Node.js bindings to the ICU (50.1) project boundary analysis module. Useful for breaking/splitting up words in many many international languages.
First, install libicu from source (so you have all the needed headers and binaries)
wget http://download.icu-project.org/files/icu4c/50.1/icu4c-50_1-src.tgztar -xvzf icu4c-50_1-src.tgzcd icu/source./configuremakemake install # or sudo make installldconfig # refresh linker cache
This will take a while. After it finishes you should have all the required ICU headers and binaries in your default build path ready to be used.
Afterwards, switch to your project folder and
npm install icu-wordsplit
And you should now be able to
require('icu-wordsplit') in your code.
Right now the only function call available is exported from the module.
Splits the specified string into words using ICU-defined word boundary.
var splitWords = require('icu-wordsplit'); var results = splitWords('The quick brown fox jumps over the lazy dog.'); console.log(results); results = splitWords('th_TH', 'แยกคำภาษาไทยก็ทำได้นะจ้ะ'); console.log(results);
The first argument is the locale. You must specify an ICU-compatible locale name here.
However, this argument is now optional as you can use
en_US to split most
strings without problem.
The second argument is the string to which to split.
The function will returns an array of words. Whitespaces and some common punctuations may be automatically removed from the list.
All contributions welcome!
Compile and run tests with
Here's something to do: