none
Sharepoint Custom Search Metadata handling RRS feed

  • Question

  • Hi

    I like to paint you my pie in the sky idea and am after input how easy this is likely to be ..

    Have a Sharepoint 2010 Server + Fast Search Server

    Problem Domain : We have a large collection of vms and “projects” for our applications. These projects are all the custom graphics, scripts etc to run our application. All these combined are large, many terrabytes. Thus this information has to live on network drives … not as files under sharepoint.

    Also if you were to index these files, it would serve little purpose to identify the purpose of the content. Ideally we would like to be able to search for candidates to be re-used for new projects.

    Possible solution : We could develop an app that used streams (http://msdn.microsoft.com/en-us/library/windows/desktop/aa364404(v=vs.85).aspx) to add extra tag info and a description, to big zip or file (at this stage I don’t know if a directory can have a file stream, e.g. for vms).

    Write a custom parser for Fast Search to extract and store the tag information so that searches can work over network areas. Now in this case this custom item can’t be file suffix related as no specific suffix would apply, e.g. you could let the normal *.docx, *pdf apply but run this custom item as the default no match case..

    Question : Does this sound feasible and could you provides some pointers on how I would go about this ?

    NB: IDE of choice vs2010 C#

    Thanks

    p.s. i tried, and yes a folder can have streams attached to it

    p.s.s. i found a few articles on customisation, and they reference in / out guids, but the posts i have seen don't seem to explain where these guid's came from, if you know, this will help

    Thursday, April 12, 2012 10:51 AM

Answers

All replies

  • Hi,

    Using streams to store metadata, or store them alongside the files as xml are both options you can consider. The xml approach might be better to make sure the data is there if you copy/move the files.

    Take a look at the file crawler BCS sample at http://msdn.microsoft.com/en-us/library/ff953161.aspx / http://code.msdn.microsoft.com/SharePoint-2010-Creating-5b6a3be6. You should be able to modify it to suit your need.

    Regards,
    Mikael Svenson


    Search Enthusiast - SharePoint MVP/WCF4/ASP.Net4
    http://techmikael.blogspot.com/

    Sunday, April 15, 2012 10:56 AM
  • Mikael

    I checked and stream data does move with the file, e.g. a directory or a file copied (under Windows NTFS assumed) does move the stream with it. However separate XML file has merits and can always be opened, typed to manually see data.

    As for the other URLs (thanks), i don't see the clear connection between a "Custom Business Connector" and info in a Search Window ? e.g. i would like the Word document search indexing to be similar to the data here so if i searched for "vm windows 7 sp1 apollo" and this was in a Word document as well as being tags for a VM, these would be shown.

    Haven't read yet, but i assume http://msdn.microsoft.com/en-us/library/gg512148.aspx is the answer...

    Thanks

    Monday, April 16, 2012 1:00 AM
  • Hi,

    You could also create a custom extensibility stage which for your files will either read the stream data or the related xml file with the meta data. This saves you from creating a custom connector/crawler.

    And I know stream data follows when moving between NTFS, but are you sure it's being kept if the files is opened and saved by an arbitrary end-user program?

    Thanks,
    Mikael Svenson


    Search Enthusiast - SharePoint MVP/WCF4/ASP.Net4
    http://techmikael.blogspot.com/

    Monday, April 16, 2012 6:49 AM