TIKA-4274: improve ExtractReaderException and logging

apache · Jul 7, 2024 · 1039898 · 1039898
1 parent 5167832
commit 1039898
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 3 deletions.
diff --git a/tika-eval/tika-eval-app/src/main/java/org/apache/tika/eval/app/io/ExtractReader.java b/tika-eval/tika-eval-app/src/main/java/org/apache/tika/eval/app/io/ExtractReader.java
@@ -111,17 +111,21 @@ public List<Metadata> loadExtract(Path extractFile) throws ExtractReaderExceptio
         try {
             length = Files.size(extractFile);
         } catch (IOException e) {
-            throw new ExtractReaderException(ExtractReaderException.TYPE.IO_EXCEPTION);
+            throw new ExtractReaderException(ExtractReaderException.TYPE.IO_EXCEPTION, e);
         }
 
         if (length == 0L) {
             throw new ExtractReaderException(ExtractReaderException.TYPE.ZERO_BYTE_EXTRACT_FILE);
         }
 
         if (minExtractLength > IGNORE_LENGTH && length < minExtractLength) {
+            LOG.info("minExtractLength {} > IGNORE_LENGTH {} and length {} < minExtractLength {}", 
+                    minExtractLength, IGNORE_LENGTH, length, minExtractLength);
             throw new ExtractReaderException(ExtractReaderException.TYPE.EXTRACT_FILE_TOO_SHORT);
         }
         if (maxExtractLength > IGNORE_LENGTH && length > maxExtractLength) {
+            LOG.info("maxExtractLength {} > IGNORE_LENGTH {} and length {} > maxExtractLength {}", 
+                    maxExtractLength, IGNORE_LENGTH, length, maxExtractLength);
             throw new ExtractReaderException(ExtractReaderException.TYPE.EXTRACT_FILE_TOO_LONG);
         }
 
@@ -148,7 +152,7 @@ public List<Metadata> loadExtract(Path extractFile) throws ExtractReaderExceptio
             }
             reader = new BufferedReader(new InputStreamReader(is, StandardCharsets.UTF_8));
         } catch (IOException e) {
-            throw new ExtractReaderException(ExtractReaderException.TYPE.IO_EXCEPTION);
+            throw new ExtractReaderException(ExtractReaderException.TYPE.IO_EXCEPTION, e);
         }
 
         try {
@@ -177,7 +181,7 @@ public List<Metadata> loadExtract(Path extractFile) throws ExtractReaderExceptio
                 metadataList = generateListFromTextFile(reader, fileSuffixes);
             }
         } catch (IOException e) {
-            throw new ExtractReaderException(ExtractReaderException.TYPE.IO_EXCEPTION);
+            throw new ExtractReaderException(ExtractReaderException.TYPE.IO_EXCEPTION, e);
         } finally {
             IOUtils.closeQuietly(reader);
             IOUtils.closeQuietly(is);

diff --git a/...-eval/tika-eval-app/src/main/java/org/apache/tika/eval/app/io/ExtractReaderException.java b/...-eval/tika-eval-app/src/main/java/org/apache/tika/eval/app/io/ExtractReaderException.java
@@ -26,6 +26,12 @@ public class ExtractReaderException extends IOException {
     private final TYPE type;
 
     public ExtractReaderException(TYPE exceptionType) {
+        super(exceptionType.toString());
+        this.type = exceptionType;
+    }
+
+    public ExtractReaderException(TYPE exceptionType, Throwable t) {
+        super(exceptionType.toString(), t);
         this.type = exceptionType;
     }