A powerful and flexible text search library for JavaScript that enables you to build a simple text search engine.
npm install @basd/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.
Install the package with:
``bash`
npm install @basd/search
First, import the Search library.
`js`
import Search from '@basd/search'
orjs`
const Search = require('@basd/search')
Here's how to create a simple search evaluator and use it.
`js
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:
`js
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
`
#### SearchFactory
Factory class to produce instances of Tokenizer, Parser, and Interpreter.
`js`
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.
`js`
const normalizedText = SearchNormalizer.normalize('some text')
#### SearchTokenizer
Tokenizes the normalized query.
`js`
const tokenizer = new SearchTokenizer()
const tokens = tokenizer.tokenize('apple AND orange')
#### SearchParser
Parses the tokens into an AST.
`js`
const parser = new SearchParser(tokens)
const ast = parser.parse()
#### SearchInterpreter
Interprets the AST against a given text.
`js`
const interpreter = new SearchInterpreter(ast)
const result = interpreter.interpret('I have an apple.')
#### Search
The main class that combines all the functionalities.
`js`
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.
The library is designed to be easily extendable. You can extend SearchTokenizer, SearchParser, and SearchInterpreter to add additional functionalities.
#### 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.
Takes a query string and returns an array of tokens.
Takes an array of tokens and returns a binary AST.
Takes a string of text and returns a boolean indicating whether it matches the AST.
In order to run the test suite, simply clone the repository and install its dependencies:
`bash`
git clone https://gitlab.com/frenware/framework/plaindb/search.git
cd search
npm install
To run the tests:
`bash`
npm test
Thank you! Please see our contributing guidelines for details.
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
@basd/search is MIT licensed.