Simple Spellchecker

A simple and fast spellchecker with spelling suggestions and Electron's integration

Features

Simple Spellchecker is a spellchecker module for Node.js, that allows to check if a word is misspelled and to get spelling suggestions.

It comes with dictionaries for English, Spanish, French, German and Dutch, but you can easily add more languages by simply defining a text file with a list of valid words.

It also has a CLI tool that allows to check words from the command line.

Usage

In order to use the module, you must first install it using NPM.

npm install simple-spellchecker

And then require the module, get a Dictionary object and invoke their methods.

var SpellChecker = require('simple-spellchecker');
SpellChecker.getDictionary("fr-FR", function(err, dictionary) {
    if(!err) {
        var misspelled = ! dictionary.spellCheck('maisonn');
        if(misspelled) {
            var suggestions = dictionary.getSuggestions('maisonn');
        }
    }
});    

Methods

Module

The module has three public methods: getDictionary(), getDictionarySync() and normalizeDictionary().

getDictionary(fileName [, folderPath], callback)

This method allows to get a Dictionary instance from a file.

Parameters:

• fileName: The name of the dictionary's file. The method is going to first search a file with .dic extension, if not found, then is going to search a .zip and is going to unzip it.
• folderPath: The folder in which the dictionary's file is located. This parameter is optional, by default it assumes that the file is in the dict folder.
• callback: A function which is going to be invoked for return the Dictionary object.

Example:

var SpellChecker = require('simple-spellchecker');
SpellChecker.getDictionary("fr-FR", function(err, result) {
    var dictionary = result;
});    

getDictionarySync(fileName [, folderPath])

This method allows to get a Dictionary instance from a file, in a synchronous way.

Parameters:

• fileName: The name of the dictionary's file. The file must have .dic extension.
• folderPath: The folder in which the dictionary's file is located. This parameter is optional, by default it assumes that the file is in the dict folder.

Returns:

• A Dictionary object.

Example:

var SpellChecker = require('simple-spellchecker');
var dictionary = SpellChecker.getDictionarySync("fr-FR");    

normalizeDictionary(inputPath [, outputPath], callback)

This methods reads a UTF-8 dictionary file, removes the BOM and \r characters and sorts the list of words.

Parameters:

• inputPath: The path of the dictionary file.
• outputPath: The path for the normalized dictionary file. This parameter is optional, by deafult the original file is overwritten.
• callback: A function which is going to be invoked when the process has finished.

Example:

var SpellChecker = require('simple-spellchecker');
SpellChecker.normalizeDictionary(inputFile, outputFile, function(err, success) {
    if(success) console.log("The file was normalized");
}); 

Dictionary

The Dictionary class has six public methods: getLength(), setWordlist(), spellCheck(), isMisspelled(), getSuggestions() and checkAndSuggest()

getLength()

This method allows to get the quantity of words that the dictionary has.

Returns:

• An integer with the number of words.

setWordlist(wordlist)

This method allows to set the words of the dictionary.

Parameter:

• wordlist: an array of strings.

spellCheck(word)

This method allows to verify is a word is correctly written or not.

Parameter:

• word: the word to verify.

Returns:

• true if the word is in the dictionary, false if not.

isMisspelled(word)

This method allows to verify is a word is misspelled or not.

Parameter:

• word: the word to verify.

Returns:

• true if the word is misspelled, false if not

getSuggestions(word [, limit] [, maxDistance])

This method allows to get spelling suggestions for a word.

Parameters:

• word: the word used to generate the suggestions.
• limit: the maximum number of suggestions to get (by default, 5).
• maxDistance: the maximum edit distance that a word can have from the word parameter, in order to being considered as a valid suggestion (by default, 2).

Returns:

• An array of strings.

checkAndSuggest(word [, limit] [, maxDistance])

This method allows to verify if a word is misspelled and to get spelling suggestions.

Parameters:

• word: the word to verify.
• limit: the maximum number of suggestions to get (by default, 5).
• maxDistance: the maximum edit distance that a word can have from the word parameter, in order to being considered as a valid suggestion (by default, 2).

Returns:

• An object with the fields misspelled, which contains a boolean, and suggestions, which contains an array of strings.

addRegex(regex)

This method adds a regular expression that will be used to verify if a word is valid even though is not in the dictionary.

This might be useful when avoiding marking special words as misspelled, such as numbers, emails, or URL addresses.

Parameters:

• regex: a regular expression object.

clearRegex

This method removes all previous regular expressions added using the method addRegex().

Add dictionaries

In order to use custom dictionaries, you must define a text file with a list of valid words, where each word is separated by a new line.

File's name

The file's extension must be .dic, and the name should (preferably) be composed by the language code and the region designator (e.g. es-AR if the language is Spanish and the region is Argentina).

Optionally you can also pack the file in a ZIP package, the module is going to be able to unzip it and read it as long as the .zip file has the same name has the .dic file (e.g. a file es-AR.zip that contains the file es-AR.dic).

File's encoding

The file must be encoded in UTF8 (without BOM), the words must be separated with a Line Feed (i.e. \n) and not with a Carriage Return plus a Line Feed (i.e. \r\n), and the words must be sorted in ascending order.

The module can remove all unwanted characters and sort the words, if you either invoke the normalize() method or pack the file in a ZIP file (the module automatically calls the normalize() method after unzip it).

CLI tools

The module comes with a script that allows to normalize dictionaries and to test them.

Test

In order to test a dictionary file, you must execute the script indicating the folder and the name of the dictionary's file and the word that you want to test.

For example, the following sentence will search in the dict folder a dictionary which is either in the file en-GB.dic or en-GB.zip, and is going to verify if the word house is misspelled or not and is going to search some spelling suggestions.

node cli.js check "./dict" en-GB house

Normalize

In order to normalize a dictionary file, you must execute the script indicating the file's path:

node cli.js normalize "./dict/en-GB.dic"

If you don't want to override the original file, you can specify the path for an output file:

node cli.js normalize "./dict/en-GB.dic" "C:/output/en-GB.dic"

Electron's integration

There are several ways in which you can integrate this module with Electron. But if you want a starting point, the following example shows how to enable the English dictionary:

1) Install the module as a dependency of your project (npm install simple-spellchecker --save).

2) In the main process, require the module, load the dictionary, and define a function to access to his methods:

// Initialization.
var SpellChecker = require('simple-spellchecker');
var myDictionary = null;

// Load dictionary.
SpellChecker.getDictionary("en-US", "./node_modules/simple-spellchecker/dict", function(err, result) {
    if(!err) {
        myDictionary = result;
    }
});

// Define function for consult the dictionary.
ipcMain.on('checkspell', function(event, word) {
    var res = null;
    if(myDictionary != null && word != null) {
        res = myDictionary.spellCheck(word);
    }
    event.returnValue = res;
});

3) In the renderer process, define a spell checker provider that uses the previously loaded dictionary.

// Get web frame.
var webFrame = require('electron').webFrame;

webFrame.setSpellCheckProvider("en-US", false, {
    spellCheck: function(text) {      
        var res = ipcRenderer.sendSync('checkspell', text);
        return res != null? res : true;
    }
});

Unit testing

The module is partially tested by the file test.js (using Mocha and Chai). In order to run the tests, just execute the command npm test.

License

Simple Spellchecker is free software; you can redistribute it and/or modify it under the terms of the Mozilla Public License v2.0. You should have received a copy of the MPL 2.0 along with this library, otherwise you can obtain one at http://mozilla.org/MPL/2.0/.