diff --git a/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractor.java b/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractor.java
index 20f3dca0..b757d5da 100644
--- a/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractor.java
+++ b/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractor.java
@@ -21,7 +21,6 @@
 
 import org.apache.poi.hpbf.extractor.PublisherTextExtractor;
 import org.codelibs.fess.crawler.entity.ExtractData;
-import org.codelibs.fess.crawler.exception.CrawlerSystemException;
 import org.codelibs.fess.crawler.exception.ExtractException;
 
 /**
@@ -41,21 +40,23 @@ public MsPublisherExtractor() {
 
     /**
      * Extracts text from the Publisher input stream.
+     * <p>
+     * The {@link PublisherTextExtractor} is wrapped in a try-with-resources
+     * block so that the underlying {@code POIFSFileSystem} (and therefore the
+     * provided input stream) is always closed even when extraction fails
+     * partway through.
+     * </p>
      * @param in The input stream.
      * @param params The parameters.
      * @return The extracted data.
      */
     @Override
     public ExtractData getText(final InputStream in, final Map<String, String> params) {
-        if (in == null) {
-            throw new CrawlerSystemException("Microsoft Publisher input stream is null. Cannot extract text from null input.");
-        }
-        try {
-            @SuppressWarnings("resource")
-            final PublisherTextExtractor publisherTextExtractor = new PublisherTextExtractor(in);
+        validateInputStream(in);
+        try (PublisherTextExtractor publisherTextExtractor = new PublisherTextExtractor(in)) {
             return new ExtractData(publisherTextExtractor.getText());
         } catch (final IOException e) {
-            throw new ExtractException(e);
+            throw new ExtractException("Failed to extract text from Publisher document.", e);
         }
     }
 
diff --git a/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractor.java b/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractor.java
index b6c2c6b6..d29a72a5 100644
--- a/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractor.java
+++ b/fess-crawler/src/main/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractor.java
@@ -21,7 +21,6 @@
 
 import org.apache.poi.hdgf.extractor.VisioTextExtractor;
 import org.codelibs.fess.crawler.entity.ExtractData;
-import org.codelibs.fess.crawler.exception.CrawlerSystemException;
 import org.codelibs.fess.crawler.exception.ExtractException;
 
 /**
@@ -41,21 +40,23 @@ public MsVisioExtractor() {
 
     /**
      * Extracts text from the Visio input stream.
+     * <p>
+     * The {@link VisioTextExtractor} is wrapped in a try-with-resources block
+     * so that the underlying {@code POIFSFileSystem} (and therefore the
+     * provided input stream) is always closed even when extraction fails
+     * partway through.
+     * </p>
      * @param in The input stream.
      * @param params The parameters.
      * @return The extracted data.
      */
     @Override
     public ExtractData getText(final InputStream in, final Map<String, String> params) {
-        if (in == null) {
-            throw new CrawlerSystemException("Microsoft Visio input stream is null. Cannot extract text from null input.");
-        }
-        try {
-            @SuppressWarnings("resource")
-            final VisioTextExtractor visioTextExtractor = new VisioTextExtractor(in);
+        validateInputStream(in);
+        try (VisioTextExtractor visioTextExtractor = new VisioTextExtractor(in)) {
             return new ExtractData(visioTextExtractor.getText());
         } catch (final IOException e) {
-            throw new ExtractException(e);
+            throw new ExtractException("Failed to extract text from Visio document.", e);
         }
     }
 
diff --git a/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractorTest.java b/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractorTest.java
index 738ae3ba..558f35dc 100644
--- a/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractorTest.java
+++ b/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsPublisherExtractorTest.java
@@ -16,7 +16,10 @@
 package org.codelibs.fess.crawler.extractor.impl;
 
 import java.io.ByteArrayInputStream;
+import java.io.FilterInputStream;
+import java.io.IOException;
 import java.io.InputStream;
+import java.util.Random;
 
 import org.codelibs.fess.crawler.container.StandardCrawlerContainer;
 import org.codelibs.fess.crawler.exception.CrawlerSystemException;
@@ -40,7 +43,7 @@ protected void setUp(final TestInfo testInfo) throws Exception {
     }
 
     @Test
-    public void test_getText_null() {
+    public void test_nullInput_throwsCrawlerSystemException() {
         try {
             msPublisherExtractor.getText(null, null);
             fail();
@@ -50,8 +53,24 @@ public void test_getText_null() {
     }
 
     @Test
-    public void test_getText_invalidData() {
-        // Test with invalid (non-Publisher) data
+    public void test_corruptedInput_throwsExtractException() {
+        // Random bytes are not a valid Publisher document; the extractor
+        // must surface this as an ExtractException rather than leaking the
+        // underlying POI resources.
+        final byte[] randomBytes = new byte[4096];
+        new Random(42L).nextBytes(randomBytes);
+        final InputStream in = new ByteArrayInputStream(randomBytes);
+        try {
+            msPublisherExtractor.getText(in, null);
+            fail();
+        } catch (final ExtractException e) {
+            // Expected - random bytes are not a valid Publisher file
+        }
+    }
+
+    @Test
+    public void test_textInput_throwsExtractException() {
+        // Plain text bytes are not a valid Publisher document.
         final InputStream in = new ByteArrayInputStream("This is not a valid Publisher file".getBytes());
         try {
             msPublisherExtractor.getText(in, null);
@@ -62,7 +81,7 @@ public void test_getText_invalidData() {
     }
 
     @Test
-    public void test_getText_emptyStream() {
+    public void test_emptyInput_throwsExtractException() {
         final InputStream in = new ByteArrayInputStream(new byte[0]);
         try {
             msPublisherExtractor.getText(in, null);
@@ -72,10 +91,47 @@ public void test_getText_emptyStream() {
         }
     }
 
+    @Test
+    public void test_corruptedInput_closesUnderlyingStream() throws IOException {
+        // Verify that even when extraction fails, the underlying input stream
+        // is closed by the try-with-resources block (no resource leak).
+        final byte[] randomBytes = new byte[4096];
+        new Random(7L).nextBytes(randomBytes);
+        final CloseTrackingInputStream tracking = new CloseTrackingInputStream(new ByteArrayInputStream(randomBytes));
+        try {
+            msPublisherExtractor.getText(tracking, null);
+            fail();
+        } catch (final ExtractException e) {
+            // Expected
+        }
+        assertTrue(tracking.isClosed());
+    }
+
     @Test
     public void test_constructor() {
         // Verify the extractor can be instantiated
         final MsPublisherExtractor extractor = new MsPublisherExtractor();
         assertNotNull(extractor);
     }
+
+    /**
+     * Test helper that records whether {@link #close()} has been invoked.
+     */
+    private static final class CloseTrackingInputStream extends FilterInputStream {
+        private boolean closed;
+
+        CloseTrackingInputStream(final InputStream in) {
+            super(in);
+        }
+
+        @Override
+        public void close() throws IOException {
+            closed = true;
+            super.close();
+        }
+
+        boolean isClosed() {
+            return closed;
+        }
+    }
 }
diff --git a/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractorTest.java b/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractorTest.java
index dbc6cf70..38993c99 100644
--- a/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractorTest.java
+++ b/fess-crawler/src/test/java/org/codelibs/fess/crawler/extractor/impl/MsVisioExtractorTest.java
@@ -16,7 +16,10 @@
 package org.codelibs.fess.crawler.extractor.impl;
 
 import java.io.ByteArrayInputStream;
+import java.io.FilterInputStream;
+import java.io.IOException;
 import java.io.InputStream;
+import java.util.Random;
 
 import org.codelibs.fess.crawler.container.StandardCrawlerContainer;
 import org.codelibs.fess.crawler.exception.CrawlerSystemException;
@@ -40,7 +43,7 @@ protected void setUp(final TestInfo testInfo) throws Exception {
     }
 
     @Test
-    public void test_getText_null() {
+    public void test_nullInput_throwsCrawlerSystemException() {
         try {
             msVisioExtractor.getText(null, null);
             fail();
@@ -50,8 +53,24 @@ public void test_getText_null() {
     }
 
     @Test
-    public void test_getText_invalidData() {
-        // Test with invalid (non-Visio) data
+    public void test_corruptedInput_throwsExtractException() {
+        // Random bytes are not a valid Visio document; the extractor must
+        // surface this as an ExtractException rather than leaking the
+        // underlying POI resources.
+        final byte[] randomBytes = new byte[4096];
+        new Random(42L).nextBytes(randomBytes);
+        final InputStream in = new ByteArrayInputStream(randomBytes);
+        try {
+            msVisioExtractor.getText(in, null);
+            fail();
+        } catch (final ExtractException e) {
+            // Expected - random bytes are not a valid Visio file
+        }
+    }
+
+    @Test
+    public void test_textInput_throwsExtractException() {
+        // Plain text bytes are not a valid Visio document.
         final InputStream in = new ByteArrayInputStream("This is not a valid Visio file".getBytes());
         try {
             msVisioExtractor.getText(in, null);
@@ -62,7 +81,7 @@ public void test_getText_invalidData() {
     }
 
     @Test
-    public void test_getText_emptyStream() {
+    public void test_emptyInput_throwsExtractException() {
         final InputStream in = new ByteArrayInputStream(new byte[0]);
         try {
             msVisioExtractor.getText(in, null);
@@ -72,10 +91,47 @@ public void test_getText_emptyStream() {
         }
     }
 
+    @Test
+    public void test_corruptedInput_closesUnderlyingStream() throws IOException {
+        // Verify that even when extraction fails, the underlying input stream
+        // is closed by the try-with-resources block (no resource leak).
+        final byte[] randomBytes = new byte[4096];
+        new Random(7L).nextBytes(randomBytes);
+        final CloseTrackingInputStream tracking = new CloseTrackingInputStream(new ByteArrayInputStream(randomBytes));
+        try {
+            msVisioExtractor.getText(tracking, null);
+            fail();
+        } catch (final ExtractException e) {
+            // Expected
+        }
+        assertTrue(tracking.isClosed());
+    }
+
     @Test
     public void test_constructor() {
         // Verify the extractor can be instantiated
         final MsVisioExtractor extractor = new MsVisioExtractor();
         assertNotNull(extractor);
     }
+
+    /**
+     * Test helper that records whether {@link #close()} has been invoked.
+     */
+    private static final class CloseTrackingInputStream extends FilterInputStream {
+        private boolean closed;
+
+        CloseTrackingInputStream(final InputStream in) {
+            super(in);
+        }
+
+        @Override
+        public void close() throws IOException {
+            closed = true;
+            super.close();
+        }
+
+        boolean isClosed() {
+            return closed;
+        }
+    }
 }