TIKA-4223 -- add detection of stl (#1691)

* TIKA-4223 -- add detection for binary and text based stl

(cherry picked from commit 9d45b69)

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

@@ -2072,7 +2072,10 @@
     <glob pattern="*.cat"/>
   </mime-type>
   <mime-type type="application/vnd.ms-pki.stl">
-    <glob pattern="*.stl"/>
+    <!-- on TIKA-4223, we moved this glob to model/x.stl-binary.
+    We think this pki.stl is a subtype of pkcs7-signature?!
+    -->
+    <!--<glob pattern="*.stl"/> -->
   </mime-type>
   <mime-type type="application/vnd.ms-playready.initiator+xml"/>
 
@@ -7041,7 +7044,17 @@
     <glob pattern="*.mesh"/>
     <glob pattern="*.silo"/>
   </mime-type>
-
+  <mime-type type="model/x.stl-ascii">
+    <magic priority="60">
+      <match value="solid " offset="0" type="string">
+        <match value="facet " offset="7:256" type="string"/>
+      </match>
+    </magic>
+  </mime-type>
+  <mime-type type="model/x.stl-binary">
+    <_comment>no magic available</_comment>
+    <glob pattern="*.stl"/>
+  </mime-type>
   <mime-type type="model/vnd.dwf">
     <acronym>DWF</acronym>
     <_comment>AutoCAD Design Web Format</_comment>

tika-core/src/test/java/org/apache/tika/TikaDetectionTest.java

@@ -354,7 +354,7 @@ public void testHttpServerFileExtensions() {
         assertEquals("application/vnd.ms-ims", tika.detect("x.ims"));
         assertEquals("application/vnd.ms-lrm", tika.detect("x.lrm"));
         assertEquals("application/vnd.ms-pki.seccat", tika.detect("x.cat"));
-        assertEquals("application/vnd.ms-pki.stl", tika.detect("x.stl"));
+        assertEquals("model/x.stl-binary", tika.detect("x.stl"));
         assertEquals("application/vnd.ms-powerpoint", tika.detect("x.ppt"));
         assertEquals("application/vnd.ms-powerpoint", tika.detect("x.pps"));
         assertEquals("application/vnd.ms-powerpoint", tika.detect("x.pot"));

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

@@ -212,6 +212,12 @@ public void testSRT() throws Exception {
         assertTypeByNameAndData("application/x-subrip", "test_subrip.srt");
     }
 
+    @Test
+    public void testSTL() throws Exception {
+        assertTypeByNameAndData("model/x.stl-binary", "testSTL-binary.stl");
+        assertTypeByNameAndData("model/x.stl-ascii", "testSTL-ascii.stl");
+    }
+
     @Test
     public void testTTML() throws Exception {
         assertTypeByData("application/ttml+xml", "test_ttml.ttml");

tika-parsers/tika-parsers-standard/tika-parsers-standard-package/src/test/resources/test-documents/testSTL-ascii.stl

@@ -0,0 +1,16 @@
+solid OpenSCAD_Model
+  facet normal 0 0 -1
+    outer loop
+      vertex -10 -35 0
+      vertex 10 -25 0
+      vertex 10 -35 0
+    endloop
+  endfacet
+  facet normal -0 0 -1
+    outer loop
+      vertex 10 -25 0
+      vertex -10 -35 0
+      vertex -10 -25 0
+    endloop
+  endfacet
+endsolid OpenSCAD_Model