Access Unicode character data from UnicodeData.txt

Downloads in past


719.0.0-alpha.68 years ago8 years agoMinified + gzip package size for unicharadata in KB


unicharadata module
JavaScript module for accessing Unicode character data from UnicodeData.txt.

For nodejs

Install with npm command as:
npm install unicharadata

then, load it in js files as:
const unicharadata = require("unicharadata");

Option: Using prebuilt database version

NOTE: For file size reducing, The current module implementation embeds UnicodeData.txt source and builds complete character database on load time.
This method uses large pre built database json file. It would be about 1sec fatser than building database (around 4sec).
const unicharadata = require("unicharadata/trial/unicharadata-load");

For browsers

Load with standard script tag as:
<script src="unicharadata.js"></script>

see: browser-example.html ( demo )
Links via CDN
  • https://raw.githack.com/bellbind/unicharadata/master/unicharadata.js
  • https://rawgit.com/bellbind/unicharadata/master/unicharadata.js


NOTE: Detail of the character propeties are in http://www.unicode.org/reports/tr44/#UnicodeData.txt
  • unicharadata.lookup(name, defaultch = ""): search character from its name
- use "Name" otherwise "Unicode 1 Name" property
  • unicharadata.lookupname(ch, defaultname = ""): get the name used lookup(name)
  • unicharadata.name(ch, defaultname = ""): get "Name" string
  • unicharadata.category(ch): get "General Category" string
- see [value table](http://www.unicode.org/reports/tr44/#General_Category_Values)
  • unicharadata.combining(ch): get "Canonical Combining Class" number
- see [value table](http://www.unicode.org/reports/tr44/#Canonical_Combining_Class_Values)
  • unicharadata.bidirectional(ch): get "Bidi Class" string
- see [value table](http://www.unicode.org/reports/tr44/#Bidi_Class_Values)
  • unicharadata.decomposition(ch): get "Decomposition Mapping" text
- return `ch` itself when no decomposition mapping
  • unicharadata.decompositionTag(ch): get "Decomposition Type" tag string
- see [value table](http://www.unicode.org/reports/tr44/#Character_Decomposition_Mappings)
  • unicharadata.decimal(ch, defaultval = NaN): get "Numeric Value" integer as single decimal type
  • unicharadata.digit(ch, defaultval = NaN): get "Numeric Value" integer as single digit type
- overwrapped `decimal(ch)`. this includes as superscript (e.g. U+00B9)
  • unicharadata.numeric(ch, defaultval = NaN): get "Numeric Value" float number as numeric type
- overwrapped `digit(ch)`, this includes non single digit numbers.
  • unicharadata.mirrored(ch): get "Bidi Mirrored" string
- `"Y"` or `"N"`
  • unicharadata.unicode1name(ch, defaultname = ""): get "Unicode 1 Name" string
  • unicharadata.isocomment(ch, defaultcomment = ""): get "ISO Comment" string
  • unicharadata.upper(ch, defaultch = ""): get "Simple Uppercase Mapping" character
  • unicharadata.lower(ch, defaultch = ""): get "Simple Lowercase Mapping" character
  • unicharadata.title(ch, defaultch = ""): get "Simple Titlecase Mapping" character
  • unicharadata.splitCombined(text): split a text to an array of strings that contains a character and following combining characters.


"use strict";

const unicharadata = require("unicharadata");

// API
console.assert(unicharadata.name("A") === "LATIN CAPITAL LETTER A");
console.assert(unicharadata.lookup("LATIN CAPITAL LETTER A") === "A");
console.assert(unicharadata.category("A") === "Lu");
console.assert(unicharadata.combining("A") === 0);
console.assert(unicharadata.bidirectional("A") === "L");
console.assert(unicharadata.decomposition("A") === "A");
console.assert(unicharadata.decompositionTag("A") === "");
console.assert(unicharadata.mirrored("A") === "N");
console.assert(unicharadata.unicode1name("A") === "");
console.assert(unicharadata.lookupname("A") === "LATIN CAPITAL LETTER A");
console.assert(unicharadata.isocomment("A") === "");
console.assert(unicharadata.upper("A") === "");
console.assert(unicharadata.lower("A") === "a");
console.assert(unicharadata.title("A") === "");

const splitted = unicharadata.splitCombined("γ“γ‚ŒγŒπŸ”‘γ§γ™".normalize("NFD"));
console.assert(splitted[0] === "こ");
console.assert(splitted[1] === "γ‚Œ");
console.assert(splitted[2] === "が".normalize("NFD"));
console.assert(splitted[3] === "πŸ”‘");
console.assert(splitted[4] === "で".normalize("NFD"));
console.assert(splitted[5] === "す");

// specific
console.assert(unicharadata.decomposition("㍍") === "γƒ‘γƒΌγƒˆγƒ«");
console.assert(unicharadata.decompositionTag("㍍") === "<square>");

console.assert(unicharadata.decimal("οΌ‘") === 1);
console.assert(unicharadata.digit("β‘£") === 4);
console.assert(unicharadata.numeric("β…›") === 1 / 8);
console.assert(unicharadata.unicode1name("\n") === "LINE FEED (LF)");
console.assert(unicharadata.lookupname("\n") === "LINE FEED (LF)");


  1. npm run download: update UnicodeData.json from UnicodeData.txt on the web
  2. npm run build: update unicharadata.js from unicharadata-raw.js and
  1. npm test: run tests
  2. npm run eslint: check coding style with eslint
- setup with `npm install` required
The package version is based on the Unicode "Version" of UnicodeData.txt.
  • e.g. 9.0.0-alpha.2: the Unicode version is 9.0.0


ISC License


  • unicharadata website: https://github.com/bellbind/unicharadata
  • Unicode Character Data: http://unicode.org/ucd/
- Spec (Unicode Standard Annex #44): http://unicode.org/reports/tr44/
- Data: http://unicode.org/Public/UCD/latest/ucd/
  • Python builtin unicodedata lib: https://docs.python.org/3/library/unicodedata.html
- function names borrowed from the library.