From 039162fbf0eca808afd350d360c3bcfe62dc4195 Mon Sep 17 00:00:00 2001 From: reger Date: Mon, 22 May 2017 01:34:08 +0200 Subject: [PATCH] Change warc importer to use defaultsurrogate-crawl profile, as reported by LA_FORGE http://forum.yacy-websuche.de/viewtopic.php?f=5&t=5990 and analysed by @luccioman (see comment https://github.com/yacy/yacy_search_server/commit/510f11d3745e14841420781376b733fd248d51f3) it creates conflict using a other crawlprofile without setting originator. --- source/net/yacy/document/importer/WarcImporter.java | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/source/net/yacy/document/importer/WarcImporter.java b/source/net/yacy/document/importer/WarcImporter.java index 5ad4582b3..3022ea7c4 100644 --- a/source/net/yacy/document/importer/WarcImporter.java +++ b/source/net/yacy/document/importer/WarcImporter.java @@ -150,15 +150,15 @@ public class WarcImporter extends Thread implements Importer { requestHeader.referer() == null ? null : requestHeader.referer().hash(), "warc", responseHeader.lastModified(), - Switchboard.getSwitchboard().crawler.defaultRemoteProfile.handle(), // use remote profile (to index text & media, without writing to cache + Switchboard.getSwitchboard().crawler.defaultSurrogateProfile.handle(), 0, - Switchboard.getSwitchboard().crawler.defaultRemoteProfile.timezoneOffset()); + Switchboard.getSwitchboard().crawler.defaultSurrogateProfile.timezoneOffset()); final Response response = new Response( request, requestHeader, responseHeader, - Switchboard.getSwitchboard().crawler.defaultRemoteProfile, + Switchboard.getSwitchboard().crawler.defaultSurrogateProfile, false, content );