Indexing dynamic content based on keys stored in database table

Question

We are researching how best to use SharePoint Search 2010 to crawl and index an external SQL Server database table. 

Our goal is as follows:

1. For each row in our table, index the article abstract database column.

2. Pass back to the search service the fully qualified URL to the article using the primary key for the article being indexed. (ie Article.aspx?&ArticleID=999).

3. We will also need to index the actual article contents (which are persisted as either a database blob column or file system based file).

4. Incrementally index new articles during scheduled crawls.

To accomplish this will we need to develop a custom indexing connector since the entity type seems to be static as it is always dynamic HTML served up thru the ASPX.

What is the best approach for doing this and have I posted this question in the correct forum?

Thanks in advance
Bob Baldwin aka vsdotnetguy
Trabon

---

TIA...vsdotnetguy

Answer 1

You can look at the following links to get started on creating a connnector that is optimized for searching.

http://msdn.microsoft.com/en-us/library/ee556429.aspx

http://msdn.microsoft.com/en-us/library/gg294165.aspx

http://msdn.microsoft.com/en-us/library/gg294168.aspx

---

Blog | SharePoint Field Notes Dev Tool | ClassMaster

Answer 2

Okay, I think we're starting to get this.  Let's say we want SharePoint Search to crawl a sql database table named Articles.  So among other methods, we will need to implement the IDEnumerator and SpecificFinder methods.  Within the SpecificFinder method, we return an object with properties (ie. ArticleAbstract) mapped to the table columns we want SharePoint to index on.  We can also override the Default Action so search result items point back to an external web link with querystring value based on the primary key (ArticleID).

That takes care of most everything except...we also need SP Search to index the contents of the associated Article file (persisted either as a database blob or as a file system file).  To accomplish this, can we read the file contents into a byte array and include this in the SpecificFinder object and will it actually index the byte array...OR do I need to do the Associated Attachment kind of thing?

Thanks in advance...Bob Baldwin aka vsdotnetguy
Trabon

---

TIA...vsdotnetguy