159 lines
14 KiB
Plaintext
159 lines
14 KiB
Plaintext
|
/**
|
||
|
|
* Copyright (c) 2013-present, Facebook, Inc.
|
||
|
|
* All rights reserved.
|
||
|
|
*
|
||
|
|
* This source code is licensed under the BSD-style license found in the
|
||
|
|
* LICENSE file in the root directory of this source tree. An additional grant
|
||
|
|
* of patent rights can be found in the PATENTS file in the same directory.
|
||
|
|
*
|
||
|
|
* @providesModule UnicodeBidi
|
||
|
|
* @typechecks
|
||
|
|
* @flow
|
||
|
|
*/
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Basic (stateless) API for text direction detection
|
||
|
|
*
|
||
|
|
* Part of our implementation of Unicode Bidirectional Algorithm (UBA)
|
||
|
|
* Unicode Standard Annex #9 (UAX9)
|
||
|
|
* http://www.unicode.org/reports/tr9/
|
||
|
|
*/
|
||
|
|
|
||
|
|
'use strict';
|
||
|
|
|
||
|
|
const UnicodeBidiDirection = require('./UnicodeBidiDirection');
|
||
|
|
|
||
|
|
const invariant = require('./invariant');
|
||
|
|
|
||
|
|
import type { BidiDirection } from './UnicodeBidiDirection';
|
||
|
|
|
||
|
|
/**
|
||
|
|
* RegExp ranges of characters with a *Strong* Bidi_Class value.
|
||
|
|
*
|
||
|
|
* Data is based on DerivedBidiClass.txt in UCD version 7.0.0.
|
||
|
|
*
|
||
|
|
* NOTE: For performance reasons, we only support Unicode's
|
||
|
|
* Basic Multilingual Plane (BMP) for now.
|
||
|
|
*/
|
||
|
|
const RANGE_BY_BIDI_TYPE = {
|
||
|
|
|
||
|
|
L: 'A-Za-z\u00AA\u00B5\u00BA\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u01BA\u01BB' + '\u01BC-\u01BF\u01C0-\u01C3\u01C4-\u0293\u0294\u0295-\u02AF\u02B0-\u02B8' + '\u02BB-\u02C1\u02D0-\u02D1\u02E0-\u02E4\u02EE\u0370-\u0373\u0376-\u0377' + '\u037A\u037B-\u037D\u037F\u0386\u0388-\u038A\u038C\u038E-\u03A1' + '\u03A3-\u03F5\u03F7-\u0481\u0482\u048A-\u052F\u0531-\u0556\u0559' + '\u055A-\u055F\u0561-\u0587\u0589\u0903\u0904-\u0939\u093B\u093D' + '\u093E-\u0940\u0949-\u094C\u094E-\u094F\u0950\u0958-\u0961\u0964-\u0965' + '\u0966-\u096F\u0970\u0971\u0972-\u0980\u0982-\u0983\u0985-\u098C' + '\u098F-\u0990\u0993-\u09A8\u09AA-\u09B0\u09B2\u09B6-\u09B9\u09BD' + '\u09BE-\u09C0\u09C7-\u09C8\u09CB-\u09CC\u09CE\u09D7\u09DC-\u09DD' + '\u09DF-\u09E1\u09E6-\u09EF\u09F0-\u09F1\u09F4-\u09F9\u09FA\u0A03' + '\u0A05-\u0A0A\u0A0F-\u0A10\u0A13-\u0A28\u0A2A-\u0A30\u0A32-\u0A33' + '\u0A35-\u0A36\u0A38-\u0A39\u0A3E-\u0A40\u0A59-\u0A5C\u0A5E\u0A66-\u0A6F' + '\u0A72-\u0A74\u0A83\u0A85-\u0A8D\u0A8F-\u0A91\u0A93-\u0AA8\u0AAA-\u0AB0' + '\u0AB2-\u0AB3\u0AB5-\u0AB9\u0ABD\u0ABE-\u0AC0\u0AC9\u0ACB-\u0ACC\u0AD0' + '\u0AE0-\u0AE1\u0AE6-\u0AEF\u0AF0\u0B02-\u0B03\u0B05-\u0B0C\u0B0F-\u0B10' + '\u0B13-\u0B28\u0B2A-\u0B30\u0B32-\u0B33\u0B35-\u0B39\u0B3D\u0B3E\u0B40' + '\u0B47-\u0B48\u0B4B-\u0B4C\u0B57\u0B5C-\u0B5D\u0B5F-\u0B61\u0B66-\u0B6F' + '\u0B70\u0B71\u0B72-\u0B77\u0B83\u0B85-\u0B8A\u0B8E-\u0B90\u0B92-\u0B95' + '\u0B99-\u0B9A\u0B9C\u0B9E-\u0B9F\u0BA3-\u0BA4\u0BA8-\u0BAA\u0BAE-\u0BB9' + '\u0BBE-\u0BBF\u0BC1-\u0BC2\u0BC6-\u0BC8\u0BCA-\u0BCC\u0BD0\u0BD7' + '\u0BE6-\u0BEF\u0BF0-\u0BF2\u0C01-\u0C03\u0C05-\u0C0C\u0C0E-\u0C10' + '\u0C12-\u0C28\u0C2A-\u0C39\u0C3D\u0C41-\u0C44\u0C58-\u0C59\u0C60-\u0C61' + '\u0C66-\u0C6F\u0C7F\u0C82-\u0C83\u0C85-\u0C8C\u0C8E-\u0C90\u0C92-\u0CA8' + '\u0CAA-\u0CB3\u0CB5-\u0CB9\u0CBD\u0CBE\u0CBF\u0CC0-\u0CC4\u0CC6' + '\u0CC7-\u0CC8\u0CCA-\u0CCB\u0CD5-\u0CD6\u0CDE\u0CE0-\u0CE1\u0CE6-\u0CEF' + '\u0CF1-\u0CF2\u0D02-\u0D03\u0D05-\u0D0C\u0D0E-\u0D10\u0D12-\u0D3A\u0D3D' + '\u0D3E-\u0D40\u0D46-\u0D48\u0D4A-\u0D4C\u0D4E\u0D57\u0D60-\u0D61' + '\u0D66-\u0D6F\u0D70-\u0D75\u0D79\u0D7A-\u0D7F\u0D82-\u0D83\u0D85-\u0D96' + '\u0D9A-\u0DB1\u0DB3-\u0DBB\u0DBD\u0DC0-\u0DC6\u0DCF-\u0DD1\u0DD8-\u0DDF' + '\u0DE6-\u0DEF\u0DF2-\u0DF3\u0DF4\u0E01-\u0E30\u0E32-\u0E33\u0E40-\u0E45' + '\u0E46\u0E4F\u0E50-\u0E59\u0E5A-\u0E5B\u0E81-\u0E82\u0E84\u0E87-\u0E88' + '\u0E8A\u0E8D\u0E94-\u0E97\u0E99-\u0E9F\u0EA1-\u0EA3\u0EA5\u0EA7' + '\u0EAA-\u0EAB\u0EAD-\u0EB0\u0EB2-\u0EB3\u0EBD\u0EC0-\u0EC4\u0EC6' + '\u0ED0-\u0ED9\u0EDC-\u0EDF\u0F00\u0F01-\u0F03\u0F04-\u0F12\u0F13\u0F14' + '\u0F15-\u0F17\u0F1A-\u0F1F\u0F20-\u0F29\u0F2A-\u0F33\u0F34\u0F36\u0F38' + '\u0F3E-\u0F3F\u0F40-\u0F47\u0F49-\u0F6C\u0F7F\u0F85\u0F88-\u0F8C' + '\u0FBE-\u0FC5\u0FC7-\u0FCC\u0FCE-\u0FCF\u0FD0-\u0FD4\u0FD5-\u0FD8' + '\u0FD9-\u0FDA\u1000-\u102A\u102B-\u102C\u1031\u1038\u103B-\u103C\u103F' + '\u1040-\u1049\u104A-\u104F\u1050-\u1055\u1056-\u1057\u105A-\u105D\u1061' + '\u1062-\u1064\u1065-\u1066\u1067-\u106D\u106E-\u1070\u1075-\u1081' + '\u1083-\u1084\u1087-\u108C\u108E\u108F\u1090-\u1099\u109A-\u109C' + '\u109E-\u109F\u10A0-\u10C5\u10C7\u10CD\u10D0-\u10FA\u10FB\u10FC' + '\u10FD-\u1248\u124A-\u124D\u1250-\u1256\u1258\u125A-\u125D\u1260-\u1288' + '\u128A-\u128D\u1290-\u12B0\u12B2-\u12B5\u12B8-\u12BE\u12C0\u12C2-\u12C5' + '\u12C8-\u12D6\u12D8-\u1310\u1312-\u1315\u1318-\u135A\u1360-\u1368' + '\u1369-\u137C\u1380-\u138F\u13A0-\u13F4\u1401-\u166C\u166D-\u166E' + '\u166F-\u167F\u1681-\u169A\u16A0-\u16EA\u16EB-\u16ED\u16EE-\u16F0' + '\u16F1-\u16F8\u1700-\u170C\u170E-\u1711\u1720-\u1731\u1735-\u1736' + '\u1740-\u1751\u1760-\u176C\u176E-\u1770\u1780-\u17B3\u17B6\u17BE-\u17C5' + '\u17C7-\u17C8\u17D4-\u17D6\u17D7\u17D8-\u17DA\u17DC\u17E0-\u17E9' + '\u1810-\u1819\u1820-\u1842\u1843\u1844-\u1877\u1880-\u18A8\u18AA' + '\u18B0-\u18F5\u1900-\u191E\u1923-\u1926\u1929-\u192B\u1930-\u1931' + '\u1933-\u1938\u1946-\u194F\u1950-\u196D\u1970-\u1974\u1980-\u19AB' + '\u19B0-\u19C0\u19C1-\u19C7\u19C8-\u19C9\u19D0-\u19D9\u19DA\u1A00-\u1A16' + '\u1A19-\u1A1A\u1A1E-\u1A1F\u1A20-\u1A54\u1A55\u1A57\u1A61\u1A63-\u1A64' + '\u1A6D-\u1A72\u1A80-\u1A89\
|
||
|
|
|
||
|
|
R: '\u0590\u05BE\u05C0\u05C3\u05C6\u05C8-\u05CF\u05D0-\u05EA\u05EB-\u05EF' + '\u05F0-\u05F2\u05F3-\u05F4\u05F5-\u05FF\u07C0-\u07C9\u07CA-\u07EA' + '\u07F4-\u07F5\u07FA\u07FB-\u07FF\u0800-\u0815\u081A\u0824\u0828' + '\u082E-\u082F\u0830-\u083E\u083F\u0840-\u0858\u085C-\u085D\u085E' + '\u085F-\u089F\u200F\uFB1D\uFB1F-\uFB28\uFB2A-\uFB36\uFB37\uFB38-\uFB3C' + '\uFB3D\uFB3E\uFB3F\uFB40-\uFB41\uFB42\uFB43-\uFB44\uFB45\uFB46-\uFB4F',
|
||
|
|
|
||
|
|
AL: '\u0608\u060B\u060D\u061B\u061C\u061D\u061E-\u061F\u0620-\u063F\u0640' + '\u0641-\u064A\u066D\u066E-\u066F\u0671-\u06D3\u06D4\u06D5\u06E5-\u06E6' + '\u06EE-\u06EF\u06FA-\u06FC\u06FD-\u06FE\u06FF\u0700-\u070D\u070E\u070F' + '\u0710\u0712-\u072F\u074B-\u074C\u074D-\u07A5\u07B1\u07B2-\u07BF' + '\u08A0-\u08B2\u08B3-\u08E3\uFB50-\uFBB1\uFBB2-\uFBC1\uFBC2-\uFBD2' + '\uFBD3-\uFD3D\uFD40-\uFD4F\uFD50-\uFD8F\uFD90-\uFD91\uFD92-\uFDC7' + '\uFDC8-\uFDCF\uFDF0-\uFDFB\uFDFC\uFDFE-\uFDFF\uFE70-\uFE74\uFE75' + '\uFE76-\uFEFC\uFEFD-\uFEFE'
|
||
|
|
|
||
|
|
};
|
||
|
|
|
||
|
|
const REGEX_STRONG = new RegExp('[' + RANGE_BY_BIDI_TYPE.L + RANGE_BY_BIDI_TYPE.R + RANGE_BY_BIDI_TYPE.AL + ']');
|
||
|
|
|
||
|
|
const REGEX_RTL = new RegExp('[' + RANGE_BY_BIDI_TYPE.R + RANGE_BY_BIDI_TYPE.AL + ']');
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Returns the first strong character (has Bidi_Class value of L, R, or AL).
|
||
|
|
*
|
||
|
|
* @param str A text block; e.g. paragraph, table cell, tag
|
||
|
|
* @return A character with strong bidi direction, or null if not found
|
||
|
|
*/
|
||
|
|
function firstStrongChar(str: string): ?string {
|
||
|
|
const match = REGEX_STRONG.exec(str);
|
||
|
|
return match == null ? null : match[0];
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Returns the direction of a block of text, based on the direction of its
|
||
|
|
* first strong character (has Bidi_Class value of L, R, or AL).
|
||
|
|
*
|
||
|
|
* @param str A text block; e.g. paragraph, table cell, tag
|
||
|
|
* @return The resolved direction
|
||
|
|
*/
|
||
|
|
function firstStrongCharDir(str: string): BidiDirection {
|
||
|
|
const strongChar = firstStrongChar(str);
|
||
|
|
if (strongChar == null) {
|
||
|
|
return UnicodeBidiDirection.NEUTRAL;
|
||
|
|
}
|
||
|
|
return REGEX_RTL.exec(strongChar) ? UnicodeBidiDirection.RTL : UnicodeBidiDirection.LTR;
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Returns the direction of a block of text, based on the direction of its
|
||
|
|
* first strong character (has Bidi_Class value of L, R, or AL), or a fallback
|
||
|
|
* direction, if no strong character is found.
|
||
|
|
*
|
||
|
|
* This function is supposed to be used in respect to Higher-Level Protocol
|
||
|
|
* rule HL1. (http://www.unicode.org/reports/tr9/#HL1)
|
||
|
|
*
|
||
|
|
* @param str A text block; e.g. paragraph, table cell, tag
|
||
|
|
* @param fallback Fallback direction, used if no strong direction detected
|
||
|
|
* for the block (default = NEUTRAL)
|
||
|
|
* @return The resolved direction
|
||
|
|
*/
|
||
|
|
function resolveBlockDir(str: string, fallback: ?BidiDirection): BidiDirection {
|
||
|
|
fallback = fallback || UnicodeBidiDirection.NEUTRAL;
|
||
|
|
if (!str.length) {
|
||
|
|
return fallback;
|
||
|
|
}
|
||
|
|
const blockDir = firstStrongCharDir(str);
|
||
|
|
return blockDir === UnicodeBidiDirection.NEUTRAL ? fallback : blockDir;
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Returns the direction of a block of text, based on the direction of its
|
||
|
|
* first strong character (has Bidi_Class value of L, R, or AL), or a fallback
|
||
|
|
* direction, if no strong character is found.
|
||
|
|
*
|
||
|
|
* NOTE: This function is similar to resolveBlockDir(), but uses the global
|
||
|
|
* direction as the fallback, so it *always* returns a Strong direction,
|
||
|
|
* making it useful for integration in places that you need to make the final
|
||
|
|
* decision, like setting some CSS class.
|
||
|
|
*
|
||
|
|
* This function is supposed to be used in respect to Higher-Level Protocol
|
||
|
|
* rule HL1. (http://www.unicode.org/reports/tr9/#HL1)
|
||
|
|
*
|
||
|
|
* @param str A text block; e.g. paragraph, table cell
|
||
|
|
* @param strongFallback Fallback direction, used if no strong direction
|
||
|
|
* detected for the block (default = global direction)
|
||
|
|
* @return The resolved Strong direction
|
||
|
|
*/
|
||
|
|
function getDirection(str: string, strongFallback: ?BidiDirection): BidiDirection {
|
||
|
|
if (!strongFallback) {
|
||
|
|
strongFallback = UnicodeBidiDirection.getGlobalDir();
|
||
|
|
}
|
||
|
|
invariant(UnicodeBidiDirection.isStrong(strongFallback), 'Fallback direction must be a strong direction');
|
||
|
|
return resolveBlockDir(str, strongFallback);
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Returns true if getDirection(arguments...) returns LTR.
|
||
|
|
*
|
||
|
|
* @param str A text block; e.g. paragraph, table cell
|
||
|
|
* @param strongFallback Fallback direction, used if no strong direction
|
||
|
|
* detected for the block (default = global direction)
|
||
|
|
* @return True if the resolved direction is LTR
|
||
|
|
*/
|
||
|
|
function isDirectionLTR(str: string, strongFallback: ?BidiDirection): boolean {
|
||
|
|
return getDirection(str, strongFallback) === UnicodeBidiDirection.LTR;
|
||
|
|
}
|
||
|
|
|
||
|
|
/**
|
||
|
|
* Returns true if getDirection(arguments...) returns RTL.
|
||
|
|
*
|
||
|
|
* @param str A text block; e.g. paragraph, table cell
|
||
|
|
* @param strongFallback Fallback direction, used if no strong direction
|
||
|
|
* detected for the block (default = global direction)
|
||
|
|
* @return True if the resolved direction is RTL
|
||
|
|
*/
|
||
|
|
function isDirectionRTL(str: string, strongFallback: ?BidiDirection): boolean {
|
||
|
|
return getDirection(str, strongFallback) === UnicodeBidiDirection.RTL;
|
||
|
|
}
|
||
|
|
|
||
|
|
const UnicodeBidi = {
|
||
|
|
firstStrongChar: firstStrongChar,
|
||
|
|
firstStrongCharDir: firstStrongCharDir,
|
||
|
|
resolveBlockDir: resolveBlockDir,
|
||
|
|
getDirection: getDirection,
|
||
|
|
isDirectionLTR: isDirectionLTR,
|
||
|
|
isDirectionRTL: isDirectionRTL
|
||
|
|
};
|
||
|
|
|
||
|
|
module.exports = UnicodeBidi;
|