.swi">
     <glob pattern="*.swi"/>
   </mime-type>

Modified: tika/trunk/tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java
URL: http://svn.apache.org/viewvc/tika/trunk/tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java?rev=1064232&r1=1064231&r2=1064232&view=diff
==============================================================================
--- tika/trunk/tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java (original)
+++ tika/trunk/tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java Thu Jan 27 17:54:37 2011
@@ -149,6 +149,26 @@ public class TestMimeTypes extends TestC
         assertTypeByNameAndData("application/vnd.ms-powerpoint.slideshow.macroenabled.12", "testPPT.ppsm");
     }
 
+    /**
+     * Note - detecting container formats by mime magic is very very
+     *  iffy, as we can't be sure where things will end up.
+     * People really ought to use the container aware detection...
+     */
+    public void testIWorkDetection() throws Exception {
+        // By name is easy
+       assertTypeByName("application/vnd.apple.keynote", "testKeynote.key");
+       assertTypeByName("application/vnd.apple.numbers", "testNumbers.numbers");
+       assertTypeByName("application/vnd.apple.pages", "testPages.pages");
+       
+       // We can't do it by data, as we'd need to unpack
+       //  the zip file to check the XML 
+       assertTypeByData("application/zip", "testKeynote.key");
+       
+       assertTypeByNameAndData("application/vnd.apple.keynote", "testKeynote.key");
+       assertTypeByNameAndData("application/vnd.apple.numbers", "testNumbers.numbers");
+       assertTypeByNameAndData("application/vnd.apple.pages", "testPages.pages");
+    }
+
     public void testJpegDetection() throws Exception {
         assertType("image/jpeg", "testJPEG.jpg");
         assertTypeByData("image/jpeg", "testJPEG.jpg");