Marown/yacy_search_server
1
0
Fork 0
mirror of https://github.com/yacy/yacy_search_server.git synced 2024-09-19 00:01:41 +02:00
Code Issues Actions Packages Projects Releases Wiki Activity

Fixed remaining blacklist entries improper decoding of '+' character

Browse Source 
In the blacklist cleaner and import/export administration pages.
This commit is contained in:
luccioman 2018-10-08 16:50:02 +02:00
parent 61c337f29a
commit 753bda1409
2 changed files with 14 additions and 17 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

23
htroot/BlacklistCleaner_p.java
View File

@ -45,6 +45,7 @@ import net.yacy.cora.document.id.Punycode.PunycodeException;
import net.yacy.cora.protocol.RequestHeader;
import net.yacy.cora.util.ConcurrentLog;
import net.yacy.data.ListManager;
import net.yacy.document.parser.html.CharacterCoding;
import net.yacy.kelondro.util.FileUtils;
import net.yacy.repository.Blacklist;
import net.yacy.repository.BlacklistHostAndPath;
@ -123,9 +124,7 @@ public class BlacklistCleaner_p {
}
// list illegal entries
final Map<String, BlacklistError> illegalEntries = getIllegalEntries(
blacklistToUse, Switchboard.urlBlacklist,
allowRegex);
final Map<String, BlacklistError> illegalEntries = getIllegalEntries(blacklistToUse, allowRegex);
prop.put(RESULTS + "blList", blacklistToUse);
prop.put(RESULTS + "entries", illegalEntries.size());
prop.putHTML(RESULTS + "blEngine", Blacklist.getEngineInfo());
@ -141,8 +140,9 @@ public class BlacklistCleaner_p {
key = entry.getKey();
prop.put(RESULTS + DISABLED + ENTRIES + i + "_error",
entry.getValue().getLong());
prop.putHTML(RESULTS + DISABLED + ENTRIES + i
+ "_entry", key);
/* We do not use here putHTML as we don't want '+' characters to be decoded as spaces by application/x-www-form-urlencoded encoding */
prop.put(RESULTS + DISABLED + ENTRIES + i
+ "_entry", CharacterCoding.unicode2html(key, true));
i++;
}
}
@ -284,8 +284,6 @@ public class BlacklistCleaner_p {
*
* @param blacklistToUse
* The blacklist to be checked.
* @param blEngine
* The blacklist engine which is used to check
* @param allowRegex
* Set to true to allow regular expressions in host part of
* blacklist entry.
@ -294,18 +292,17 @@ public class BlacklistCleaner_p {
* error code as value.
*/
private static Map<String, BlacklistError> getIllegalEntries(
final String blacklistToUse, final Blacklist blEngine,
final String blacklistToUse,
final boolean allowRegex) {
final Map<String, BlacklistError> illegalEntries = new HashMap<String, BlacklistError>();
final Set<String> legalEntries = new HashSet<String>();
final Map<String, BlacklistError> illegalEntries = new HashMap<>();
final Set<String> legalEntries = new HashSet<>();
final List<String> list = FileUtils.getListArray(new File(
ListManager.listsPath, blacklistToUse));
final Map<String, String> properties = new HashMap<String, String>();
final Map<String, String> properties = new HashMap<>();
properties.put("allowRegex", String.valueOf(allowRegex));
BlacklistError err = BlacklistError.NO_ERROR;
BlacklistError err;
for (String element : list) {
element = element.trim();

8
htroot/sharedBlacklist_p.java
View File

@ -285,9 +285,9 @@ public class sharedBlacklist_p {
while (otherBlacklist.hasNext()) {
final String tmp = otherBlacklist.next();
if( !Blacklist.contains(tmp) && (!tmp.equals("")) ){
//newBlacklist.add(tmp);
prop.put("page_urllist_" + count + "_dark", count % 2 == 0 ? "0" : "1");
prop.putHTML("page_urllist_" + count + "_url", tmp);
/* We do not use here putHTML as we don't want '+' characters to be decoded as spaces by application/x-www-form-urlencoded encoding */
prop.put("page_urllist_" + count + "_url", CharacterCoding.unicode2html(tmp, true));
prop.put("page_urllist_" + count + "_count", count);
count++;
}
@ -310,9 +310,9 @@ public class sharedBlacklist_p {
for (final String element : sortedlist) {
final String tmp = element;
if(!tmp.equals("")){
//newBlacklist.add(tmp);
prop.put("page_urllist_" + count + "_dark", count % 2 == 0 ? "0" : "1");
prop.putHTML("page_urllist_" + count + "_url", tmp);
/* We do not use here putHTML as we don't want '+' characters to be decoded as spaces by application/x-www-form-urlencoded encoding */
prop.put("page_urllist_" + count + "_url", CharacterCoding.unicode2html(tmp, true));
prop.put("page_urllist_" + count + "_count", count);
count++;
}