Search basics

A query tree structure, along with order and paging specifiers, allows a search to be performed against indexed documents held in ElasticSearch. The query API allows any required sub-query structure to be defined and a comprehensive selection of Operators enable individual field level evaluation.

• Queries
• Location searches
• Sub-queries
• Ordering
• Paging
• Weighting
• Specifying fields
• Ordering by distance
• Complete example
• HTTP GET queries

This example demonstrates a simple search using HTTP POST.

Searches can also be carried out using HTTP GET queries.

Search for locations within a radius of a specified location.

Find all entries which have a location within 10.5 miles of Ludlow Castle's location. Append the search value at the end of the distance specified, so for example "10.5mi" or "10.5miles".

When searching for a location field, you can also order the results by distance.

A sub-query is a query within another query that is used as a condition to further restrict the results. Effectively they are defined by an explicit nesting of logical operators.

This example demonstrates a simple search with a sub-query:

Results can be ordered by one or more fields in an ascending or descending direction. Order clauses are prioritised in the order that they are added. By default, if no order clauses are specified then the entry results are ordered by:

• A relevancy 'score' for each entry for the search query (from ElasticSearch) in a descending direction
• The EntryTitle in an ascending direction.

This is the best solution for searches using a search term of some kind as it will promote the most relevant results to the top of the results. However, this can occasionally mean that an odd entry might appear on more than one page as the relevancy score is dynamically calculated on each paged request. As such, if you intend to page through the complete result set without any duplicates you should always specify an order by field.

Order by releaseDate in an ascending direction.

Order by releaseDate in a descending direction.

Paging allows the number of results to be restricted to a defined count so that the results are easier to handle and ensures a response is returned quickly. If you intend to page through the complete result set without any duplicates you should always specify an order by field.

The page number can also be specified to allow which set of results is to be returned. The page size is limited to a maximum of 10,000 however this is not recommended.

System fields such as id, contentTypeId, projectId, versionNo etc. are under the sys object and can be accessed using a dot notation, e.g. sys.id, sys.contentTypeId, sys.projectId, sys.version.versionNo.

The entryTitle and entryDescription fields are dynamic values, determined by the EntryTitleField and EntryDescriptionField values in the content type.

Fields defined in the content type for the entry can be accessed by their API id.

All fields can be searched by specifying an asterisk (*) in the field id. Note, the only supported operator is FreeText.

Searching on array fields is the same as searching on a child property of a field, you just use the dot notation. All operators support searching across array fields.

This example searches for a quote source of "Bruce Willis" within a quote array field called movieQuote.

When searching by location, to return the search results according to the distance of the location field in each entry from the distance specified in the distanceWithin search, add the location field to the orderBy clause, e.g.

The search results will include the distance from the location specified in the distanceWithin search to the location in the entry, in the same units that were specified in the search (but without the distance units). e.g.

The following example combines the ordering, paging and weighting concepts.

If you have large entries and only require a subset of fields it is worth limiting the fields returned in the results. This will reduce the size of the payload from the API which in turn will improve performance. The fields returned in the results can be specified using a "fields" clause. You can include fields by specifying the field API ID.

Example of returning the title and synopsis

If no fields are specified then all the data fields and system fields for an entry will be returned.

Fields defined in the content type for the entry can be specified using their API ID.

By default all system fields will be returned for an entry.

If you specify -sys, then the following system fields will still be returned as they are the minimum set:

• sys.id
• sys.dataFormat
• sys.language
• sys.contentTypeId
• sys.uri

Specific system fields can be specified to be returned by being prefixed with sys., e.g.

The optional system fields that can be specified are:

• sys.projectId
• sys.owner
• sys.slug
• sys.isPublished
• sys.availableLanguages
• sys.allUris
• sys.workflow
• sys.workflow.id
• sys.workflow.state
• sys.metadata
• sys.properties
• sys.version
• sys.version.publishedBy
• sys.version.createdBy
• sys.version.created
• sys.version.versionNo
• sys.version.modified
• sys.version.modifiedBy
• sys.version.published

It is not possible to specify individual fields within the properties and metadata fields.

Searches can also be carried out using an HTTP GET query, with the JSON for the individual sections of the query (where, orderBy, pageSize, pageIndex, fields) specified as querystring parameters.

This example demonstrates a simple search using HTTP GET:

This example demonstrates a more complex search using all the search options (including fieldLinkDepths which is only available in Contensis version 16+):

N.B. The querystring parameters will need to be URL encoded, this has not been done in the above examples for clarity. When URL encoded the examples look like this: