From cfb69dfa28af542261562e0d5c7ba962f2419e32 Mon Sep 17 00:00:00 2001
From: =?UTF-8?q?Ant=C3=B3nio=20Lindo?= <alindo@keep.pt>
Date: Tue, 28 Nov 2023 09:09:36 +0000
Subject: [PATCH] Environment variable to disable Tika AutoDetectParser
 characters limit

---
 .../transformers/ToolkitStructure2ViewerStructure.java   | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/src/main/java/com/databasepreservation/common/transformers/ToolkitStructure2ViewerStructure.java b/src/main/java/com/databasepreservation/common/transformers/ToolkitStructure2ViewerStructure.java
index 8c1ade52..435b20a7 100644
--- a/src/main/java/com/databasepreservation/common/transformers/ToolkitStructure2ViewerStructure.java
+++ b/src/main/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();