Lexicor #

Lexicor is a strictly typed, highly optimized, offline interface for the Open English WordNet.

It provides a high-performance, embedded SQL engine to query English definitions, synonyms, antonyms, hypernyms, and more, without relying on external APIs or raw text parsing.

✨ Features #

• 🚀 Ultra Fast: Microsecond-level lookups (~20µs on disk). Powered by a custom SQLite database using WITHOUT ROWID optimizations and specific covering indexes.
• 🔒 Strictly Typed: No magic strings or integers. Work with Concept, SpeechPart, RelationType, and DomainCategory objects.
• 🧠 Morphology Aware: Automatically handles stem resolution. Searching for "ran" matches "run"; "better" matches "good".
• ⚡ Dual Modes:
  • Disk Mode: Instant startup (<25ms), low memory usage.
  • Memory Mode: Loads DB into RAM for nanosecond-level query speeds.
• 🔗 Rich Relations: Distinguishes between Semantic relations (Concept-to-Concept) and Lexical relations (Word-to-Word).

📦 Installation #

Add lexicor to your pubspec.yaml:

dependencies:
  lexicor: ^0.1.0

🚀 Quick Start #

import 'package:lexicor/lexicor.dart';

void main() async {
  // 1. Initialize (Disk mode is default)
  final lexicor = await Lexicor.init();

  // 2. Lookup a word
  final result = lexicor.lookup('bank');
  
  print('Found ${result.concepts.length} concepts for "${result.query}"');

  // 3. Iterate concepts
  for (final concept in result.concepts) {
    print('[${concept.part.label}] ${concept.category.label}');
    
    // 4. Get relationships (Hypernyms, Parts, Antonyms...)
    final relations = lexicor.related(concept);
    
    for (final rel in relations.withRelation(RelationType.hypernym)) {
      print('  -> is a type of: ${rel.word}');
    }
  }

  // 5. Cleanup
  lexicor.close();
}

📖 Usage Guide #

Initialization Modes #

Lexicor offers two ways to load the database via StorageMode:

// 1. StorageMode.onDisk (Default)
// Instant startup (~25ms). Queries take ~20-50µs. 
// Best for CLI tools and Mobile apps.
final db = await Lexicor.init(mode: StorageMode.onDisk);

// 2. StorageMode.inMemory
// Slower startup (~100ms copy time) but faster queries (~15µs). 
// Best for backend servers or heavy batch processing.
final db = await Lexicor.init(mode: StorageMode.inMemory);

Flutter Integration #

Because Flutter assets are packed into the app bundle, sqlite3 cannot open them directly from the bundle. You must copy the asset to a file path first (e.g., using path_provider).

// 1. Copy 'dictionary.sqlite' from assets to Application Documents Directory.
// 2. Pass that path to Lexicor:
final lexicor = await Lexicor.init(
  customPath: '/path/to/app_documents/dictionary.sqlite',
);

Morphology #

Lexicor automatically resolves word forms. You don't need to manually stem words.

// The user types "running"
final result = lexicor.lookup('running');

// Lexicor automatically searches for "run"
print(result.resolvedForms); // ['running', 'run']

If you need raw access to morphology:

final root = lexicor.morphology('better', SpeechPart.adjective);
print(root); // "good"

Relationships #

WordNet distinguishes between two types of links:

1. Semantic (Concept-to-Concept): e.g., A Dog is an Animal.
2. Lexical (Word-to-Word): e.g., Slow is the antonym of Fast.

lexicor.related() returns both, but you can filter them:

final rels = lexicor.related(concept);

// Get synonyms, hypernyms, etc.
final semantic = rels.semanticOnly;

// Get antonyms, derivations, etc.
final lexical = rels.lexicalOnly;

// Filter by specific type
final parts = rels.byType(RelationType.partMeronym);

📊 Database Stats #

Lexicor uses a highly optimized database structure. Unlike raw SQL dumps (often 100MB+), Lexicor is compressed to ~27 MB while maintaining full relationship graphs.

📂 Database Source & License #

This package includes a compressed, optimized version of Open English WordNet 2025 (v2.3.2).

• Source: Open English Wordnet in Sqlite form
• Database License: CC BY 4.0 (Open English WordNet).
• Package License: MIT.

Using this package in your commercial app requires you to attribute the Open English WordNet project in your app's About/License section.