Merge pull request #373 from keeps/alindo-dev

Environment variable to disable Tika AutoDetectParser characters limit
keeps · Nov 28, 2023 · 4015b89 · 4015b89
2 parents f11f282 + cfb69df
commit 4015b89
Showing 1 changed file with 8 additions and 1 deletion.
diff --git a/...n/java/com/databasepreservation/common/transformers/ToolkitStructure2ViewerStructure.java b/...n/java/com/databasepreservation/common/transformers/ToolkitStructure2ViewerStructure.java
@@ -922,7 +922,14 @@ private static void detectMimeType(ViewerRow row, ViewerCell cell, String databa
           AutoDetectParser parser = new AutoDetectParser();
           Metadata metadata = new Metadata();
 
-          parser.parse(inputStream, new BodyContentHandler(), metadata, new ParseContext());
+          Boolean autoDetectParserNoLimit = ViewerFactory.getEnvBoolean("AUTO_DETECT_PARSER_NO_LIMIT", false);
+
+          if (autoDetectParserNoLimit) {
+            parser.parse(inputStream, new BodyContentHandler(-1), metadata, new ParseContext());
+          } else {
+            parser.parse(inputStream, new BodyContentHandler(), metadata, new ParseContext());
+          }
+
           mimeType = metadata.get("Content-Type");
           fileExtension = MimeTypes.getDefaultMimeTypes().forName(mimeType).getExtension();