In addition to understanding how to use search
functionality from an end-user perspective, it’s also helpful to
understanding the moving parts behind the scenes. This section covers
two key components: the front-end components that drive the user
experience and the back-end components that power the indexing and
search query results.
The User Experience
In a SharePoint Site Collection, search pages,
search Web Parts, and navigation elements are combined to create the
search user experience. From an end-user perspective, the search box
might be considered the primary component of search. The search box Web
Part is one of seventeen or more search Web Parts which ships with the
product. There are search box Web
Parts such as Advanced Search Box, People Search Box, and the Search
Box, as well as search results Web Parts such as Federated Results,
People Search Core Results, Search Core Results, and Top Federated
Results. Each Web Part is configurable using a Web Part tool pane. In
addition to search Web Parts, SharePoint also includes search page
layouts for Advanced Search, People Search Results, Search Box, and
Each search box is configured to point to a
corresponding search results page, where the search results will
display when an end user performs a query. The results page contains
multiple Web Parts for displaying the results, as well as other
elements, such as action links, to provide a rich user experience.
As the Site Collection components make up the user
experience from a look-and-feel perspective, the search services
actually determine which items are most relevant to a user’s search
query. Relevance is a measure of how well items in the index meet the
user’s criteria. With each search request, SharePoint retrieves
matches, calculates a rank value for results from the index, and then
returns the results to the search results Web Part in an XML format.
The search results Web Part formats the raw XML into something more
user-friendly by way of its XSL style sheet. The search results style
sheet is configurable through the Web Part tool pane and customizable
in SharePoint Designer. By default, search results are sorted in an
order of rank value in descending order. In other words, the most
relevant items display at the top of the search results page. Relevancy
and rank are calculated by SharePoint using complex algorithms, which
weigh and process a variety of ranking parameters such as frequency of
search term, file type, click distance, URL depth, and language
SharePoint 2010 provides three site templates that
can be used for search. These templates provide the navigation, page
layouts, and Web Parts required to create a visual search experience. Figure 1 shows the search center site template options.
Enterprise Search Center.
A site for delivering the search experience. The Welcome page includes
a box with two tabs: one for general searches and another for searches
for information about people. You can add and customize tabs to focus
on other search scopes or results types.
Basic Search Center. A site for delivering the search experience. The site includes pages for search results and advanced searches.
FAST Search Center.
A site for delivering the FAST search experience. The Welcome page
includes a search box with two tabs: one for general searches and
another for searches for information about people. You can add and
customize tabs to focus on other search scopes or result types.
Figure 1. The three Search Center site templates are available in the Enterprise group of site templates
Index and Query Components
When a user performs a search query, SharePoint is
able to locate and retrieve the content from an index. Acting much like
an index you might find in the back of a text book, SharePoint
maintains an index of text from the
crawled content. By default, the index is located in the Windows file
system of the servers, which are running either the SharePoint search
index service or the SharePoint search query service. The default path
to the directory is C:\Program Files\Microsoft Office
Servers\14.0\Data\Applications\GUID. Within a subdirectory, files store
the unique keywords from crawled items as well as other pointer
information. Gather logs, noise word files, and thesaurus files can
also be located within subfolders of this directory location.
Where the index files are stored in the file system,
a SQL Server database, called the Metadata Property Store, stores
property information (used in scopes and filters) as well as security
definitions for the crawled content.
Content sources contain address information (URL,
UNC, and so on) and instruct SharePoint where to crawl. The index and
property stores are populated by the index engine when content sources
are being crawled. Crawling can be initiated manually or according to a
schedule. Content sources are configurable within the Search Service
Application, which is accessible from the Central Administration Web
application. SharePoint 2010 supports the following types of content