Unified Search

Iotellect Server features an integrated search engine that performs indexing of the unified data model and offers a unified search interface that can be used to lookup various contexts, variables, functions, events and class instances.

The unified search is accessed via Search action defined in the Root context. The action opens a special dialog those search field can use used to search the whole Iotellect Server context tree and database.

Indexing settings can be accessed via the Search plugin configuration. The settings include several configuration items:

This table includes several rows that define what is indexed in various groups of contexts. Each row defines:

• Context Mask of contexts to index.

If the context path of the resource being accessed matches or extends the context mask defined in the current line of the indexing settings table, the indexing settings defined in this line will be used as the effective indexing settings for this path. Once a matching line is found, no other lines are matched. So, for example, if you start the table with the * context mask (i.e, that's the first line in the table) no other lines will ever be checked, because that line matches all contexts. That's why the * context mask should be always the last line in the table.

In deployments with a large number of contexts, using the * context mask to index all contexts can result in degraded performance and lengthy indexing. In this case, performance can be improved by determining narrower specifications for what should be included in searches, creating targeted context masks for these entities, and removing the * context from indexing settings.

• Context Definitions flag enables indexing of context basic properties (such as names, descriptions, types, etc.)

• Variable Definition flag enables indexing of variable metadata (names, descriptions, field parameters, etc.)

• Function Definitions flag enables indexing of function metadata (names, descriptions, input and output field parameters, etc.)

• Events flag enables indexing of event metadata (names, descriptions, field parameters, etc.)

• Actions flag enables indexing of action metadata (names, descriptions, etc.)

• Variable Values table defines a list of variables whose values will be indexed. Note that Variable Definitions must be enabled to allow any variable values to be indexed

• Event Instances table defines a list of Events types whose instances will be indexed (level, event data, acknowledgements, and enrichments are indexed)

This property defines parameters for the indexes and performance in a deployment with a distributed architecture:

• Index Storage Folder settings that specifies a folder (or a sub-folder of Iotellect Server installation folder) that will contain the search index.

• Event Instance Cleaning Period indicates how frequently the index of event instances should be cleaned. When cleaning is performed, all event instances which have passed their Expiration Time will be removed from the index.

• Include External Search Sources into Distributed Search Results applies to deployments with a distributed architecture.

  • When enabled on a Provider, the Provider will include search results from External Data Sources in response to search requests from Consumers. The time Consumers wait for search results will be dependent on the time it takes for the Provider to receive and processes all responses from External Data Sources.

  • When disabled, the Provider will not query External Data Sources when processing search requests from a Consumer. This can result in reduced response times to Consumer search queries.

• Custom Search Results Dashboard allows a custom dashboard to be opened when the user opens the Unified Search bar in the user interface. When the custom dashboard is opened this way, a URL parameter with key searchValue will be included in the dashboard URL. The value of the parameter will be the search term entered by the user.

In the Class Indexing Settings table, you can configure which classes and class instances to index. This table includes the following parameters:

• Class Context is used to define a class context whose instances need to be indexed.

• Class Fields is used to define fields which need to be indexed.

The External Data Sources table allows for any number of external data sources to be seamlessly integrated into the Unified Search results. Any data source that can be accessed from Iotellect can be integrated, including external databases, responses from APIs, and so on.

Each row in the External Data Sources table defines how queries will be processed against an individual data source. The expressions in each field are described in detail further on, but the basic order of operations is as follows:

• Search Expression uses the query string to query the external data source, and must resolve to a data table containing search results. This data table of search results is then used by the Entity Value Expression.

• Entity Value Expression is evaluated for each row of the data table generated by the Search Expression, and must resolve to a String which represents a full response to a query. For example, if a was query against an external database containing blog information, the result would be the full blog post.

• Entity Type Expression is evaluated for each row of the data table generated by the Search Expression, and must resolve to a String which represents the response type. The type is user defined for external data sources, and in the example above, might be “Blog Post”.

• Reference Expression is evaluated for each row of the data table generated by the Search Expression, and must resolve to a String which indicates how to retrieve the query result. In the example above, a URL to the blog post could be appropriate.

• Fragment Expression is evaluated for each row of the data table generated by the Search Expression, and must resolve to a String. The fragment typically shows a small part of a full query response, and could be displayed in a search results page to give the user a quick understanding of each result.

The fields of the External Data Sources are described in more detail below:

• Data Source Name defines the name of the external data source.

• Search Expression extracts data from the external data source. The environment variable searchQuery contains the string provided by the user during the search operation. The expression must evaluate to a data table containing the query results. Possible methods could include querying a database device or querying a REST endpoint using an HTTP device.

• Entity Value Expression must evaluate to a String for each row of the data table from the Search Expression evaluation. The string should represent the main text of the result of the search query.

• Entity Type Expression must evaluate to a String for each row of the data table from the Search Expression evaluation. The string should represent the type of result of the search query. This value is user-defined, and is typically used as a label to group or filter query results on the custom search results dashboard. For comparison, the entity types used by Iotellect to define internal search results are variable, event, function, and context.

• Reference Expression must evaluate to a String for each row of the data table from the Search Expression evaluation. The string should represent the reference of result of the search query. For a search result which can be accessed by internet, this would generally be a URL.

• Fragment Expression must evaluate to a String for each row of the data table from the Search Expression evaluation. The string should represent a fragment of the result of the search query. This would often be a subset of the result value, containing the query term. The fragment should be suitable for use as a preview of the search result when displayed in the user interface.

Was this page helpful?

Tell us more

Send

Thank you!