lexigram-search

Full-text search and indexing for Lexigram Framework — Elasticsearch, Meilisearch, Typesense, and OpenSearch

Overview

lexigram-search provides a unified SearchEngineProtocol interface over Meilisearch, Elasticsearch, OpenSearch, Typesense, PostgreSQL full-text, MySQL, MongoDB, and SQLite. It supports typo-tolerant search, faceting, fuzzy matching, result caching, and analytics. All services are wired via SearchProvider, which registers the search engine protocol with the DI container.

Full documentation: docs.lexigram.dev

Install

uv add lexigram-search
# Optional extras
uv add "lexigram-search[meilisearch]"     # Meilisearch
uv add "lexigram-search[elasticsearch]"   # Elasticsearch 8.x
uv add "lexigram-search[typesense]"       # Typesense
uv add "lexigram-search[postgres,mysql,sqlite,mongodb]"  # Database backends

Quick Start

from lexigram import Application
from lexigram.di.module import Module, module

# Import the module from the package
from lexigram.search import SearchModule


@module(imports=[SearchModule.configure(...)])
class AppModule(Module):
    pass


async with Application.boot(modules=[AppModule]) as app:
    # use app.container to resolve services
    ...

Configuration

Default config: Pass SearchConfig() explicitly to use all defaults (in-memory backend). SearchModule.configure() with no arguments raises ValueError — a config or engine must be specified. The in-memory NullBackend is fine for development and tests.

Option 1 — YAML file

# application.yaml
search:
  backend_type: meilisearch
  meilisearch:
    url: http://localhost:7700
    api_key: "${MEILI_API_KEY}"
  query:
    strategy: fuzzy
    default_limit: 10

Option 2 — Profiles + Environment Variables (recommended)

export LEX_SEARCH__ENABLED=true
# Environment variables for each field

Option 3 — Python

from lexigram.search.config import SearchConfig, BackendType
from lexigram.search import SearchModule

config = SearchConfig(backend_type=BackendType.MEILISEARCH, ...)
SearchModule.configure(config)

Config reference

Module Factory Methods

Key Features

• Protocol abstraction — Swap backends without changing application code
• Meilisearch — Typo-tolerant, instant search with faceting
• Elasticsearch — Full Lucene query DSL; aggregations; multi-index
• Typesense — Fast, schema-enforced search with scoped API keys
• PostgreSQL FTS — tsvector/tsquery via lexigram-sql; no extra infra
• MySQL FTS — FULLTEXT index support for MySQL / MariaDB
• MongoDB Text — Native $text operator with language stemming
• SQLite FTS5 — Local development with zero dependencies
• Cached search — Transparent result caching via CacheBackend
• Analytics — Query recording and hit-rate analytics for ranking improvement

Testing

async with Application.boot(modules=[SearchModule.stub()]) as app:
    # your test code
    ...

Key Source Files