Search

Bosca Search helps users find the right content fast. It combines standard full-text search with optional vector search for semantic relevance, and it integrates tightly with permissions, workflows, and metadata.

What you get:

• Relevance tuned by your content models (Collections, Metadata, fields)
• Facets and filters for rich exploration
• Permission-aware results that only return what a user can see
• Incremental indexing driven by Workflows
• Optional vector search via Meilisearch's embeddings

How indexing works:

• Automatic Updates: Content moves through Workflows. When items are Published, they are indexed; when Unpublished or Deleted, they are removed.
• Secure by Design: Only public or permitted content is discoverable. Bosca checks permissions at query time so users only see what they are allowed to see.
• Smart Search: Vector fields (optional) allow for "semantic search"---finding results based on meaning, not just matching keywords.
• Customizable Indexing: Jsonata transformations let you define exactly how content fields map to search index attributes, giving fine-grained control over what gets indexed and how it is structured.

Search Settings

The Administration interface includes a dedicated Search Settings page where you can configure index settings, manage search attributes, and review indexing status.

For developers

Bosca's search is built on Meilisearch for speed and flexibility. It uses a defined IndexConfiguration to map your Collections to search indexes.

Key features:

• GraphQL API: Query search results directly with filters, facets, and pagination.
• Custom Indexing: Define exactly which fields and attributes are searchable, with Jsonata transformations for flexible field mapping.
• Vector Support: Seamless integration with embeddings for AI-powered search experiences.
• Collection Indexing: Collections are indexed alongside metadata for comprehensive search results.

Related:

• Architecture: Search in Architecture
• Workflows: Publishing lifecycle
• Source code: backend/framework/core-search, backend/framework/search