Query Pipeline StagesConfiguration specifications
A query pipeline is made up of a series of query stages that process incoming search queries.
A pipeline stage definition associates a unique ID with a set of properties. These definitions are registered with the Managed Fusion API service and stored in ZooKeeper for re-use across pipelines and search applications.
Managed Fusion includes a number of specialized query stages as well as a JavaScript stage that allows advanced processing via a JavaScript program.
The detailed configuration article about each Managed Fusion query pipeline stage is provided in this section.
To configure a query pipeline stage, sign in to Managed Fusion and click Querying > Query Workbench. Then click Add a Stage and select the stage you want to configure. The stages are displayed under these categories:
-
Set Up. This section displays options that let you configure stages such as Facets, Graph Security Trimming, and Query Fields.
-
Results Relevancy. This section displays options that let you configure stages such as Block Documents, Boost Documents, Recommend Items for Item, and Recommend Items for User.
-
Fetch Data. This section displays options that let you configure stages such as Apply Rules, JDBC Lookup, and REST Query.
-
Responses. This section displays options that let you configure stages such as QnA Result Document Field Sorting, Response Diagnostics, and Response Document Exclusion.
-
Troubleshooting. This section displays the option that lets you configure the Logging stage.
-
Advanced. This section displays options that let you configure stages such as Additional Query Parameters, Managed JavaScript, and QnA Filter Stop Words.
-
Custom. This section displays options that let you configure stages such as Custom Graph Security Trimming and JavaScript.
-
AI. This section displays options that let you configure stages such as Hybrid Query, LWAI Prediction, and Machine Learning.
-
Other. This section displays options that let you configure stages such as Experiment Stage, Return Query Parameters, and Text Tagger.
For information about how to configure stages in a query pipeline, see Query Workbench.
For conceptual information, see Query pipeline stages.
Global pipeline stage properties
- Required
-
-
type: The type of the pipeline stage. This must be one of the defined Managed Fusion pipeline stage types, such as
index-logging. If the Managed Fusion UI is used to define the stage, this property is filled in automatically.
-
- Optional
-
-
label: A string field with a maximum length of 255 characters. The label is displayed on the Managed Fusion UI.
-
skip: A boolean value. If true, pipeline processing bypasses this stage. The default is false.
-
condition: A JavaScript expression that evaluates to true (1) or false (0). If this condition evaluates to false, this stage is skipped. The default is true.
-
Pipeline Condition Expression
The JavaScript expression specified in the condition property of a pipeline stage has access the pipeline objects.
Index Pipeline Stage Condition Example
An index PipelineDocument has two available variables: doc and ctx.
To check whether pipeline document contains a named field:
doc.hasField("acl_ss")Query pipeline stage Condition Example
To process a query based the query request object:
request.hasParam("fusion-user-name") && request.getFirstParam("fusion-user-name").equals("SuperUser");
!request.hasParam("isFusionPluginQuery")The first condition checks that the request parameter fusion-user-name is present and has the value SuperUser. The second condition checks that the request parameter isFusionPluginQuery is not present.