Search

A powerful and flexible text search library for JavaScript that enables you to build a simple text search engine. It provides a set of classes to tokenize, parse, and interpret queries using a binary AST (Abstract Syntax Tree). The library supports various grouping operators (and/or/&/|) and any degree of parenthesis nesting.

Features

• Tokenization of search queries
• Parsing to Abstract Syntax Trees (AST)
• Interpretation to evaluate search queries against text
• Normalization of text and query strings
• Abstract factory for easy extension

Installation

Install the package with:

npm install @basd/search

Usage

First, import the Search library.

import Search from '@basd/search'

or

const Search = require('@basd/search')

Quick Start

Here's how to create a simple search evaluator and use it.

const Search = require('@basd/search')

const search = new Search()
const evaluator = search.evaluator('apple AND orange')

const result = evaluator('I have an apple and an orange.')
// Returns true

Here's a basic example of how you can use @basd/search to perform a text search:

const { Tokenizer, Parser, Interpreter } = require('@basd/search')

const query = 'apple AND orange OR pear'
const tokenizer = new Tokenizer()
const tokens = tokenizer.tokenize(query)

const parser = new Parser(tokens)
const ast = parser.parse()

const interpreter = new Interpreter(ast)
const result = interpreter.interpret('apple orange') // true

Documentation

• API Reference

API Reference

Classes

SearchFactory

Factory class to produce instances of Tokenizer, Parser, and Interpreter.

const factory = new SearchFactory(registry)

Methods

• createTokenizer(...args): Creates a SearchTokenizer instance.
• createParser(...args): Creates a SearchParser instance.
• createInterpreter(...args): Creates a SearchInterpreter instance.

SearchNormalizer

Normalizes text to be used in tokenization and interpretation.

const normalizedText = SearchNormalizer.normalize('some text')

SearchTokenizer

Tokenizes the normalized query.

const tokenizer = new SearchTokenizer()
const tokens = tokenizer.tokenize('apple AND orange')

SearchParser

Parses the tokens into an AST.

const parser = new SearchParser(tokens)
const ast = parser.parse()

SearchInterpreter

Interprets the AST against a given text.

const interpreter = new SearchInterpreter(ast)
const result = interpreter.interpret('I have an apple.')

Search

The main class that combines all the functionalities.

const search = new Search()

Methods

• evaluator(needle): Returns an evaluator function for a given search query.
• evaluate(needle, haystack): Evaluates a search query against a given text.

Extending the Library

The library is designed to be easily extendable. You can extend SearchTokenizer, SearchParser, and SearchInterpreter to add additional functionalities.

Classes

TextNormalizer

Normalizes text by removing punctuations, converting to uppercase, and replacing multiple spaces with a single space.

Tokenizer

Tokenizes a query into distinct elements such as words, operators, and parentheses.

Parser

Takes the tokens and turns them into a binary AST.

Interpreter

Takes the AST and matches a given text string against it.

API Reference

Tokenizer.tokenize(query: string): Token[]

Takes a query string and returns an array of tokens.

Parser.parse(): ASTNode

Takes an array of tokens and returns a binary AST.

Interpreter.interpret(data: string): boolean

Takes a string of text and returns a boolean indicating whether it matches the AST.

Tests

In order to run the test suite, simply clone the repository and install its dependencies:

git clone https://gitlab.com/frenware/framework/plaindb/search.git
cd search
npm install

To run the tests:

npm test

Contributing

Thank you! Please see our contributing guidelines for details.

Donations

If you find this project useful and want to help support further development, please send us some coin. We greatly appreciate any and all contributions. Thank you!

Bitcoin (BTC):

1JUb1yNFH6wjGekRUW6Dfgyg4J4h6wKKdF

Monero (XMR):

46uV2fMZT3EWkBrGUgszJCcbqFqEvqrB4bZBJwsbx7yA8e2WBakXzJSUK8aqT4GoqERzbg4oKT2SiPeCgjzVH6VpSQ5y7KQ

License

@basd/search is MIT licensed.