From c650b112eaa5c728f667d91566033c612f8a15cf Mon Sep 17 00:00:00 2001
From: theli <theli@6c8d7289-2bf4-0310-a012-ef5d649a1542>
Date: Sun, 27 Nov 2005 06:35:23 +0000
Subject: [PATCH] *) Bugfix for relative URL Bug in Crawler    See:
 http://www.yacy-forum.de/viewtopic.php?p=13266#13266

git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@1130 6c8d7289-2bf4-0310-a012-ef5d649a1542
---
 source/de/anomic/plasma/plasmaCrawlWorker.java | 9 ++++++++-
 1 file changed, 8 insertions(+), 1 deletion(-)

diff --git a/source/de/anomic/plasma/plasmaCrawlWorker.java b/source/de/anomic/plasma/plasmaCrawlWorker.java
index 14670d236..ccb1421cb 100644
--- a/source/de/anomic/plasma/plasmaCrawlWorker.java
+++ b/source/de/anomic/plasma/plasmaCrawlWorker.java
@@ -362,7 +362,13 @@ public final class plasmaCrawlWorker extends Thread {
 
                 // reserve cache entry
                 plasmaHTCache.Entry htCache = cacheManager.newEntry(requestDate, depth, url, name, requestHeader, res.status, res.responseHeader, initiator, profile);
-
+                if (!htCache.cacheFile.getCanonicalPath().startsWith(cacheManager.cachePath.getAbsolutePath())) {
+                    // if the response has not the right file type then reject file
+                    remote.close();
+                    log.logInfo("REJECTED URL " + url.toString() + " because of an invalid file path '" + htCache.cacheFile.getAbsolutePath() + "'.");
+                    return;                    
+                }
+                
                 // request has been placed and result has been returned. work off response
                 File cacheFile = cacheManager.getCachePath(url);
                 try {
@@ -386,6 +392,7 @@ public final class plasmaCrawlWorker extends Thread {
                         // if the response has not the right file type then reject file
                         remote.close();
                         log.logInfo("REJECTED WRONG MIME/EXT TYPE " + res.responseHeader.mime() + " for URL " + url.toString());
+                        return;
                     }
                     // enQueue new entry with response header
                     if (profile != null) {