none
Search Crawling a child list

    问题

  • Given:

    1. There exists a basic DocLib named 'Docs'
    2. There exists a basic List named Docs_MetaData
    3. 'Docs_MetaData' contains a Linked Column pointing to the ID of 'Docs'

    Question:

    Based on the Given, when the Search Crawler indexes a record in 'Docs_MetaData', does the information get associated to the value in 'Docs', such that when the user searches for a MetaData value, the Document in 'Docs' is returned in the search results?

    If not, is something like this possible in some other way?

    (The requirement is to store data in a pseudo-normalized fashion, so hoping that a MultiValue Field is NOT the only answer.)


    Steve Clark, MCTS | Twin-Soft.com

    2012年6月7日 19:19

答案

  • I think I understand what you are trying to do. Search really isn't "relational". You can try to manage the necessary properties so that you can return links to the Assets and Cases.

    Another option that I usually rely on is to use JQuery to perform a second search based on the results of the first search. Once you are returned your search results you can tag the specific search results by type and then perform a second search based on the first result. If a Document is returned, the JQuery would go and fetch the Assets and Cases related to the document. This is significantly easier if you are using Content Types.

    Then for Assets you could return cases, and Cases could return documents and Assets.

    This all depends on your ability to define the data and the queries. You can also look at creating Crawlable Entities in the BCS (Scot Hillier's book discusses this). I have been meaning to try crawling child entities, I just don't have enough experience, JQuery has always been the solution of choice.

    HTH,

    Matt


    Matthew McDermott, MVP SharePoint

    2012年6月11日 17:21

全部回复

  • Steve,

    It depends on how you are performing the linking. I have done this with the Managed Metadata Service and Taxonomy FIelds and it works well. With list lookups it is less effective. All data is crawled by the indexer, though it does not make associations. You can make the associations in the Search Results.

    Can you decribe what you want the Search User Experience to be?

    Matt


    Matthew McDermott, MVP SharePoint

    2012年6月10日 20:17
  • Documents are tied to Assets and Court Cases, and can have multiple of each type.  The loose specs are that the user searches for an AssetID and/or CaseID, and sees the documents for them, vice just the metadata record(s) associated.

    We're setting up a test bed today for it, so any tips are appreciated.

    ...

    Just reviewed the Managed MetaData concept again, and it is difficult as new AssetID and CourtCase are made when needed, so the user would need to update the Managed terms prior to associating to the Document. Thus, I'm thinking non-starter, but feel free to enlighten.


    Steve Clark, MCTS | Twin-Soft.com


    • 已编辑 SClark_EasyBins 2012年6月11日 14:27 I'm a little smarter now.
    2012年6月11日 14:21
  • I think I understand what you are trying to do. Search really isn't "relational". You can try to manage the necessary properties so that you can return links to the Assets and Cases.

    Another option that I usually rely on is to use JQuery to perform a second search based on the results of the first search. Once you are returned your search results you can tag the specific search results by type and then perform a second search based on the first result. If a Document is returned, the JQuery would go and fetch the Assets and Cases related to the document. This is significantly easier if you are using Content Types.

    Then for Assets you could return cases, and Cases could return documents and Assets.

    This all depends on your ability to define the data and the queries. You can also look at creating Crawlable Entities in the BCS (Scot Hillier's book discusses this). I have been meaning to try crawling child entities, I just don't have enough experience, JQuery has always been the solution of choice.

    HTH,

    Matt


    Matthew McDermott, MVP SharePoint

    2012年6月11日 17:21