+1-888-365-2779
Try Now
More in this section

Forums / Bugs & Issues / In-accurate Search Indexing

In-accurate Search Indexing

16 posts, 0 answered
  1. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    07 Aug 2009
    Link to this post
    I have created three different search indexes. They have all been the same setup; all pages, news, and blogs. The first index shows 2724 words and 110 pages. The second index shows 1832 words and 80 pages, and the last shows 2495 words and 110 pages.

    We have been trying to figure out why a search term is not showing results when a page name has the term, the page has a title, Generic Content, with the term and also has the term in a Generic Content block of text. Out of all three indexes, this term is not found. The term by the way is "EnSite". The page with this term in it has the indexing option selected.

    What I really don't understand is the inconsistency in the indexing. If all indexes are the same you should get the exact results. Now I know it could be off one or so words as content can be changed and such, but this is a development site and the content is not being changed. I have deleted the search index folder and re-indexed and still get odd results.

    We are running a 3.5 site. I know this is old but it is what we are working with. Any ideas??
  2. Georgi
    Georgi avatar
    3583 posts
    Registered:
    28 Oct 2016
    12 Aug 2009
    Link to this post
    Hello Richard Baugh,

    I just spoke with our QAs and they have let me know that the count number of the indexing is valid for the page index only. In other words, the News index and Blog index providers are not counting the indexed words correctly. This is logged and is in our ToDo list.

    As for the term not appearing in the search results, you can try to use this tool - http://www.getopt.org/luke/ to see the contents of the index, and check if the word is indexed. You can also check if there is a rule in the fieldsInfoProvider.xml file, which prevents the indexer from reaching the word.

    Let me know about the results from your investigation.

    Greetings,
    Georgi
    the Telerik team

    Instantly find answers to your questions on the newTelerik Support Portal.
    Check out the tipsfor optimizing your support resource searches.
  3. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    12 Aug 2009
    Link to this post
    I used the tool and the terms are not in the index. These terms are on the site in the page name and title along with the content of the page.

    Here is the contents of the fieldsInfoProvider.xml file:
    <fields> 
      <field name="cmsWrapper" weight="3" indexAttribute="" filterTag="div" filterAttributes="id:cmsWrapper"/> 
      <field name="title" weight="1" indexAttribute="" filterTag="title" filterAttributes="" /> 
      <field name="keywords" weight="1" indexAttribute="content" filterTag="meta" filterAttributes="name:keywords;" /> 
      <field name="description" weight="1" indexAttribute="content" filterTag="meta" filterAttributes="name:description;" /> 
      <field name="script" weight="-1" indexAttribute="" filterTag="script" filterAttributes="" /> 
      <field name="style" weight="-1" indexAttribute="" filterTag="style" filterAttributes="" /> 
    </fields> 

    Richard
  4. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    12 Aug 2009
    Link to this post
    I did some more searching and found that I am also not getting the index for the parent page. The terms that are not showing up in the index, are on two separate pages. These pages are under another page that is not getting indexed. I looked at the properties for this page, and it is set to allow indexing. It is a public page and navigable.

    It seems that it won't index the two child pages if it doesn't index the parent page. All adjacent pages to the parent page are getting indexed.

    Richard
  5. Parvan
    Parvan avatar
    151 posts
    Registered:
    24 Sep 2012
    17 Aug 2009
    Link to this post
    Hi Richard Baugh,

    Thank you for using our services.

    I suppose that you are using Sitefinity 3.5 (from your first post).

    1. First try to use the default fieldsinfoProvider.xml (just delete the file and it will created), because the changes you have made can cause the differences.

    2. This problem should have been fixed in Sitefinity 3.6 (3.7). If the problem occurs again, upgrade to Sitefinity 3.7.

    I note that you do not specify the index field name (see the green text):
    <fields>  
      <field name="cmsWrapper" weight="3" indexAttribute="content" filterTag="div" filterAttributes="id:cmsWrapper"/>  
      <field name="title" weight="1" indexAttribute="" filterTag="title" filterAttributes="" />  
      <field name="keywords" weight="1" indexAttribute="content" filterTag="meta" filterAttributes="name:keywords;" />  
      <field name="description" weight="1" indexAttribute="content" filterTag="meta" filterAttributes="name:description;" />  
      <field name="script" weight="-1" indexAttribute="" filterTag="script" filterAttributes="" />  
      <field name="style" weight="-1" indexAttribute="" filterTag="style" filterAttributes="" />  
    </fields>  

    I hope this helps you.

    Greetings,
    Parvan
    the Telerik team

    Instantly find answers to your questions on the newTelerik Support Portal.
    Check out the tipsfor optimizing your support resource searches.
  6. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    24 Aug 2009
    Link to this post
    Sorry for the late post, I was out last week. I have deleted the fieldsinfoProvider.xml file and had the search recreate the file. I have also added the index field name. Neither of which have resulted in the search working correctly. As I mentioned, it won't index the parent page either. If it won't index the parent page I don't expect it to index the child pages.

    All pages are published, navigable, and are set to be indexed.
  7. Parvan
    Parvan avatar
    151 posts
    Registered:
    24 Sep 2012
    27 Aug 2009
    Link to this post
    Hi Richard Baugh,

    The problem is really strange. We have tried but cannot reproduce the issue yet.

    The only possible reason I can find is that the parent page that is not indexed is not public (the Allow anonymous access property of the page is set to false) and the PageIndex's property Index non-public pages is set to false. 
    Then its children will not be indexed, but the siblings can be indexed (if they are public). 

    I know that you said that the page is public but this is what we can suggest by knowing how Sitefinity PageIndex works.

    All the best,
    Parvan
    the Telerik team

    Instantly find answers to your questions on the new Telerik Support Portal.
    Watch a video on how to optimize your support resource searches and check out more tips on the blogs.
  8. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    27 Aug 2009
    Link to this post
    The page is published, anonymous access is allowed, and the index is true. The page is visible to the public and shows up in the menu. I'll try to re-create the pages in hope that it was just a random issue.
  9. Georgi
    Georgi avatar
    3583 posts
    Registered:
    28 Oct 2016
    27 Aug 2009
    Link to this post
    Hi Richard Baugh,

    Is there a way for you to check if the child page has a ParentID different than null?

    Greetings,
    Georgi
    the Telerik team

    Instantly find answers to your questions on the new Telerik Support Portal.
    Watch a video on how to optimize your support resource searches and check out more tips on the blogs.
  10. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    27 Aug 2009
    Link to this post
    Here is a new twist. I duplicated the parent page and then ran a search. The duplicated page shows up in the search. This was even with the page not set to be in the menu, non-navigable. I then moved the child pages under this page, renamed the live page and then renamed the dup page to the original live page. Basically causing the dup page to now show in the menu and have the old one set to hidden. When I ran a search, using the new index based on the auto index, I was able to see the new duplicated parent page, but the child pages were still not showing.

    During our posts back to one another, the client added some new pages in the site that also had the terms that are on these pages that are not being indexed. When running a search, these new pages were showing up. Since the moved child pages were not showing up after duplicating the parent page, I decided to manually for a new index. After forcing the new index, none of the pages showed up. I then went back into the pages that were showing up and edited them and published them. This then allowed them to show up in the search again, but the child pages that are part of the original problem are still not showing even after editing them.

    I decided to check the DB as you requested. The parentID is correct for all of the pages. This is also true as they are displaying correctly in the admin section and also on the live site.
  11. Georgi
    Georgi avatar
    3583 posts
    Registered:
    28 Oct 2016
    01 Sep 2009
    Link to this post
    Hi Richard,

    Thank you for the details. Although we cannot reproduce this here, we will continue with the testing. We will try to follow up once we find something.

    Best wishes,
    Georgi
    the Telerik team

    Instantly find answers to your questions on the new Telerik Support Portal.
    Watch a video on how to optimize your support resource searches and check out more tips on the blogs.
  12. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    01 Sep 2009
    Link to this post
    One thing I have noticed, and this may be a 3.5 bug or intended actions, is that when we manually index the site we are always redirected to the home page after some time. When going back to the admin section, it seems that the index was completed. Not sure if this has anything to do with our problem.
  13. Georgi
    Georgi avatar
    3583 posts
    Registered:
    28 Oct 2016
    01 Sep 2009
    Link to this post
    Hi Richard Baugh,

    Can you please check the log file, when you got logged out? Is there any exception from that time?

    Kind regards,
    Georgi
    the Telerik team

    Instantly find answers to your questions on the new Telerik Support Portal.
    Watch a video on how to optimize your support resource searches and check out more tips on the blogs.
  14. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    01 Sep 2009
    Link to this post
    We are not getting logged out, just redirected to the home page. If I use the browser's back button and the select "Search" from the command panel, the Search control will reload. There is not a need to re-login as we are still logged in. As for the log file, there doesn't seem to be anything in it.
  15. Georgi
    Georgi avatar
    3583 posts
    Registered:
    28 Oct 2016
    04 Sep 2009
    Link to this post
    Hi Richard Baugh,

    Do you have any Response.Redirects on the page, which for any reason can direct the requests to the login form? If yes,you could try to remove them temporary and try running the indexers again. If this does not help, I guess that we will have to ask for your database.

    Greetings,
    Georgi
    the Telerik team

    Instantly find answers to your questions on the new Telerik Support Portal.
    Watch a video on how to optimize your support resource searches and check out more tips on the blogs.
  16. Richard Baugh
    Richard Baugh avatar
    201 posts
    Registered:
    22 Aug 2012
    04 Sep 2009
    Link to this post
    Ok, I got the redirect issue worked out but I am still having problems with the indexing.

    Problems Re-Cap
    Running a manual index returns different index than the auto-index. i.e. pages indexed/words indexed
    Manual and auto indexes are not indexing all of the navigable pages.
Register for webinar
16 posts, 0 answered