svn commit: r1912568 - in /manifoldcf/trunk: ./ connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/ connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler...

Wed, 27 Sep 2023 21:20:35 -0700 

Author: kwright
Date: Thu Sep 28 04:20:30 2023
New Revision: 1912568

URL: http://svn.apache.org/viewvc?rev=1912568&view=rev
Log:
Add pr-149.

Modified:
    manifoldcf/trunk/CHANGES.txt
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConfig.java
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConnector.java
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_en_US.properties
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_es_ES.properties
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_fr_FR.properties
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_ja_JP.properties
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_zh_CN.properties
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/editSpecification_Inclusions.html.vm
    
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/viewSpecification.html.vm

Modified: manifoldcf/trunk/CHANGES.txt
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/CHANGES.txt?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- manifoldcf/trunk/CHANGES.txt (original)
+++ manifoldcf/trunk/CHANGES.txt Thu Sep 28 04:20:30 2023
@@ -3,6 +3,16 @@ $Id$
 
 ======================= 2.26-dev =====================
 
+pr-149: Add Webconnector feature:
+The "Force the inclusion of redirection” options allows you to include hosts 
redirected from original seeds. You might want to use this option if the site 
you are crawling is subject to redirections. Note that it is not required if 
the previous option is not checked. Here are the possible behaviors:
+
+   - If the user checks the “Include only hosts”, but not the “Force the 
inclusion” option, then the redirected files will be filtered if their new 
URL doesn’t match the seed.
+   - If the user checks the Include only hosts, and checks the Force the 
inclusion option, then when the job finds a url that is not in the same domain, 
it is dropped EXCEPT if the url is originated by a 301 or 302 redirection in 
the document queue.
+   - If the user does NOT check the include only hosts, but checks the Force 
the inclusion option, then the job will crawl any url found, even if it is 
originated by a 301 or 302 redirection.
+   - If the user does not check anything, then the behavior is the same as the 
previous case.
+
+   If the admin checks the second option AND if the first option is checked, 
then the job will check any host added in the Set. If a host is subject to 
redirection, then we add the destination URL in the Set.
+(Emeric Bernet-Rollande)
 
 ======================= Release 2.25 =====================
 

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConfig.java
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConfig.java?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConfig.java
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConfig.java
 Thu Sep 28 04:20:30 2023
@@ -147,6 +147,10 @@ public class WebcrawlerConfig
 
   /** Limit to seeds.  When value attribute is true, only seed domains will be 
permitted. */
   public static final String NODE_LIMITTOSEEDS = "limittoseeds";
+
+  /** Force the inclusion of redirections.  When value attribute is true, 
redirected URL will be included. */
+  public static final String NODE_FORCEINCLUSION = "forceinclusionofredirects";
+
   /** Canonicalization rule.  Attributes are regexp, description, reorder, 
   *javasessionremoval, aspsessionremoval, phpsessionremoval, bvsessionremoval 
*/
   public static final String NODE_URLSPEC = "urlspec";

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConnector.java
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConnector.java?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConnector.java
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/java/org/apache/manifoldcf/crawler/connectors/webcrawler/WebcrawlerConnector.java
 Thu Sep 28 04:20:30 2023
@@ -2506,36 +2506,37 @@ public class WebcrawlerConnector extends
     String inclusions = ".*\n";
     String inclusionsIndex = ".*\n";
     boolean includeMatching = true;
+    boolean forceInclusion = true;
+
 
     i = 0;
-    while (i < ds.getChildCount())
-    {
+    while (i < ds.getChildCount()) {
       SpecificationNode sn = ds.getChild(i++);
-      if (sn.getType().equals(WebcrawlerConfig.NODE_INCLUDES))
-      {
+      if (sn.getType().equals(WebcrawlerConfig.NODE_INCLUDES)) {
         inclusions = sn.getValue();
         if (inclusions == null)
           inclusions = "";
-      }
-      else if (sn.getType().equals(WebcrawlerConfig.NODE_INCLUDESINDEX))
-      {
+      } else if (sn.getType().equals(WebcrawlerConfig.NODE_INCLUDESINDEX)) {
         inclusionsIndex = sn.getValue();
         if (inclusionsIndex == null)
           inclusionsIndex = "";
-      }
-      else if (sn.getType().equals(WebcrawlerConfig.NODE_LIMITTOSEEDS))
-      {
+      } else if (sn.getType().equals(WebcrawlerConfig.NODE_LIMITTOSEEDS)) {
         String value = sn.getAttributeValue(WebcrawlerConfig.ATTR_VALUE);
         if (value == null || value.equals("false"))
           includeMatching = false;
         else
           includeMatching = true;
+      } else if (sn.getType().equals(WebcrawlerConfig.NODE_FORCEINCLUSION)) {
+        final String value = sn.getAttributeValue(WebcrawlerConfig.ATTR_VALUE);
+        forceInclusion = value != null && !"false".equals(value);
       }
     }
 
     velocityContext.put("INCLUSIONS",inclusions);
     velocityContext.put("INCLUSIONSINDEX",inclusionsIndex);
     velocityContext.put("INCLUDEMATCHING",includeMatching);
+    velocityContext.put("FORCEINCLUSION", forceInclusion);
+
 
   }
 
@@ -2862,7 +2863,26 @@ public class WebcrawlerConnector extends
       
cn.setAttribute(WebcrawlerConfig.ATTR_VALUE,(matchingHosts==null||matchingHosts.equals("false"))?"false":"true");
       ds.addChild(ds.getChildCount(),cn);
     }
-    
+
+    // Handle the force-inclusion switch
+    final String forceInclusionPresent = 
variableContext.getParameter(seqPrefix + "forceinclusion_present");
+    if (forceInclusionPresent != null) {
+      // Delete existing switch record first
+      int i = 0;
+      while (i < ds.getChildCount()) {
+        final SpecificationNode sn = ds.getChild(i);
+        if (sn.getType().equals(WebcrawlerConfig.NODE_FORCEINCLUSION))
+          ds.removeChild(i);
+        else
+          i++;
+      }
+
+      final String forceInclusion = variableContext.getParameter(seqPrefix + 
"forceInclusion");
+      final SpecificationNode cn = new 
SpecificationNode(WebcrawlerConfig.NODE_FORCEINCLUSION);
+      cn.setAttribute(WebcrawlerConfig.ATTR_VALUE, (forceInclusion == null || 
forceInclusion.equals("false")) ? "false" : "true");
+      ds.addChild(ds.getChildCount(), cn);
+    }
+
     // Get the exclusions
     String exclusions = variableContext.getParameter(seqPrefix+"exclusions");
     if (exclusions != null)
@@ -5701,6 +5721,7 @@ public class WebcrawlerConnector extends
       List<String> packList = new ArrayList<String>();
       String[] packStuff = new String[2];
       boolean limitToSeeds = false;
+      boolean forceInclusion = false;
       int i = 0;
       while (i < spec.getChildCount())
       {
@@ -5768,6 +5789,9 @@ public class WebcrawlerConnector extends
             limitToSeeds = false;
           else
             limitToSeeds = true;
+        } else if (sn.getType().equals(WebcrawlerConfig.NODE_FORCEINCLUSION)) {
+          final String value = 
sn.getAttributeValue(WebcrawlerConfig.ATTR_VALUE);
+          forceInclusion = value != null && 
!value.equals(WebcrawlerConfig.ATTRVALUE_FALSE);
         }
         else if (sn.getType().equals(WebcrawlerConfig.NODE_URLSPEC))
         {
@@ -5883,8 +5907,18 @@ public class WebcrawlerConnector extends
 
             String host = url.getHost();
 
-            if (host != null)
+            if (host != null) {
               seedHosts.add(host);
+              if (forceInclusion) {
+                // In case of redirection, if "Force the inclusion of 
redirects" is set to true, we add the redirected url to seedHosts
+                try {
+                  seedHosts.add(getFinalURL(urlCandidate));
+                } catch (IOException e) {
+                  // Skip the entry
+                }
+              }
+            }
+
           }
           catch (java.net.URISyntaxException e)
           {
@@ -6083,6 +6117,28 @@ public class WebcrawlerConnector extends
     public Map<String,List<String>> headerData = null;
 
   }
+
+  /**
+   * If the initial url is permanently or temporarly redirected (code 301 or 
302), the method returns the destination url
+   * @param url The initial url
+   * @return the url after redirection
+   */
+  public static String getFinalURL(String url) throws IOException, 
URISyntaxException {
+    URL formattedUrl = new URL(url);
+    URLConnection urlConnection = formattedUrl.openConnection();
+    HttpURLConnection con = (HttpURLConnection) urlConnection;
+    con.setInstanceFollowRedirects(false);
+    con.connect();
+    con.getInputStream();
+
+    if (con.getResponseCode() == HttpURLConnection.HTTP_MOVED_PERM || 
con.getResponseCode() == HttpURLConnection.HTTP_MOVED_TEMP) {
+      String redirectUrl = con.getHeaderField("Location");
+      return getFinalURL(redirectUrl);
+    }
+
+    final java.net.URI uri = new URI(url);
+    return uri.getHost();
+  }
   
 }
 

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_en_US.properties
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_en_US.properties?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_en_US.properties
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_en_US.properties
 Thu Sep 28 04:20:30 2023
@@ -75,6 +75,7 @@ WebcrawlerConnector.AddUrlRegexp=Add url
 WebcrawlerConnector.IncludeInCrawl=Include in crawl:
 WebcrawlerConnector.IncludeInIndex=Include in index:
 WebcrawlerConnector.IncludeOnlyHostsMatchingSeeds=Include only hosts matching 
seeds?
+WebcrawlerConnector.ForceInclusionOfRedirects=Force the inclusion of 
redirections (overwrites the Include only hosts option above)
 WebcrawlerConnector.ExcludeFromCrawl=Exclude from crawl:
 WebcrawlerConnector.ExcludeFromIndex=Exclude from index:
 WebcrawlerConnector.ExcludeContentFromIndex=Exclude content from index:

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_es_ES.properties
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_es_ES.properties?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_es_ES.properties
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_es_ES.properties
 Thu Sep 28 04:20:30 2023
@@ -72,12 +72,13 @@ WebcrawlerConnector.RemoveBVSessions=Ret
 WebcrawlerConnector.DeleteUrlRegexp2=Eliminar regexp url
 WebcrawlerConnector.NoCanonicalizationSpecified=No canonicalización 
especificada - todas las direcciones URL se reordenan y se han eliminado todas 
las sesiones
 WebcrawlerConnector.AddUrlRegexp=Añadir regexp url
-WebcrawlerConnector.IncludeInCrawl=Incluir en rastreo:
-WebcrawlerConnector.IncludeInIndex=Incluir en el índice:
-WebcrawlerConnector.IncludeOnlyHostsMatchingSeeds=Incluya sólo los hosts que 
emparejan semillas?
-WebcrawlerConnector.ExcludeFromCrawl=Excluir de rastreo:
-WebcrawlerConnector.ExcludeFromIndex=Excluir del índice:
-WebcrawlerConnector.ExcludeContentFromIndex=Excluir contenido del índice:
+WebcrawlerConnector.IncludeInCrawl=Incluir en rastreo:
+WebcrawlerConnector.IncludeInIndex=Incluir en el índice:
+WebcrawlerConnector.IncludeOnlyHostsMatchingSeeds=Incluya sólo los hosts que 
emparejan semillas?
+WebcrawlerConnector.ForceInclusionOfRedirects=Force the inclusion of 
redirections (overwrites the Include only hosts option above)
+WebcrawlerConnector.ExcludeFromCrawl=Excluir de rastreo:
+WebcrawlerConnector.ExcludeFromIndex=Excluir del índice:
+WebcrawlerConnector.ExcludeContentFromIndex=Excluir contenido del índice:
 WebcrawlerConnector.DeleteToken=eliminar símbolo #
 WebcrawlerConnector.NoAccessTokensPresent=No hay tokens de acceso actuales
 WebcrawlerConnector.AddAccessToken=Añadir token de acceso

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_fr_FR.properties
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_fr_FR.properties?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_fr_FR.properties
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_fr_FR.properties
 Thu Sep 28 04:20:30 2023
@@ -75,6 +75,7 @@ WebcrawlerConnector.AddUrlRegexp=Ajouter
 WebcrawlerConnector.IncludeInCrawl=Inclure dans le crawl:
 WebcrawlerConnector.IncludeInIndex=Inclure dans l'index:
 WebcrawlerConnector.IncludeOnlyHostsMatchingSeeds=Inclure uniquement les seeds 
de matching d'hôtes (hosts matching seeds)?
+WebcrawlerConnector.ForceInclusionOfRedirects=Forcer l'inclusion des 
redirections (écrase l'option ci-dessus)
 WebcrawlerConnector.ExcludeFromCrawl=Exclure du crawl:
 WebcrawlerConnector.ExcludeFromIndex=Exclure de l'index:
 WebcrawlerConnector.DeleteToken=Supprimer le jeton #

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_ja_JP.properties
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_ja_JP.properties?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_ja_JP.properties
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_ja_JP.properties
 Thu Sep 28 04:20:30 2023
@@ -75,6 +75,7 @@ WebcrawlerConnector.AddUrlRegexp=URLæ­
 WebcrawlerConnector.IncludeInCrawl=クロールに含める：
 WebcrawlerConnector.IncludeInIndex=索引に含める：
 
WebcrawlerConnector.IncludeOnlyHostsMatchingSeeds=シードと一致するホストのみ対象にする
+WebcrawlerConnector.ForceInclusionOfRedirects=Force the inclusion of 
redirections (overwrites the Include only hosts option above)
 WebcrawlerConnector.ExcludeFromCrawl=クロールから除外：
 WebcrawlerConnector.ExcludeFromIndex=索引が除外：
 WebcrawlerConnector.DeleteToken=トークンを削除 #

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_zh_CN.properties
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_zh_CN.properties?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_zh_CN.properties
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/native2ascii/org/apache/manifoldcf/crawler/connectors/webcrawler/common_zh_CN.properties
 Thu Sep 28 04:20:30 2023
@@ -72,12 +72,13 @@ WebcrawlerConnector.RemoveBVSessions=å�
 WebcrawlerConnector.DeleteUrlRegexp2=删除URL正则表达式
 WebcrawlerConnector.NoCanonicalizationSpecified=规范化未指定 - 
所有URL将被重排，会话被排除
 WebcrawlerConnector.AddUrlRegexp=添加URL正则表达式
-WebcrawlerConnector.IncludeInCrawl=包含于爬虫内: 
-WebcrawlerConnector.IncludeInIndex=包含于索引内: 
-WebcrawlerConnector.IncludeOnlyHostsMatchingSeeds=只包含和种子匹é…
çš„主机
-WebcrawlerConnector.ExcludeFromCrawl=排除于爬虫外: 
-WebcrawlerConnector.ExcludeFromIndex=排除于索引外: 
-WebcrawlerConnector.DeleteToken=删除令牌 #
+WebcrawlerConnector.IncludeInCrawl=包含于爬虫内: 
+WebcrawlerConnector.IncludeInIndex=包含于索引内: 
+WebcrawlerConnector.IncludeOnlyHostsMatchingSeeds=只包含和种子匹é…
çš„主机
+WebcrawlerConnector.ForceInclusionOfRedirects=Force the inclusion of 
redirections (overwrites the Include only hosts option above)
+WebcrawlerConnector.ExcludeFromCrawl=排除于爬虫外: 
+WebcrawlerConnector.ExcludeFromIndex=排除于索引外: 
+WebcrawlerConnector.DeleteToken=删除令牌 #
 WebcrawlerConnector.ExcludeContentFromIndex=Exclude content from index:
 WebcrawlerConnector.NoAccessTokensPresent=访问令牌不存在
 WebcrawlerConnector.AddAccessToken=添加访问令牌

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/editSpecification_Inclusions.html.vm
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/editSpecification_Inclusions.html.vm?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/editSpecification_Inclusions.html.vm
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/editSpecification_Inclusions.html.vm
 Thu Sep 28 04:20:30 2023
@@ -33,11 +33,18 @@
         <input type="hidden" name="${SEQPREFIX}matchinghosts_present" 
value="true"/>
       </div>
     </div>
+    <div class="form-group">
+      <div class="checkbox">
+        <label><input type="checkbox" name="${SEQPREFIX}forceInclusion" 
value="true" #if($FORCEINCLUSION) checked="yes" #end 
/>$Encoder.bodyEscape($ResourceBundle.getString('WebcrawlerConnector.ForceInclusionOfRedirects'))</label>
+        <input type="hidden" name="${SEQPREFIX}forceinclusion_present" 
value="true"/>
+      </div>
+    </div>
   </div>
 </div>
 #else
 <input type="hidden" name="${SEQPREFIX}inclusions" 
value="$Encoder.attributeEscape($INCLUSIONS)"/>
 <input type="hidden" name="${SEQPREFIX}inclusionsindex" 
value="$Encoder.attributeEscape($INCLUSIONSINDEX)"/>
 <input type="hidden" name="${SEQPREFIX}matchinghosts" 
value="#if($INCLUDEMATCHING)true#{else}false#end" />
+<input type="hidden" name="${SEQPREFIX}forceInclusion" 
value="#if($FORCEINCLUSION)true#{else}false#end" />
 <input type="hidden" name="${SEQPREFIX}matchinghosts_present" value="true"/>
 #end

Modified: 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/viewSpecification.html.vm
URL: 
http://svn.apache.org/viewvc/manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/viewSpecification.html.vm?rev=1912568&r1=1912567&r2=1912568&view=diff
==============================================================================
--- 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/viewSpecification.html.vm
 (original)
+++ 
manifoldcf/trunk/connectors/webcrawler/connector/src/main/resources/org/apache/manifoldcf/crawler/connectors/webcrawler/viewSpecification.html.vm
 Thu Sep 28 04:20:30 2023
@@ -78,10 +78,13 @@
     
<td>#if($INCLUDEMATCHING)$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.yes"))#{else}$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.no"))#{end}</td>
   </tr>
   <tr>
-    
<th>$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.IncludeInCrawl"))</th>
-    <td><pre>$Encoder.bodyEscape($INCLUSIONS)</pre></td>
+    
<th>$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.ForceInclusionOfRedirects"))</th>
+    
<td>#if($FORCEINCLUSION)$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.yes"))#{else}$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.no"))#{end}</td>
   </tr>
   <tr>
+    
<th>$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.IncludeInCrawl"))</th>
+    <td><pre>$Encoder.bodyEscape($INCLUSIONS)</pre></td>  </tr>
+  <tr>
     
<th>$Encoder.bodyEscape($ResourceBundle.getString("WebcrawlerConnector.IncludeInIndex"))</th>
     <td><pre>$Encoder.bodyEscape($INCLUSIONSINDEX)</pre></td>
   </tr>

Reply via email to