This guide has the three sections regarding problem, solution and results. For better understanding, review them in order, as each section builds on concepts presented in previous sections.
Audience should have a prior working knowledge of SharePoint 2010 and FAST Search for SharePoint Server 2010.
This guide is designed to highlight the solution for crawling the related data in SharePoint Server 2010 Fast Search. It provides an overview of the problem statement, a solution for the issue, and concludes with implementation of the solution suggested.
This section describes why exactly we need relational crawled properties. It refers to a practical scenario where we need to address the crawled property for related data.
Suppose we have 4 SharePoint lists names Cities, Universities, Courses & Colleges
As seen from above schema, there is no information related to city in the Colleges list. So now when user searches any record related to any city, FAST Search will retrieve the record from Cities list only.
We want to achieve that, when a user searches for any city, then related records must also be retrieved (i.e. from Universities list and Colleges list also, as City is a lookup column in Universities list and University column is a lookup column in Colleges list.)
The pipeline extensibility of FAST Search has been customized and is called Processing Component. The item processing pipeline prepares an item from a content source for indexing and searching. Additional searchable metadata has been generated from the content. This preparation includes text extraction, language detection, and tokenization. The following diagram depicts the architecture of how pipeline extensibility in FAST Search works. It shows the standard sequence of ‘stages’ from crawl to Index.