Skip to contents

Search Resources Using Elasticsearch Query

query.Rd

Executes a search query against the improve platform's Elasticsearch index to find resources matching the specified criteria. Uses the same query syntax as the improve web UI search interface, supporting field-based searches, wildcards, and logical operators.

Usage

query(query, fetchSize = 100, fetchOffset = 0)

Arguments

query

Character. Search query using improve's query grammar. Examples:

  • "name='analysis'" - exact name match

  • "name='*analysis*'" - name contains "analysis"

  • "nodeType='Step' AND runStatus='FINISHED'" - combined criteria

  • "description='*population*'" - search in description field

fetchSize

Numeric. Maximum number of results to return. Defaults to 100. Use for pagination or limiting large result sets.

fetchOffset

Numeric. Number of results to skip before returning matches. Defaults to 0. Use with fetchSize to implement pagination.

Value

A data frame containing the matched resources with their metadata, or NULL if no resources match the query or if the query contains errors. The data frame includes standard resource fields like resourceId, entityId, name, path, nodeType, and any additional fields returned by the search (depending on the query).

Details

The function submits the query to the improve platform's Elasticsearch backend, which indexes resource metadata for fast searching. Search results include the resourceId for each match, which is then used to load full resource details.

Query syntax supports:

  • Field searches: fieldName='value'

  • Wildcards: fieldName='*partial*'

  • Logical operators: AND, OR

  • Date fields: use ISO format for date comparisons

  • Path-based searches: path='/Some/Folder/*'

If the query contains errors, a warning is logged with the error message from the server, and NULL is returned.

For searches within a specific folder hierarchy, use queryFolder which automatically restricts results to a folder and its descendants.

References

ics1143

See also

queryFolder for folder-scoped searches, loadResource to load resources by known identifiers

Examples

if (FALSE) { # \dontrun{
# Find all finished steps
query("nodeType='Step' AND runStatus='FINISHED'")

# Search for resources by name pattern
query("name='*baseline*'")

# Paginate through large result sets
page1 <- query("nodeType='File'", fetchSize = 50, fetchOffset = 0)
page2 <- query("nodeType='File'", fetchSize = 50, fetchOffset = 50)

# Find resources in a specific path
query("path='/Modeling/PopPK/*'")
} # }