TIKA-4303: Handle OneNotePropertyEnum.CachedTitleString as RichEditTe…

…xtUnicode TIKA-4303: Handle OneNotePropertyEnum.CachedTitleString as RichEditTextUnicode
apache · Jan 25, 2025 · f59f6c5 · f59f6c5
1 parent bc6fd9d
commit f59f6c5
Show file tree

Hide file tree

Showing 2 changed files with 10 additions and 0 deletions.
diff --git a/...soft-module/src/test/java/org/apache/tika/parser/microsoft/onenote/OneNoteParserTest.java b/...soft-module/src/test/java/org/apache/tika/parser/microsoft/onenote/OneNoteParserTest.java
@@ -288,4 +288,14 @@ public void testDupeText() throws Exception {
 
         assertEquals(1, StringUtils.countMatches(txt, "Sunday morning"));
     }
+
+    /**
+     * TIKA-4303 - test extract Chinese
+     */
+    @Test
+    public void testExtractChinese() throws Exception {
+        Metadata metadata = new Metadata();
+        XMLResult xml = getXML("test-tika-4303-Chinese-notes.one", metadata);
+        assertContains("<p>中文标题</p>", xml.xml);
+    }
 }
diff --git a/...arser-microsoft-module/src/test/resources/test-documents/test-tika-4303-Chinese-notes.one b/...arser-microsoft-module/src/test/resources/test-documents/test-tika-4303-Chinese-notes.one