Merge pull request #44 from levigo/feature/JF-1083_more_consistent_config

Feature/jf 1083 more consistent config

Author: welschsn

src/main/resources/magic-all.xml

@@ -35,6 +35,7 @@ public static void init() throws AnalyzerException {
   void testRtfEmbeddedFiles() throws IOException {
     for (File f : nullSafe(EMBEDDED_FILES_FOLDER.listFiles())) {
       final Map<String, Object> result = ANALYZER.analyze(f);
+      TestVariousTypes.printResult(result);
       assertNotNull(result, f + " could not be analyzed");
       assertEquals("text/rtf", result.get(MimeTypeAction.KEY), f + " is recognized as RTF");
       assertValidDetails(result); 
@@ -46,6 +47,7 @@ void testRtfEmbeddedFiles() throws IOException {
   void testNormalRtfFiles() throws IOException {
     for (File f : nullSafe(NORMAL_FOLDER.listFiles())) {
       final Map<String, Object> result = ANALYZER.analyze(f);
+      TestVariousTypes.printResult(result);
       assertNotNull(result, f + " could not be analyzed");
       assertEquals("text/rtf", result.get(MimeTypeAction.KEY), f + " is recognized as RTF");
       assertValidDetails(result); 

src/main/resources/magic.xml

@@ -61,10 +61,37 @@ void testEmptyStream() throws Exception {
     assertNotNull(results, "empty stream could not be analyzed");
     assertEquals("text/plain", results.get(MimeTypeAction.KEY));
     assertEquals("txt", results.get(ExtensionAction.KEY));
-    assertEquals("Binary data, ASCII Text Document", results.get(DescriptionAction.KEY));
+    assertEquals("ASCII Text Document", results.get(DescriptionAction.KEY));
   }
 
-  public static Stream<Arguments> dataProvider() {
+  public static Stream<Arguments> dataProviderVarious() {
+    return Stream.of(
+        arguments("/various_types/test.png", "image/png", "PNG image data, colormap,", "png"),
+        arguments("/various_types/File-PNG_8pbc_GRAY.png", "image/png", "PNG image data, grayscale,", "png"),
+        arguments("/various_types/File-PNG_8bpc_RGBA.png", "image/png", "PNG image data, \\b/color RGBA,", "png"),
+        arguments("/various_types/File-PNG_16bpc_RGBA.png", "image/png", "PNG image data, \\b/color RGBA,", "png"),
+        arguments("/various_types/sample-rtf-files-sample2.rtf", "text/rtf", "Rich Text Format data", "rtf")
+    );
+  }
+
+  @ParameterizedTest
+  @MethodSource("dataProviderVarious")
+  void testVariousTypesExplicitly(String resource, String expectedMimeType, String expectedDescription, String expectedExtension) throws Exception {
+    final URL url = getClass().getResource(resource);
+    assertNotNull(url);
+    final File file = new File(url.toURI());
+    final Map<String, Object> results = analyzer.analyze(file);
+    printResult(results);
+    assertNotNull(results, file + " could not be analyzed");
+    assertNotNull(results.get(MimeTypeAction.KEY), "mimeType missing");
+    assertEquals(expectedMimeType, results.get(MimeTypeAction.KEY), "wrong mimeType");
+    assertNotNull(results.get(DescriptionAction.KEY), "description missing");
+    assertEquals(expectedDescription, results.get(DescriptionAction.KEY), "wrong description");
+    assertNotNull(results.get(ExtensionAction.KEY), "could not be analyzed");
+    assertEquals(expectedExtension, results.get(ExtensionAction.KEY), "wrong extension");
+  }
+
+  public static Stream<Arguments> dataProviderXRechnung() {
     return Stream.of(
         arguments("/various_types/BASIC_Einfach.pdf", "application/pdf"),
         arguments("/various_types/EN16931_Einfach.pdf", "application/pdf"),
@@ -75,7 +102,7 @@ public static Stream<Arguments> dataProvider() {
   }
 
   @ParameterizedTest
-  @MethodSource("dataProvider")
+  @MethodSource("dataProviderXRechnung")
   void testXRechnung(String resource, String expectedMimeType) throws Exception {
     final URL url = getClass().getResource(resource);
     assertNotNull(url);
@@ -109,7 +136,7 @@ private void ensureXRechnungIsTrue(final Map<String, Object> results, final Stri
   }
 
 
-  private static void printResult(final Map<String, Object> results) {
+  public static void printResult(final Map<String, Object> results) {
     for (final Map.Entry<String, Object> e : results.entrySet()) {
       LOGGER.info("   {}={}", e.getKey(), e.getValue());
     }

src/test/java/TestRTFMatcher.java

@@ -0,0 +1,18 @@
+{\rtf1\mac\deff2 {\fonttbl{\f0\fswiss Chicago;}{\f2\froman New York;}{\f3\fswiss Geneva;}{\f4\fmodern Monaco;}{\f11\fnil Cairo;}{\f13\fnil Zapf Dingbats;}{\f14\fnil Bookman;}{\f15\fnil N Helvetica Narrow;}{\f16\fnil Palatino;}{\f18\fnil Zapf Chancery;}
+{\f20\froman Times;}{\f21\fswiss Helvetica;}{\f22\fmodern Courier;}{\f23\ftech Symbol;}{\f33\fnil Avant Garde;}{\f34\fnil New Century Schlbk;}{\f1297\fnil GoudyHundred;}{\f1602\fnil BlackChancery;}{\f2515\fnil MT Extra;}{\f4950\fnil TTYFont;}
+{\f11132\fnil InsigniaLQmono;}{\f11133\fnil InsigniaLQprop;}{\f32500\fnil VT320;}{\f32525\fnil VT100;}}{\colortbl\red0\green0\blue0;\red0\green0\blue255;\red0\green255\blue255;\red0\green255\blue0;\red255\green0\blue255;\red255\green0\blue0;
+\red255\green255\blue0;\red255\green255\blue255;}{\stylesheet{\s250\li720 \f21\fs20\ul \sbasedon0\snext0 heading 6;}{\s251\li720 \b\f21\fs20 \sbasedon0\snext0 heading 5;}{\s252\li360 \f21\ul \sbasedon0\snext0 heading 4;}{\s253\li360 \b\f21 
+\sbasedon0\snext0 heading 3;}{\s254\sb120 \b\f21 \sbasedon0\snext0 heading 2;}{\s255\sb240 \b\f21\ul \sbasedon0\snext0 heading 1;}{\f21 \sbasedon222\snext0 Normal;}{\s2 \b\f21\cf1 \sbasedon0\snext2 Anchor;}}{\info{\author Chuck Shotton}}
+\margl720\margr720\ftnbj\fracwidth \sectd \sbknone\linemod0\linex0\cols1\endnhere \pard\plain \s255\sb240 \b\f21\ul Heading 1\par 
+\pard\plain \f21 This is the first normal paragraph!\par 
+\pard\plain \s254\sb120 \b\f21 Heading 2\par 
+\pard\plain \s253\li360 \b\f21 Heading 3\par 
+\pard\plain \s252\li360 \f21\ul Heading 4\par 
+\pard\plain \s251\li720 \b\f21\fs20 Heading 5\par 
+\pard\plain \s250\li720 \f21\fs20\ul Heading 6\par 
+\pard\plain \f21 This is a chunk of normal text.\par 
+This is a chunk of normal text with specials, &, <, and >.\par 
+This is a second paragraph.\par 
+This is text with embedded {\b bold}, {\i italic}, and {\ul underline} styles.\par 
+Here is the {\cf1 anchor} style. And here is the {\cf5 Image} style.\par 
+}