From a3d651ba460c321bab34197e3d9ed3f85c45ed42 Mon Sep 17 00:00:00 2001
From: Alex Osborne <aosborne@nla.gov.au>
Date: Mon, 13 Dec 2021 15:41:38 +0900
Subject: [PATCH] FetchHTTP: Handle null characters in the Content-Length
 header

If a null character appears in the Content-Length header Firefox and
Chrome seem to ignore the rest of the value so let's do the same.
Let's also handle the exception when parsing the content-length fails.

Fixes #449
---
 .../java/org/archive/modules/fetcher/FetchHTTP.java | 13 +++++++++++--
 1 file changed, 11 insertions(+), 2 deletions(-)

diff --git a/modules/src/main/java/org/archive/modules/fetcher/FetchHTTP.java b/modules/src/main/java/org/archive/modules/fetcher/FetchHTTP.java
index 45b2bf72..a089e678 100644
--- a/modules/src/main/java/org/archive/modules/fetcher/FetchHTTP.java
+++ b/modules/src/main/java/org/archive/modules/fetcher/FetchHTTP.java
@@ -685,8 +685,17 @@ public class FetchHTTP extends Processor implements Lifecycle {
         
         long contentLength = -1l;
         Header h = response.getLastHeader("content-length");
-        if (h != null && h.getValue().trim().length()>0) {
-            contentLength = Long.parseLong(h.getValue());
+        if (h != null) {
+            // browsers ignore everything after a null character and some buggy web servers rely on this
+            String contentLengthHeader = StringUtils.substringBefore(h.getValue(), "\0").trim();
+            if (!contentLengthHeader.isEmpty()) {
+                try {
+                    contentLength = Long.parseLong(contentLengthHeader);
+                } catch (NumberFormatException e) {
+                    cleanup(curi, e, "invalid content-length header", S_CONNECT_LOST);
+                    return;
+                }
+            }
         }
         try {
             if (!req.request.isAborted()) {