Auto-completion provides the user with real-time suggestions as query terms are typed into the search box.
Several types of auto-completion are supported:
- Simple auto-completion
is quick and easy to set up and returns suggestions that are words and phrases found within the content.
- Structured auto-completion
- Recent searches
Recent searches are suggested from the current user’s search history. This is available when search sessions are enabled on a search.
- Search-based auto-completion
Uses a results page, configured to return auto-completion format JSON, to provide search-based auto-completions. This approach can be used if additional ranking is required for suggestions. Due to performance overheads, this is not recommended for most implementations.
Auto-completion, when displayed to the end user can consist of several independent auto-completion data sets from the forms of auto-completion listed above.
By default, a single data set of simple auto-completion is generated for each results page.
Auto-completion is enabled by default using the auto-completion collection.cfg setting.
Configuring auto-completion data sets Before any auto-completion can be added one or more auto-completion datasets must be generated.
This is achieved by creating one or more results pages that generate an auto-completion index for use by any of the results pages within the same search package. Auto-completion is generated in one of the following ways:
As simple auto-completion that provides a set of keywords suggestions based on words within documents that belong to the results page.
As structured auto-completion, generated from an auto-completion.csv file uploaded to the results page.
As structured auto-completion, generated using the auto-completion plugin.
Auto-completion in Funnelback is delivered via a web service. The web service is accessed using a REST-style request and returns a JSON packet containing the auto-completion suggestions.
Existing templates (either based on the old jQuery UI widget, or the prototype/GitHub Concierge widget) can be upgraded to use the Concierge widget with minimal changes to existing templates.
v15 and earlier structured auto-completion was normally generated from the Funnelback search index using custom workflow commands. In v16 the generation process needs to be upgraded to use the auto-completion plugin.