Split text into sentences with a vanilla
rule based approach (i.e working ~95% of the time).
http://tessmore.github.io/sbd/
$ npm install sbd
$ yarn add sbd
var tokenizer = require('sbd');
var optional_options = {};
var text = "On Jan. 20, former Sen. Barack Obama became the 44th President of the U.S. Millions attended the Inauguration.";
var sentences = tokenizer.sentences(text, optional_options);
// [
// 'On Jan. 20, former Sen. Barack Obama became the 44th President of the U.S.',
// 'Millions attended the Inauguration.',
// ]
var options = {
"newline_boundaries" : false,
"html_boundaries" : false,
"sanitize" : false,
"allowed_tags" : false,
"preserve_whitespace" : false,
"abbreviations" : null
};
newline_boundaries
, force sentence split at newlineshtml_boundaries
, force sentence split at specific tags (br, and closing p, div, ul, ol)sanitize
: If you don't expect nor want html in your text.allowed_tags
: To sanitize html, the library santize-html is used. You can pass the allowed tags option.preserve_whitespace
: Preserve the literal whitespace between words and sentences (otherwise, internal spaces are normalized to a single space char, and inter-sentence whitespace is omitted). Preserve whitespace has no effect if either newline_boundaries or html_boundaries is specified.abbreviations
: list of abbreviations to override the original ones for use with other languages. Don't put dots in your custom abbreviations.You can run unit tests with npm test
.
If you feel something is missing, you can open an issue stating the problem sentence and desired result. If code is unclear give me a @mention. Pull requests are welcome.
npm install -g browserify
npm run-script build