Search Engine

This topic explains how to adjust the settings for an XperienCentral channel. In XperienCentral, navigate to Configuration > Channel Configuration. See the Setup Tool for information on configuring all other global XperienCentral settings. If your XperienCentral project contains multiple channels, you can select the channel to configure from the "Select Channel" drop-down list at the top of the Channel Configuration panel. Most changes you make to the channel configuration will become available after a browser refresh. Some configuration changes require a restart of XperienCentral (the web server).

In This Topic

Starting the Search Service

Adding basic search functionality to a website is straightforward. The following are the prerequisites:

• a search service that indexes the website
• a form for entering the search query
• a page on which to display the search results

If your installation already contains a working search service you can skip this section.

The search engine is not a default part of the XperienCentral application; it runs in a separate (Java) process, therefore it has to be started by running either a batch file (Windows) or a shell script (Unix). Because the search engine process is executed in a separate process, it needs to have a basic form of authentication. This is configured in the credentials.xml file. Open credentials.xml in a text editor. It is stored in the search engine conf directory:

• For normal GX WebManager installations on Windows computers this is <xperiencentral-root>\webmanager-searchengine\target\classes\conf.
• For UNIX servers this is normally something like /vol/www/search/indexer-mywebsite/conf.

Add a valid username and password in the second line of the file:

<credentials>
   <credential pattern=".*localhost.*" type="postform" username="gxsearch" password="Search987">
</credentials>

The First Site Index

There are two main ways to index the website: manually and automatically. The second is configured by running the search engine at a scheduled time from a cron job or scheduled service. How to configure this is covered in the installation manuals for Windows and Linux product environments. This chapter will describe how to index a website manually, and also on local computers where the indexing is not scheduled.

1. Navigate to the Setup Tool. Login in as Administrator and then click the Search Tools tab.
   
   
   
   If the search engine is running correctly then the lower part of the page under ‘Server Status’ contains the number of indexed pages plus some other information. Normally when the search engine has been started for the first time than the number of pages will be 0, as starting the search engine does not automatically mean that the indexing is started:
   
   
   
   
   
   

   Right above the "Search status" information is the "Index URL" pane. This pane contains three fields that tell the search engine

   • where to start crawling the website (the "URL" field).
   • which domains are allowed (the "host" field). This should contain at least the front-end hostname, the backend hostname. Optionally this can contain front-end and backend alias hostnames. The hostnames must be separated by a comma.
   • how deep the crawler goes. Normally this is set to 1, which means the crawler will index the page in the ‘URL’ field plus all the pages that are linked from the "URL" page.
   
   
   
   
   

   By default the "URL" field contains the URL of the so called indexer page. This indexer page contains links to all accessible XperienCentral pages, documents and articles. The URL is normally something like http://<domain>:<port>/web/webmanager?id=39016. To see all the URLs that will be indexed open this URL in a separate window.
   
   

2. Click on the [Index] button to start the indexing process. The search engine crawler will open the indexer page and start indexing all the pages and documents. This process will take anywhere from several seconds for a small website to several hours for a large website. For a detailed description of the indexing process see *** Crawling.

   
   
   
   
   

   In the upper part of the ‘Search Tools’ page the search engine can be tested by entering one or more search terms and hitting ‘<Enter>’ or clicking on the [Relevance] or [Date] buttons:
   
   
   
   

   In a normal situation this should return results. If there are zero documents in the index, or if there are documents in the index that are not returned in the search field, or if there is a certain document that’s missing, see ***troubleshooting.

Back to top

Search Element and Search Results

The Search Element

The basic search element is added to the page by adding the Search element to a page from the Insert menu:

1. Select the size of the "Search" box in pixels.
2. Select whether to restrict the search to pages that are on the same level as the page on which the Search element has been added.
3. Select whether to restrict the search to only the subpages of the page on which the Search element has been added.

4. Select where to display the search results. If you select "Yes", the results will be shown on the page containing the Search element. If you select "No" and then click [Apply], you can then select the page on which to display the results.
   
   

   On most websites the Search element is not placed on one or more pages but rather on a central page section. This is a better solution because you only have to do it once and can centrally manage the Search element.

Search Results Page

The default search results are displayed with the following basic set of information:

• A number and the title of the page or article
• The bread crumb path or navigation path (for pages only)
• A summary of the contents

Some presentation properties can be changed by creating a design template variant of the "WM Advanced search element".

1. Navigate to Configuration > Design Templates.
2. Set the scope to "SearchElement" and choose WM Advanced search element".
3. Click on the button [New presentation variant]:
   
   
   
   
   
4. Modify the new design template as desired.
5. Navigate to the page containing the Search element and select the new design template variant for the Search element  in order to preview the look and feel of the Search element.
   
   

Central Search Results Page

You can show search results on a special page. This is convenient because you can use a special presentation or add special content to accompany the results. To assign a special search results page, follow these steps:

1. Navigate to Configuration > Channel Configuration.
2. Click [Special Pages].
3. Click [Search] next to "Search-result page" and select the search results page from Advanced Search.

4. Click [Select] and then [Apply]. The search results page is assigned. For example:
   
   
   
   
   

   Be sure that the search results page that you select is not visible in the website navigation.

Back to top

Additional Indexing Options

Include/Exclude Pages

By default the search engine will index all pages and Content Repository items. It is possible, however to exclude certain pages from the index. Follow these steps:

1. Clear the search index in the Search Tools tab of the Setup Tool. Check the status: there should be 0 pages in the index.
2. Create some pages and content on the website to test the indexing.
3. To exclude a page from the indexing, click [Edit] in the Properties widget and navigate to the "Basic" tab.
4. Clear the "Include in Search Engine" checkbox:
   
   
   
   
   
5. Reindex the website from the Setup Tool.
   
   By clearing the "Include in Search Engine" option the page HTML will contain a special META tag that prevents search engines from indexing the website. The code of the tag is
   <meta name="robots" content="noindex" />. This will not only prevent the XperienCentral search engine from indexing that page but all other internet search engines such as Google as well.

Include/Exclude Page Parts

Page parts can be included or excluded in the search index. To specify whether a page part is included or excluded from the search index, follow these steps:

1. Navigate to Configuration > Design Templates.
2. Click [Page Parts].
3. Select a page part from the drop-down list.
4. Next to "Index", select either "Yes" or "No" to include its content in the search index. For example:
   
   
   
   
   

   The result of this configuration depends on the design template of the website. In normal situations, indexing is enabled for the main content page parts, therefore setting the indexing to "No" doesn’t prevent the page part from being indexed. Technically, the "No" setting does nothing and by setting the index to "Yes" the page part is wrapped with <!-- indexer: on --/> and <!-- indexer: off --/> tags. As a result, the indexing configuration in page parts will only be effective for websites where all content is wrapped in indexer: off tags and where individual page parts can be enabled for indexing.

Back to top

Meta Keywords

Before the rise of internet search engines such as Google, a lot of search engines relied on meta information such as keywords and descriptions on pages. Nowadays meta information such as keywords and descriptions hardly play a role for internet search engines. However, they can still play a role for internal search engines. GX WebManager contains options to configure meta keywords on three levels: 1) for the entire website, 2) for page templates and 3) for individual pages.

On most public websites it is considered bad practice to have identical keywords for a lot of pages. This can result in a lower ranking in internet search engines such as Google. Google expects content and meta information to be relevant, and by assigning a static keyword to all pages this can be considered to be less relevant. Therefore the general advice would be not to use the general keywords for an entire website or in page templates unless there is a special purpose.

Meta Keywords for the Entire Website

To enter keywords that will be added on every page and article go to Configuration > Channel configuration [General]. In the field "Default meta keywords" one or more keywords can be added. Multiple keywords have to be separated by commas.

Meta Keywords for Pages

Meta keywords can be assigned to individual pages. When you assign one or more meta keywords to a page, they override the website-wide meta keyword(s). To assign a meta keyword to a page, follow these steps:

1. Navigate to the page.
2. Click [Edit] in the Properties widget.
3. Click [SEO].
4. Enter the keyword in the "Meta Keywords" field. If you enter multiple keywords, separate them with a comma (no space).
5. Click [Save].

Back to top