From 23a39698a7c32c45ee6cc3c69a565b4255d178e9 Mon Sep 17 00:00:00 2001 From: borg-0300 Date: Mon, 9 Jan 2006 19:07:11 +0000 Subject: [PATCH] last commit removed git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@1301 6c8d7289-2bf4-0310-a012-ef5d649a1542 --- source/de/anomic/htmlFilter/htmlFilterContentScraper.java | 5 +---- 1 file changed, 1 insertion(+), 4 deletions(-) diff --git a/source/de/anomic/htmlFilter/htmlFilterContentScraper.java b/source/de/anomic/htmlFilter/htmlFilterContentScraper.java index 06ea067d8..ffd66871e 100644 --- a/source/de/anomic/htmlFilter/htmlFilterContentScraper.java +++ b/source/de/anomic/htmlFilter/htmlFilterContentScraper.java @@ -160,7 +160,7 @@ public class htmlFilterContentScraper extends htmlFilterAbstractScraper implemen } else if (url.getProtocol().equals("https")) { if (url.getPort() < 0 || url.getPort() == 443) { defaultPort = true; } } - String path = url.getPath(); + String path = url.getFile(); // (this is different from previous normal forms where a '/' must not appear in root paths; here it must appear. Makes everything easier.) if (path.length() == 0 || path.charAt(0) != '/') path = "/" + path; @@ -172,9 +172,6 @@ public class htmlFilterContentScraper extends htmlFilterAbstractScraper implemen matcher.reset(path); } - String query = url.getQuery().replaceAll("[\"\\/:*?<>|]", "_"); - if (query != null) { path = path.concat("_").concat(query); } - if (defaultPort) return url.getProtocol() + "://" + url.getHost() + path; return url.getProtocol() + "://" + url.getHost() + ":" + url.getPort() + path; }