The syntax patterns in the list that follows can be combined to form rules.
Stem—(
text
)—An alphanumeric sequence of characters that matches any form of the stem. For example, the patternbank
matches the termsbanks, banking, bank, banker
, etc.Quoted string—(
"text"
)—A string delimited by quotation marks ("
) that matches its literal form. For example, the pattern"banks"
only matches the termbanks
.Percent quoted string—(
%text%
)—A string delimited by percent signs (%
) that matches any expansion of the stem, using the lexicon’s data. Since ATG Search allows users to define their own terminology in a Dictionary, this expansion data can be controlled by the user. For example, the pattern%computer%
might match the termslaptop, desktop
, etc.Wildcard—(
*text
)—A string containing one or more asterisks (*
). This pattern matches any stem that matches the wildcard pattern. For example, the pattern*computer
would match the termssupercomputer, minicomputer, computer
, etc.Macro—(
$macro_name
)—The name of a macro beginning with a dollar sign ($
). This pattern matches any of the stems listed in the macro definition. For example, the pattern$bank
would match any term in the definition such asbank, institution, credit union, savings bank
, etc.Meta tag—(
<meta name="property" content="value">
)—A<meta>
tag embedded in an indexed document. For example, the pattern<meta name="product" content="eServer">
would match any indexed document containing this <meta> tag. Meta tag rules have no text region so they cannot compete against any other rules.The properties created are controlled by the Text Processing Option set. Only document metatags that have been configured with these options are used to match these special topic patterns.
URL—(
"/URL"
)—A full or partial Internet address that points to a document or document directory. The URL can be a full URL or any relative path using the following four forms:http://www.mycorp.com/company/support/help.htm
is a full URL and only matches itself.http://www.mycorp.com/company/support
is a URL directory that matches any document contained within that directory or its subdirectories.http://www.mycorp.com/company/support/*
is a URL directory that matches any document contained within that directory but NOT its subdirectories./support/help.htm
is a relative URL, which matches any document whose URL ends with stringsupport/help.htm
.
Docset—A document set name. Sometimes the physical layout of indexed content suggests an appropriate way to categorize the content; for example, content from a specific subdirectory of a file system might map to a topic. Oracle ATG Web Commerce Search provides syntax for specifying document set patterns that should map to a topic. The syntax is similar to the URL rules. Note that this rule pattern cannot specify a single index item, only a set of them. The syntax requires a quoted string beginning with a forward slash, and has three forms:
"/dir"
– An absolute document set path, such as"/Documents/www.mycorp.com/product/help"
. Matches any index item contained within the document set, either directly or within child sets."/rel"
– A relative document set path, such as"/product/help"
. Matches all index items whose physical document set ends with the relative path."/dir/*"
– An absolute path to a document set followed by an asterisk, such as"/Documents/www.mycorp.com/product/*"
. Matches any index item that is immediately under the document set path, excluding those within child sets.