TIKA-1823 AutoCAD DWF mime magic and subtypes

Parent AutoCAD DWF mimetype and general magic, based on patch from
Luca Moretti, along with version-specific subtypes with more specific
magic

Author: Gagravarr

tika-core/src/main/resources/org/apache/tika/mime/tika-mimetypes.xml

@@ -5250,8 +5250,38 @@
   </mime-type>
 
   <mime-type type="model/vnd.dwf">
+    <acronym>DWF</acronym>
+    <_comment>AutoCAD Design Web Format</_comment>
+    <alias type="drawing/x-dwf"/>
+    <magic priority="50">
+      <match type="string" offset="0" value="(DWF V">
+         <match type="string" offset="8" value=".">
+            <match type="string" offset="11" value=")" />
+         </match>
+      </match>
+    </magic>
     <glob pattern="*.dwf"/>
   </mime-type>
+  <mime-type type="model/vnd.dwf;version=6">
+    <!-- Zip file with DWF header on the front -->
+    <magic priority="60">
+      <match type="string" offset="0" value="(DWF V06.">
+         <match type="string" offset="11" value=")PK" />
+      </match>
+    </magic>
+    <sub-class-of type="model/vnd.dwf"/>
+  </mime-type>
+  <mime-type type="model/vnd.dwf;version=5">
+    <magic priority="60">
+      <match type="string" offset="0" value="(DWF V00.55)"/>
+    </magic>
+  </mime-type>
+  <mime-type type="model/vnd.dwf;version=2">
+    <magic priority="60">
+      <match type="string" offset="0" value="(DWF V00.22)"/>
+    </magic>
+  </mime-type>
+
   <mime-type type="model/vnd.dwfx+xps">
     <_comment>AutoCAD Design Web Format</_comment>
     <glob pattern="*.dwfx"/>

tika-parsers/src/test/java/org/apache/tika/mime/TestMimeTypes.java

@@ -537,6 +537,11 @@ public void testAutoCADDetection() throws Exception {
         assertTypeByData("image/vnd.dwg", "testDWG2007.dwg");
         assertTypeByData("image/vnd.dwg", "testDWG2010.dwg");
 
+        // From name, gets the common parent type
+        assertTypeByName("model/vnd.dwf", "x.dwf");
+        // With the data, can work out it's the v6 zip-based flavour
+        assertTypeByData("model/vnd.dwf; version=6", "testDWF2010.dwf");
+        
         // From name, gets the common parent type
         assertTypeByName("image/vnd.dxf", "x.dxf");
         // With the data, can work out it's the ASCII flavour