Search

Full-text search with Elasticsearch, Meilisearch, Typesense, or in-memory backend

Overview

github.com/xraph/forge/extensions/search provides a unified full-text search interface backed by Elasticsearch, Meilisearch, Typesense, or an in-memory implementation. It registers a SearchService in the DI container that manages connection lifecycle and exposes index management, document CRUD, search, suggest, and autocomplete operations.

What It Registers

Service	DI Key	Type
Search service	`search`	`*SearchService` (also satisfies `Search`)

The service is managed by Vessel. Vessel calls Start (connects to the backend) and Stop (disconnects) automatically.

Quick Start

package main

import (
    "context"
    "fmt"

    "github.com/xraph/forge"
    "github.com/xraph/forge/extensions/search"
)

func main() {
    app := forge.NewApp(forge.AppConfig{Name: "my-app", Version: "1.0.0"})

    // Register with defaults (in-memory backend)
    app.RegisterExtension(search.NewExtension())

    // Or connect to Meilisearch
    app.RegisterExtension(search.NewExtension(
        search.WithDriver("meilisearch"),
        search.WithURL("http://localhost:7700"),
        search.WithAPIKey("your-api-key"),
    ))

    ctx := context.Background()
    app.Start(ctx)
    defer app.Stop(ctx)

    s := search.MustGetSearch(app.Container())

    // Create an index with a schema
    s.CreateIndex(ctx, "products", search.IndexSchema{
        Fields: []search.FieldSchema{
            {Name: "title", Type: "text", Searchable: true, Sortable: true},
            {Name: "price", Type: "float", Filterable: true, Sortable: true},
            {Name: "category", Type: "keyword", Filterable: true, Faceted: true},
        },
    })

    // Index a document
    s.Index(ctx, "products", search.Document{
        ID:     "prod-1",
        Fields: map[string]any{"title": "Wireless Mouse", "price": 29.99, "category": "electronics"},
    })

    // Search with filters and facets
    results, _ := s.Search(ctx, search.SearchQuery{
        Index:   "products",
        Query:   "wireless",
        Filters: []search.Filter{{Field: "price", Operator: "<=", Value: 50.0}},
        Facets:  []string{"category"},
        Limit:   10,
    })

    for _, hit := range results.Hits {
        fmt.Printf("  %s (score: %.2f)\n", hit.ID, hit.Score)
    }
}

Using Search in Your Services

Inject *search.SearchService into your constructors for automatic DI resolution:

type ProductSearch struct {
    search search.Search
    logger forge.Logger
}

func NewProductSearch(s *search.SearchService, logger forge.Logger) *ProductSearch {
    return &ProductSearch{search: s, logger: logger}
}

func (ps *ProductSearch) Find(ctx context.Context, query string, maxPrice float64) ([]search.Hit, error) {
    results, err := ps.search.Search(ctx, search.SearchQuery{
        Index:     "products",
        Query:     query,
        Filters:   []search.Filter{{Field: "price", Operator: "<=", Value: maxPrice}},
        Highlight: true,
        Limit:     20,
    })
    if err != nil {
        return nil, err
    }
    return results.Hits, nil
}

forge.ProvideConstructor(app.Container(), NewProductSearch)

Key Concepts

Drivers -- select inmemory, elasticsearch, meilisearch, or typesense via config. The in-memory driver is fully functional and useful for development and testing.
Index management -- create, delete, list, and inspect indexes with typed schemas including field definitions, analyzers, synonyms, and ranking rules.
Document operations -- index single or bulk documents, retrieve by ID, update, and delete.
Search -- query with filters, sorting, faceting, highlighting, fuzzy matching, field boosting, and score thresholds.
Suggest and Autocomplete -- dedicated query types for search-as-you-type experiences.

Important Runtime Notes

Extension health delegates to the backend's Ping() method via Vessel.
The Stop() method logs disconnect errors but does not propagate them to avoid blocking shutdown.
Backend stubs for Elasticsearch, Meilisearch, and Typesense are declared. Verify your target driver is fully implemented before production use.

Detailed Pages

Features

Search capabilities, supported operations, and backend comparison.

Configuration

Config reference, YAML example, and programmatic options.

Functions

Public API grouped by purpose with descriptions.