Ufal/Preview issues (#764)

* Ensure the content preview doesn't overload maximum length of the column in the database. And encode the input stream into UTF-8.

* Do not store HTML content in the database because it could be longer than the limit of the database column.

dspace-server-webapp/src/main/java/org/dspace/app/rest/repository/MetadataBitstreamRestRepository.java

@@ -12,6 +12,7 @@
 import java.io.InputStream;
 import java.io.InputStreamReader;
 import java.io.UnsupportedEncodingException;
+import java.nio.charset.StandardCharsets;
 import java.nio.file.FileSystem;
 import java.nio.file.FileSystems;
 import java.nio.file.Files;
@@ -82,6 +83,10 @@
 public class MetadataBitstreamRestRepository extends DSpaceRestRepository<MetadataBitstreamWrapperRest, Integer> {
     private static Logger log = org.apache.logging.log4j.LogManager.getLogger(MetadataBitstreamRestRepository.class);
 
+    // This constant is used to limit the length of the preview content stored in the database to prevent
+    // the database from being overloaded with large amounts of data.
+    private static final int MAX_PREVIEW_COUNT_LENGTH = 2000;
+
     @Autowired
     HandleService handleService;
 
@@ -170,8 +175,12 @@ public Page<MetadataBitstreamWrapperRest> findByHandle(@Parameter(value = "handl
                     // Generate new content if we didn't find any
                     if (prContents.isEmpty()) {
                         fileInfos = getFilePreviewContent(context, bitstream, fileInfos);
-                        for (FileInfo fi : fileInfos) {
-                            createPreviewContent(context, bitstream, fi);
+                        // Do not store HTML content in the database because it could be longer than the limit
+                        // of the database column
+                        if (!StringUtils.equals("text/html", bitstream.getFormat(context).getMIMEType())) {
+                            for (FileInfo fi : fileInfos) {
+                                createPreviewContent(context, bitstream, fi);
+                            }
                         }
                     } else {
                         for (PreviewContent pc : prContents) {
@@ -311,8 +320,11 @@ private List<FileInfo> processInputStreamToFilePreview(Context context, Bitstrea
                                                            List<FileInfo> fileInfos, InputStream inputStream)
             throws IOException, SQLException, ParserConfigurationException, SAXException, ArchiveException {
         String bitstreamMimeType = bitstream.getFormat(context).getMIMEType();
-        if (bitstreamMimeType.equals("text/plain") || bitstreamMimeType.equals("text/html")) {
-            String data = getFileContent(inputStream);
+        if (bitstreamMimeType.equals("text/plain")) {
+            String data = getFileContent(inputStream, true);
+            fileInfos.add(new FileInfo(data, false));
+        } else if (bitstreamMimeType.equals("text/html")) {
+            String data = getFileContent(inputStream, false);
             fileInfos.add(new FileInfo(data, false));
         } else {
             String data = "";
@@ -462,17 +474,44 @@ public String extractFile(InputStream inputStream, String fileType) {
      * @return content of the inputStream as a String
      * @throws IOException
      */
-    public String getFileContent(InputStream inputStream) throws IOException {
+    public String getFileContent(InputStream inputStream, boolean cutResult) throws IOException {
         StringBuilder content = new StringBuilder();
-        BufferedReader reader = new BufferedReader(new InputStreamReader(inputStream));
-
-        String line;
-        while ((line = reader.readLine()) != null) {
-            content.append(line).append("\n");
+        // Generate the preview content in the UTF-8 encoding
+        BufferedReader reader = new BufferedReader(new InputStreamReader(inputStream, StandardCharsets.UTF_8));
+        try {
+            String line;
+            while ((line = reader.readLine()) != null) {
+                content.append(line).append("\n");
+            }
+        } catch (UnsupportedEncodingException e) {
+            log.error("UnsupportedEncodingException during creating the preview content because: ", e);
+        } catch (IOException e) {
+            log.error("IOException during creating the preview content because: ", e);
         }
 
         reader.close();
-        return content.toString();
+        return cutResult ? ensureMaxLength(content.toString()) : content.toString();
+    }
+
+    /**
+     * Trims the input string to ensure it does not exceed the maximum length for the database column.
+     * @param input The original string to be trimmed.
+     * @return A string that is truncated to the maximum length if necessary.
+     */
+    private static String ensureMaxLength(String input) {
+        if (input == null) {
+            return null;
+        }
+
+        // Check if the input string exceeds the maximum preview length
+        if (input.length() > MAX_PREVIEW_COUNT_LENGTH) {
+            // Truncate the string and append " . . ."
+            int previewLength = MAX_PREVIEW_COUNT_LENGTH - 6; // Subtract length of " . . ."
+            return input.substring(0, previewLength) + " . . .";
+        } else {
+            // Return the input string as is if it's within the preview length
+            return input;
+        }
     }
 
     /**