Fixed some HTML validation errors : Illegal character in query

Now encode space characters in URLs query part.
yacy · Sep 30, 2016 · 60df09f · 60df09f
1 parent a76a46a
commit 60df09f
Show file tree

Hide file tree

Showing 4 changed files with 28 additions and 6 deletions.
diff --git a/htroot/yacysearch.html b/htroot/yacysearch.html
@@ -3,7 +3,7 @@
 <head>
   <title>#[former]# - YaCy '#[clientname]#': Search Page</title>
   #%env/templates/metas.template%#
-  <link rel="alternate" type="application/rss+xml" title="Search for #[former]#" href="yacysearch.rss?query=#[former]#" />
+  <link rel="alternate" type="application/rss+xml" title="Search for #[former]#" href="yacysearch.rss?query=#[formerEncoded]#" />
   <link rel="search" type="application/opensearchdescription+xml" title="YaCy '#[clientname]#'" href="http://#[thisaddress]#/opensearchdescription.xml" />
   <link rel="stylesheet" type="text/css" media="screen" href="env/highslide.css" />
   <script type="text/javascript" src="js/ajax.js"></script>

diff --git a/htroot/yacysearch.java b/htroot/yacysearch.java
@@ -28,6 +28,9 @@
 // if the shell's current path is HTROOT
 
 import java.io.IOException;
+import java.io.UnsupportedEncodingException;
+import java.net.URLEncoder;
+import java.nio.charset.StandardCharsets;
 import java.util.ArrayList;
 import java.util.Collection;
 import java.util.ConcurrentModificationException;
@@ -852,7 +855,14 @@ public static serverObjects respond(
         }
         prop.put("focus", focus ? 1 : 0); // focus search field
         prop.put("searchagain", global ? "1" : "0");
-        prop.putHTML("former", originalquerystring.replaceAll(Segment.catchallString, "*"));
+        String former = originalquerystring.replaceAll(Segment.catchallString, "*");
+        prop.putHTML("former", former);
+        try {
+			prop.put("formerEncoded", URLEncoder.encode(former, StandardCharsets.UTF_8.name()));
+		} catch (UnsupportedEncodingException e) {
+			ConcurrentLog.warn("LOCAL_SEARCH", "Unsupported UTF-8 encoding!");
+			prop.put("formerEncoded", former);
+		}
         prop.put("count", itemsPerPage);
         prop.put("offset", startRecord);
         prop.put("resource", global ? "global" : "local");
@@ -877,4 +887,8 @@ public static serverObjects respond(
         // return rewrite properties
         return prop;
     }
+
+    public static void main(String args[]) throws UnsupportedEncodingException {
+    	System.out.println(URLEncoder.encode("aao?+ bbibu", StandardCharsets.UTF_8.name()));
+    }
 }
diff --git a/htroot/yacysearchtrailer.java b/htroot/yacysearchtrailer.java
@@ -110,7 +110,8 @@ public static serverObjects respond(final RequestHeader header, final serverObje
                 name = navigatorIterator.next();
                 count = theSearch.namespaceNavigator.get(name);
                 if (count == 0) break;
-                nav = "inurl%3A" + name;
+                /* We use ':' character here, it will be percent encoded later, in QueryParams.navurl() function*/
+                nav = "inurl:" + name;
                 if (!theSearch.query.modifier.toString().contains("inurl:"+name)) {
                     pos++;
                     prop.put("nav-namespace_element_" + i + "_on", 1);
@@ -317,7 +318,7 @@ public static serverObjects respond(final RequestHeader header, final serverObje
                 prop.put("nav-topics_element_" + i + "_on", 1);
                 prop.put(fileType, "nav-topics_element_" + i + "_modifier", name);
                 prop.put(fileType, "nav-topics_element_" + i + "_name", name);
-                prop.put(fileType, "nav-topics_element_" + i + "_url", QueryParams.navurl(fileType, 0, theSearch.query, name, false).toString());
+                prop.put("nav-topics_element_" + i + "_url", QueryParams.navurl(fileType, 0, theSearch.query, name, false).toString());
                 prop.put("nav-topics_element_" + i + "_count", count);
                 int fontsize = TOPWORDS_MINSIZE + (TOPWORDS_MAXSIZE - TOPWORDS_MINSIZE) * (count - mincount) / (1 + maxcount - mincount);
                 fontsize = Math.max(TOPWORDS_MINSIZE, fontsize - (name.length() - 5));

diff --git a/source/net/yacy/search/query/QueryParams.java b/source/net/yacy/search/query/QueryParams.java
@@ -26,6 +26,9 @@
 
 package net.yacy.search.query;
 
+import java.io.UnsupportedEncodingException;
+import java.net.URLEncoder;
+import java.nio.charset.StandardCharsets;
 import java.util.ArrayList;
 import java.util.Collection;
 import java.util.Date;
@@ -702,15 +705,19 @@ public static StringBuilder navurlBase(final RequestHeader.FileType ext, final Q
             if (!newModifier.isEmpty()) {
                 if (!theQuery.modifier.isEmpty()) sb.append("+" + theQuery.modifier.toString());
                 if (newModifierReplacesOld) {
-                    int nmpi = newModifier.indexOf("%3A");
+                    int nmpi = newModifier.indexOf(":");
                     if (nmpi > 0) {
                         String nmp = newModifier.substring(0, nmpi) + ":";
                         int i = sb.indexOf(nmp);
                         if (i > 0) sb = new StringBuilder(sb.substring(0, i).trim());
                         if (sb.charAt(sb.length() - 1) == '+') sb.setLength(sb.length() - 1);
                     }
                 }
-                sb.append("+" + newModifier);
+                try {
+                	sb.append("+" + URLEncoder.encode(newModifier, StandardCharsets.UTF_8.name()));
+                } catch (final UnsupportedEncodingException e) {
+                	sb.append("+" + newModifier);
+                }
             }
         }