Merge pull request #59 from irockel/fix/Issue54_DetectUtf16

fix (#54): added detection for utf16

Author: irockel

tda/dependency-reduced-pom.xml

@@ -28,7 +28,7 @@
     <plugins>
       <plugin>
         <artifactId>maven-dependency-plugin</artifactId>
-        <version>3.6.0</version>
+        <version>3.10.0</version>
         <executions>
           <execution>
             <id>unpack-licenses</id>
@@ -41,15 +41,15 @@
                 <artifactItem>
                   <groupId>com.formdev</groupId>
                   <artifactId>flatlaf</artifactId>
-                  <version>3.3</version>
+                  <version>3.7.1</version>
                   <type>jar</type>
                   <includes>META-INF/LICENSE</includes>
                   <outputDirectory>${project.build.directory}/licenses/flatlaf</outputDirectory>
                 </artifactItem>
                 <artifactItem>
                   <groupId>com.formdev</groupId>
                   <artifactId>flatlaf-extras</artifactId>
-                  <version>3.3</version>
+                  <version>3.7.1</version>
                   <type>jar</type>
                   <includes>META-INF/LICENSE</includes>
                   <outputDirectory>${project.build.directory}/licenses/flatlaf-extras</outputDirectory>
@@ -69,7 +69,7 @@
       </plugin>
       <plugin>
         <artifactId>maven-antrun-plugin</artifactId>
-        <version>3.1.0</version>
+        <version>3.2.0</version>
         <executions>
           <execution>
             <id>rename-licenses</id>
@@ -89,7 +89,7 @@
       </plugin>
       <plugin>
         <artifactId>maven-compiler-plugin</artifactId>
-        <version>3.8.1</version>
+        <version>3.15.0</version>
         <configuration>
           <compilerArgs>
             <arg>--add-modules</arg>
@@ -100,7 +100,7 @@
       </plugin>
       <plugin>
         <artifactId>maven-surefire-plugin</artifactId>
-        <version>3.0.0-M5</version>
+        <version>3.5.5</version>
         <configuration>
           <systemPropertyVariables>
             <java.awt.headless>true</java.awt.headless>
@@ -109,7 +109,7 @@
       </plugin>
       <plugin>
         <artifactId>maven-jar-plugin</artifactId>
-        <version>3.2.0</version>
+        <version>3.5.0</version>
         <configuration>
           <archive>
             <manifest>
@@ -125,7 +125,7 @@
       </plugin>
       <plugin>
         <artifactId>maven-shade-plugin</artifactId>
-        <version>3.5.0</version>
+        <version>3.6.2</version>
         <executions>
           <execution>
             <phase>package</phase>
@@ -161,7 +161,7 @@
     <dependency>
       <groupId>org.junit.jupiter</groupId>
       <artifactId>junit-jupiter</artifactId>
-      <version>6.0.2</version>
+      <version>6.0.3</version>
       <scope>test</scope>
       <exclusions>
         <exclusion>

tda/src/main/java/de/grimmfrost/tda/parser/DumpParserFactory.java

@@ -24,10 +24,13 @@
 import de.grimmfrost.tda.utils.DateMatcher;
 import de.grimmfrost.tda.utils.LogManager;
 import de.grimmfrost.tda.utils.PrefManager;
+import java.io.BufferedInputStream;
 import java.io.BufferedReader;
 import java.io.IOException;
 import java.io.InputStream;
 import java.io.InputStreamReader;
+import java.nio.charset.Charset;
+import java.nio.charset.StandardCharsets;
 import java.util.Map;
 import java.util.logging.Level;
 import java.util.logging.Logger;
@@ -74,10 +77,12 @@ public DumpParser getDumpParserForLogfile(InputStream dumpFileStream, Map<String
         int readAheadLimit = PrefManager.get().getStreamResetBuffer();
         int lineCounter = 0;
         DumpParser currentDumpParser = null;
-            
+
         try {
-            bis = new BufferedReader(new InputStreamReader(dumpFileStream));
-            
+            BufferedInputStream bufferedStream = new BufferedInputStream(dumpFileStream);
+            Charset charset = detectCharset(bufferedStream);
+            bis = new BufferedReader(new InputStreamReader(bufferedStream, charset));
+
             // reset current dump parser
             DateMatcher dm = new DateMatcher();
             while (bis.ready() && (currentDumpParser == null)) {
@@ -121,5 +126,32 @@ public DumpParser getDumpParserForLogfile(InputStream dumpFileStream, Map<String
             LOGGER.log(Level.SEVERE, "IO error detecting parser for logfile", ex);
         }
         return currentDumpParser;
-    }    
+    }
+
+    private Charset detectCharset(BufferedInputStream bis) throws IOException {
+        bis.mark(4);
+        byte[] bom = new byte[4];
+        int read = bis.read(bom);
+        bis.reset();
+
+        if (read >= 2) {
+            if (bom[0] == (byte) 0xFF && bom[1] == (byte) 0xFE) {
+                return StandardCharsets.UTF_16LE;
+            } else if (bom[0] == (byte) 0xFE && bom[1] == (byte) 0xFF) {
+                return StandardCharsets.UTF_16BE;
+            }
+        }
+
+        // if no BOM, check for UTF-16LE/BE by looking for null bytes
+        // IBM VMs often don't have BOM but use UTF-16LE
+        if (read >= 4) {
+            if (bom[1] == 0 && bom[3] == 0) {
+                return StandardCharsets.UTF_16LE;
+            } else if (bom[0] == 0 && bom[2] == 0) {
+                return StandardCharsets.UTF_16BE;
+            }
+        }
+
+        return Charset.defaultCharset();
+    }
 }

tda/src/test/java/de/grimmfrost/tda/parser/DumpParserFactoryTest.java

@@ -82,4 +82,19 @@ public void testGetDumpParserForJSONLogfile() throws FileNotFoundException {
 
         assertInstanceOf(JCmdJSONParser.class, result);
     }
+
+    /**
+     * Test of getDumpParserForVersion method, of class de.grimmfrost.tda.DumpParserFactory.
+     */
+    @Test
+    public void testGetDumpParserForUTF16Logfile() throws FileNotFoundException {
+        InputStream dumpFileStream = new FileInputStream("src/test/resources/java21dump_utf16.log");
+        Map<String, Map<String, String>> threadStore = new HashMap<>();
+        DumpParserFactory instance = DumpParserFactory.get();
+
+        DumpParser result = instance.getDumpParserForLogfile(dumpFileStream, threadStore, false, 0);
+        assertNotNull(result);
+
+        assertInstanceOf(SunJDKParser.class, result);
+    }
 }