SEARCH_INTERFACE

The SEARCH_INTERFACE element is a named collection of Endeca standard attributes and/or managed attributes.

Both standard attributes and managed attributes can co-exist in a SEARCH_INTERFACE. The Endeca attributes in the group are specified in MEMBER_NAME elements.

If a standard attribute or managed attribute is not included in any SEARCH_INTERFACE element, then an implicit SEARCH_INTERFACE element is created with the same name as the standard attribute or managed attribute and that single standard attribute or managed attribute as its only member. The value for the CROSS_FIELD_RELEVANCE_RANK is set to 0.

Format

<!ELEMENT SEARCH_INTERFACE
    ( MEMBER_NAME+
    , PARTIAL_MATCH?
    )
>
<!ATTLIST SEARCH_INTERFACE
    NAME				                    CDATA		     	#REQUIRED
    DEFAULT_RELRANK_STRATEGY	   CDATA		     	#IMPLIED
    CROSS_FIELD_RELEVANCE_RANK	 CDATA		     	#IMPLIED
    CROSS_FIELD_BOUNDARY		      (ALWAYS
                                |ON_FAILURE
                                |NEVER)      "NEVER"
    STRICT_PHRASE_MATCH		       (TRUE|FALSE)	#IMPLIED
>

Attributes

The SEARCH_INTERFACE element has the following attributes.

NAME

A unique name for this search interface.

DEFAULT_RELRANK_STRATEGY

For record search, a default relevance scoring function assigned to a SEARCH_INTERFACE. For example, if your search interface is called Flavors, the DEFAULT_RELRANK_STRATEGY attribute has the value "Flavors_strategy".

CROSS_FIELD_RELEVANCE_RANK

Specifies the relevance rank score for cross-field matches. The value should be an unsigned 32-bit integer. The default value for CROSS_FIELD_RELEVANCE_RANK is 0.

CROSS_FIELD_BOUNDARY

Specifies when the search engine should try to match search queries across standard attribute/managed attribute boundaries, but within the members of the SEARCH_INTERFACE:
  • If its value is set to ON_FAILURE, the search engine will only try to match queries across standard attribute/managed attribute boundaries if it fails to find any match within a single standard attribute/managed attribute.
  • If its value is set to ALWAYS, the engine will always look for matches across standard attribute/managed attribute boundaries, in addition to matches within a standard attribute/managed attribute.
  • If its value is set to NEVER, the engine will not look across boundaries for matches. This is the default.

STRICT_PHRASE_MATCH

Specifies that the Dgraph should interpret a query strictly when comparing white space in the query with punctuation in the source text. If set to FALSE, partial word tokens connected in the source text by punctuation can be matched to a phrase query where the partial tokens are separated by spaces instead of matching punctuation. The default value of this attribute is TRUE.

Sub-elements

The following table provides a brief overview of the SEARCH_INTERFACE sub-elements.

Sub-element Brief description
MEMBER_NAME Specifies the name of an attribute that is part of a SEARCH_INTERFACE.
PARTIAL_MATCH Specifies if partial query matches should be supported for the SEARCH_INTERFACE that contains this element.

Example

This example establishes a search interface called AllFields, which contains four members. 

<SEARCH_INTERFACE CROSS_FIELD_BOUNDARY="ALWAYS" 
      CROSS_FIELD_RELEVANCE_RANK="0" 
      DEFAULT_RELRANK_STRATEGY="All" NAME="AllFields">
   <MEMBER_NAME RELEVANCE_RANK="4">ProductType</MEMBER_NAME>
   <MEMBER_NAME RELEVANCE_RANK="3">ProductName</MEMBER_NAME>
   <MEMBER_NAME RELEVANCE_RANK="2">SalesRegion</MEMBER_NAME>
   <MEMBER_NAME RELEVANCE_RANK="1">Description</MEMBER_NAME>
</SEARCH_INTERFACE>