Smart Hybrid Fallback

Users make mistakes. They misspell words, type in the wrong keyboard layout, or use partial words. A search engine that only returns exact matches will fail these users.

pg-smart-search implements a smart fallback chain to maximize result relevance.

The Fallback Chain

1. Linguistic Search (FTS): The fastest strategy. Uses dictionaries and indices to find root words (e.g., "running" → "run").
2. ILIKE: If FTS returns nothing, the engine falls back to simple substring matching. Useful for partial matches or names.
3. Trigram Fuzzy: If ILIKE fails, the engine uses PostgreSQLs pg_trgm to find words that are structurally similar. This handles typos (e.g., "laptpp" → "laptop").
4. Keyboard Layout Correction (ISO 9): If Trigram fails, the engine checks if the query makes sense if the user typed in a different layout (e.g., Russian layout on an English keyboard: "ghbdjy" → "привет").

Implementation

This happens transparently inside the engine.search() method. You dont need to write complex OR clauses or conditional logic.

const results = await engine.search({
  query: "laptpp", // Typo handled automatically
  language: "ru", // Enables ISO 9 layout correction for Russian
});

The engine will try the fastest methods first, and only engage heavier computations if the faster ones yield zero results.