mirror of
https://github.com/yacy/yacy_search_server.git
synced 2024-09-21 00:00:13 +02:00
BUGFIX for regular expression
git-svn-id: https://svn.berlios.de/svnroot/repos/yacy/trunk@1261 6c8d7289-2bf4-0310-a012-ef5d649a1542
This commit is contained in:
parent
106cd29981
commit
8eb4181737
|
@ -164,7 +164,7 @@ public class htmlFilterContentScraper extends htmlFilterAbstractScraper implemen
|
|||
|
||||
// (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;
|
||||
|
||||
|
||||
Pattern pathPattern = Pattern.compile("(/[^/\\.]+/)[.]{2}(?=/)|/\\.(?=/)|/(?=/)");
|
||||
Matcher matcher = pathPattern.matcher(path);
|
||||
while (matcher.find()) {
|
||||
|
@ -172,6 +172,10 @@ public class htmlFilterContentScraper extends htmlFilterAbstractScraper implemen
|
|||
matcher.reset(path);
|
||||
}
|
||||
|
||||
while (path.startsWith("/../")) {
|
||||
path = path.substring(3);
|
||||
}
|
||||
|
||||
if (defaultPort) return url.getProtocol() + "://" + url.getHost() + path;
|
||||
return url.getProtocol() + "://" + url.getHost() + ":" + url.getPort() + path;
|
||||
}
|
||||
|
|
Loading…
Reference in New Issue
Block a user