gfx/thebes/nsUnicodeRange.cpp
author Ehsan Akhgari <ehsan@mozilla.com>
Mon, 07 Oct 2013 19:15:59 -0400
changeset 164848 47b5355992adc942e1a059c572703becc40bcdb0
parent 108991 a16372ce30b5f6b747246b01fcd215a4bf3b6342
child 220697 3d88e3845d62f187474f6dc346f71f600cbb125f
permissions -rw-r--r--
Bug 924221 - Minimize the #includes in gfx/thebes; r=bjacob

/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* This Source Code Form is subject to the terms of the Mozilla Public
 * License, v. 2.0. If a copy of the MPL was not distributed with this
 * file, You can obtain one at http://mozilla.org/MPL/2.0/. */

#include "nsUnicodeRange.h"
#include "nsGkAtoms.h"
#include "mozilla/NullPtr.h"

// This table depends on unicode range definitions. 
// Each item's index must correspond unicode range value
// eg. x-cyrillic = LangGroupTable[kRangeCyrillic]
static nsIAtom **gUnicodeRangeToLangGroupAtomTable[] =
{
  &nsGkAtoms::x_cyrillic,
  &nsGkAtoms::el_,
  &nsGkAtoms::tr,
  &nsGkAtoms::he,
  &nsGkAtoms::ar,
  &nsGkAtoms::x_baltic,
  &nsGkAtoms::th,
  &nsGkAtoms::ko,
  &nsGkAtoms::Japanese,
  &nsGkAtoms::zh_cn,
  &nsGkAtoms::zh_tw,
  &nsGkAtoms::x_devanagari,
  &nsGkAtoms::x_tamil,
  &nsGkAtoms::x_armn,
  &nsGkAtoms::x_beng,
  &nsGkAtoms::x_cans,
  &nsGkAtoms::x_ethi,
  &nsGkAtoms::x_geor,
  &nsGkAtoms::x_gujr,
  &nsGkAtoms::x_guru,
  &nsGkAtoms::x_khmr,
  &nsGkAtoms::x_mlym,
  &nsGkAtoms::x_orya,
  &nsGkAtoms::x_telu,
  &nsGkAtoms::x_knda,
  &nsGkAtoms::x_sinh,
  &nsGkAtoms::x_tibt
};

/**********************************************************************
 * Unicode subranges as defined in unicode 3.0
 * x-western, x-central-euro, tr, x-baltic  -> latin 
 *  0000 - 036f 
 *  1e00 - 1eff
 *  2000 - 206f  (general punctuation)
 *  20a0 - 20cf  (currency symbols)
 *  2100 - 214f  (letterlike symbols)
 *  2150 - 218f  (Number Forms)
 * el         -> greek
 *  0370 - 03ff
 *  1f00 - 1fff
 * x-cyrillic -> cyrillic
 *  0400 - 04ff
 * he         -> hebrew
 *  0590 - 05ff
 * ar         -> arabic
 *  0600 - 06ff
 *  fb50 - fdff (arabic presentation forms)
 *  fe70 - feff (arabic presentation forms b)
 * th - thai
 *  0e00 - 0e7f
 * ko        -> korean
 *  ac00 - d7af  (hangul Syllables)
 *  1100 - 11ff    (jamo)
 *  3130 - 318f (hangul compatibility jamo)
 * ja
 *  3040 - 309f (hiragana)
 *  30a0 - 30ff (katakana)
 * zh-CN
 * zh-TW
 *
 * CJK
 *  3100 - 312f (bopomofo)
 *  31a0 - 31bf (bopomofo extended)
 *  3000 - 303f (CJK Symbols and Punctuation) 
 *  2e80 - 2eff (CJK radicals supplement)
 *  2f00 - 2fdf (Kangxi Radicals)
 *  2ff0 - 2fff (Ideographic Description Characters)
 *  3190 - 319f (kanbun)
 *  3200 - 32ff (Enclosed CJK letters and Months)
 *  3300 - 33ff (CJK compatibility)
 *  3400 - 4dbf (CJK Unified Ideographs Extension A)
 *  4e00 - 9faf (CJK Unified Ideographs)
 *  f900 - fa5f (CJK Compatibility Ideographs)
 *  fe30 - fe4f (CJK compatibility Forms)
 *  ff00 - ffef (halfwidth and fullwidth forms)
 *
 * Armenian
 *  0530 - 058f 
 * Sriac 
 *  0700 - 074f
 * Thaana
 *  0780 - 07bf
 * Devanagari
 *  0900 - 097f
 * Bengali
 *  0980 - 09ff
 * Gurmukhi
 *  0a00 - 0a7f
 * Gujarati
 *  0a80 - 0aff
 * Oriya
 *  0b00 - 0b7f
 * Tamil
 *  0b80 - 0bff
 * Telugu
 *  0c00 - 0c7f
 * Kannada
 *  0c80 - 0cff
 * Malayalam
 *  0d00 - 0d7f
 * Sinhala
 *  0d80 - 0def
 * Lao
 *  0e80 - 0eff
 * Tibetan
 *  0f00 - 0fbf
 * Myanmar
 *  1000 - 109f
 * Georgian
 *  10a0 - 10ff
 * Ethiopic
 *  1200 - 137f
 * Cherokee
 *  13a0 - 13ff
 * Canadian Aboriginal Syllabics
 *  1400 - 167f
 * Ogham
 *  1680 - 169f
 * Runic 
 *  16a0 - 16ff
 * Khmer
 *  1780 - 17ff
 * Mongolian
 *  1800 - 18af
 * Misc - superscripts and subscripts
 *  2070 - 209f
 * Misc - Combining Diacritical Marks for Symbols
 *  20d0 - 20ff
 * Misc - Arrows
 *  2190 - 21ff
 * Misc - Mathematical Operators
 *  2200 - 22ff
 * Misc - Miscellaneous Technical
 *  2300 - 23ff
 * Misc - Control picture
 *  2400 - 243f
 * Misc - Optical character recognition
 *  2440 - 2450
 * Misc - Enclose Alphanumerics
 *  2460 - 24ff
 * Misc - Box Drawing 
 *  2500 - 257f
 * Misc - Block Elements
 *  2580 - 259f
 * Misc - Geometric Shapes
 *  25a0 - 25ff
 * Misc - Miscellaneous Symbols
 *  2600 - 267f
 * Misc - Dingbats
 *  2700 - 27bf
 * Misc - Braille Patterns
 *  2800 - 28ff
 * Yi Syllables
 *  a000 - a48f
 * Yi radicals
 *  a490 - a4cf
 * Alphabetic Presentation Forms
 *  fb00 - fb4f
 * Misc - Combining half Marks
 *  fe20 - fe2f
 * Misc - small form variants
 *  fe50 - fe6f
 * Misc - Specials
 *  fff0 - ffff
 *********************************************************************/



#define NUM_OF_SUBTABLES      10
#define SUBTABLE_SIZE         16

static const uint8_t gUnicodeSubrangeTable[NUM_OF_SUBTABLES][SUBTABLE_SIZE] = 
{ 
  { // table for X---
    kRangeTableBase+1,  //u0xxx
    kRangeTableBase+2,  //u1xxx
    kRangeTableBase+3,  //u2xxx
    kRangeSetCJK,       //u3xxx
    kRangeSetCJK,       //u4xxx
    kRangeSetCJK,       //u5xxx
    kRangeSetCJK,       //u6xxx
    kRangeSetCJK,       //u7xxx
    kRangeSetCJK,       //u8xxx
    kRangeSetCJK,       //u9xxx
    kRangeTableBase+4,  //uaxxx
    kRangeKorean,       //ubxxx
    kRangeKorean,       //ucxxx
    kRangeTableBase+5,  //udxxx
    kRangePrivate,      //uexxx
    kRangeTableBase+6   //ufxxx
  },
  { //table for 0X--
    kRangeSetLatin,          //u00xx
    kRangeSetLatin,          //u01xx
    kRangeSetLatin,          //u02xx
    kRangeGreek,             //u03xx     XXX 0300-036f is in fact kRangeCombiningDiacriticalMarks
    kRangeCyrillic,          //u04xx
    kRangeTableBase+7,       //u05xx, includes Cyrillic supplement, Hebrew, and Armenian
    kRangeArabic,            //u06xx
    kRangeTertiaryTable,     //u07xx
    kRangeUnassigned,        //u08xx
    kRangeTertiaryTable,     //u09xx
    kRangeTertiaryTable,     //u0axx
    kRangeTertiaryTable,     //u0bxx
    kRangeTertiaryTable,     //u0cxx
    kRangeTertiaryTable,     //u0dxx
    kRangeTertiaryTable,     //u0exx
    kRangeTibetan            //u0fxx
  },
  { //table for 1x--
    kRangeTertiaryTable,     //u10xx
    kRangeKorean,            //u11xx
    kRangeEthiopic,          //u12xx
    kRangeTertiaryTable,     //u13xx
    kRangeCanadian,          //u14xx
    kRangeCanadian,          //u15xx
    kRangeTertiaryTable,     //u16xx
    kRangeKhmer,             //u17xx
    kRangeMongolian,         //u18xx
    kRangeUnassigned,        //u19xx
    kRangeUnassigned,        //u1axx
    kRangeUnassigned,        //u1bxx
    kRangeUnassigned,        //u1cxx
    kRangeUnassigned,        //u1dxx
    kRangeSetLatin,          //u1exx
    kRangeGreek              //u1fxx
  },
  { //table for 2x--
    kRangeSetLatin,          //u20xx
    kRangeSetLatin,          //u21xx
    kRangeMathOperators,     //u22xx
    kRangeMiscTechnical,     //u23xx
    kRangeControlOpticalEnclose, //u24xx
    kRangeBoxBlockGeometrics, //u25xx
    kRangeMiscSymbols,       //u26xx
    kRangeDingbats,          //u27xx
    kRangeBraillePattern,    //u28xx
    kRangeUnassigned,        //u29xx
    kRangeUnassigned,        //u2axx
    kRangeUnassigned,        //u2bxx
    kRangeUnassigned,        //u2cxx
    kRangeUnassigned,        //u2dxx
    kRangeSetCJK,            //u2exx
    kRangeSetCJK             //u2fxx
  },
  {  //table for ax--
    kRangeYi,                //ua0xx
    kRangeYi,                //ua1xx
    kRangeYi,                //ua2xx
    kRangeYi,                //ua3xx
    kRangeYi,                //ua4xx
    kRangeUnassigned,        //ua5xx
    kRangeUnassigned,        //ua6xx
    kRangeUnassigned,        //ua7xx
    kRangeUnassigned,        //ua8xx
    kRangeUnassigned,        //ua9xx
    kRangeUnassigned,        //uaaxx
    kRangeUnassigned,        //uabxx
    kRangeKorean,            //uacxx
    kRangeKorean,            //uadxx
    kRangeKorean,            //uaexx
    kRangeKorean             //uafxx
  },
  {  //table for dx--
    kRangeKorean,            //ud0xx
    kRangeKorean,            //ud1xx
    kRangeKorean,            //ud2xx
    kRangeKorean,            //ud3xx
    kRangeKorean,            //ud4xx
    kRangeKorean,            //ud5xx
    kRangeKorean,            //ud6xx
    kRangeKorean,            //ud7xx
    kRangeSurrogate,         //ud8xx
    kRangeSurrogate,         //ud9xx
    kRangeSurrogate,         //udaxx
    kRangeSurrogate,         //udbxx
    kRangeSurrogate,         //udcxx
    kRangeSurrogate,         //uddxx
    kRangeSurrogate,         //udexx
    kRangeSurrogate          //udfxx
  },
  { // table for fx--
    kRangePrivate,           //uf0xx 
    kRangePrivate,           //uf1xx 
    kRangePrivate,           //uf2xx 
    kRangePrivate,           //uf3xx 
    kRangePrivate,           //uf4xx 
    kRangePrivate,           //uf5xx 
    kRangePrivate,           //uf6xx 
    kRangePrivate,           //uf7xx 
    kRangePrivate,           //uf8xx 
    kRangeSetCJK,            //uf9xx 
    kRangeSetCJK,            //ufaxx 
    kRangeArabic,            //ufbxx, includes alphabic presentation form
    kRangeArabic,            //ufcxx
    kRangeArabic,            //ufdxx
    kRangeTableBase+8,       //ufexx
    kRangeTableBase+9        //uffxx, halfwidth and fullwidth forms, includes Specials
  },
  { //table for 0x0500 - 0x05ff
    kRangeCyrillic,          //u050x
    kRangeCyrillic,          //u051x
    kRangeCyrillic,          //u052x
    kRangeArmenian,          //u053x
    kRangeArmenian,          //u054x
    kRangeArmenian,          //u055x
    kRangeArmenian,          //u056x
    kRangeArmenian,          //u057x
    kRangeArmenian,          //u058x
    kRangeHebrew,            //u059x
    kRangeHebrew,            //u05ax
    kRangeHebrew,            //u05bx
    kRangeHebrew,            //u05cx
    kRangeHebrew,            //u05dx
    kRangeHebrew,            //u05ex
    kRangeHebrew             //u05fx
  },
  { //table for 0xfe00 - 0xfeff
    kRangeSetCJK,            //ufe0x
    kRangeSetCJK,            //ufe1x
    kRangeSetCJK,            //ufe2x
    kRangeSetCJK,            //ufe3x
    kRangeSetCJK,            //ufe4x
    kRangeSetCJK,            //ufe5x
    kRangeSetCJK,            //ufe6x
    kRangeArabic,            //ufe7x
    kRangeArabic,            //ufe8x
    kRangeArabic,            //ufe9x
    kRangeArabic,            //ufeax
    kRangeArabic,            //ufebx
    kRangeArabic,            //ufecx
    kRangeArabic,            //ufedx
    kRangeArabic,            //ufeex
    kRangeArabic             //ufefx
  },
  { //table for 0xff00 - 0xffff
    kRangeSetCJK,            //uff0x, fullwidth latin
    kRangeSetCJK,            //uff1x, fullwidth latin
    kRangeSetCJK,            //uff2x, fullwidth latin
    kRangeSetCJK,            //uff3x, fullwidth latin
    kRangeSetCJK,            //uff4x, fullwidth latin
    kRangeSetCJK,            //uff5x, fullwidth latin
    kRangeSetCJK,            //uff6x, halfwidth katakana
    kRangeSetCJK,            //uff7x, halfwidth katakana
    kRangeSetCJK,            //uff8x, halfwidth katakana
    kRangeSetCJK,            //uff9x, halfwidth katakana
    kRangeSetCJK,            //uffax, halfwidth hangul jamo
    kRangeSetCJK,            //uffbx, halfwidth hangul jamo
    kRangeSetCJK,            //uffcx, halfwidth hangul jamo
    kRangeSetCJK,            //uffdx, halfwidth hangul jamo
    kRangeSetCJK,            //uffex, fullwidth symbols
    kRangeSpecials,          //ufffx, Specials
  },
};

// Most scripts between U+0700 and U+16FF are assigned a chunk of 128 (0x80) 
// code points  so that the number of entries in the tertiary range
// table for that range is obtained by dividing (0x1700 - 0x0700) by 128.
// Exceptions: Ethiopic, Tibetan, Hangul Jamo and Canadian aboriginal 
// syllabaries take multiple chunks and Ogham and Runic share  a single chunk.
#define TERTIARY_TABLE_SIZE ((0x1700 - 0x0700) / 0x80)

static const uint8_t gUnicodeTertiaryRangeTable[TERTIARY_TABLE_SIZE] =
{ //table for 0x0700 - 0x1600 
    kRangeSyriac,            //u070x
    kRangeThaana,            //u078x
    kRangeUnassigned,        //u080x  place holder(resolved in the 2ndary tab.)
    kRangeUnassigned,        //u088x  place holder(resolved in the 2ndary tab.)
    kRangeDevanagari,        //u090x
    kRangeBengali,           //u098x
    kRangeGurmukhi,          //u0a0x
    kRangeGujarati,          //u0a8x
    kRangeOriya,             //u0b0x
    kRangeTamil,             //u0b8x
    kRangeTelugu,            //u0c0x
    kRangeKannada,           //u0c8x
    kRangeMalayalam,         //u0d0x
    kRangeSinhala,           //u0d8x
    kRangeThai,              //u0e0x  
    kRangeLao,               //u0e8x
    kRangeTibetan,           //u0f0x  place holder(resolved in the 2ndary tab.)
    kRangeTibetan,           //u0f8x  place holder(resolved in the 2ndary tab.)
    kRangeMyanmar,           //u100x
    kRangeGeorgian,          //u108x
    kRangeKorean,            //u110x  place holder(resolved in the 2ndary tab.)
    kRangeKorean,            //u118x  place holder(resolved in the 2ndary tab.)
    kRangeEthiopic,          //u120x  place holder(resolved in the 2ndary tab.)
    kRangeEthiopic,          //u128x  place holder(resolved in the 2ndary tab.)
    kRangeEthiopic,          //u130x  
    kRangeCherokee,          //u138x
    kRangeCanadian,          //u140x  place holder(resolved in the 2ndary tab.)
    kRangeCanadian,          //u148x  place holder(resolved in the 2ndary tab.)
    kRangeCanadian,          //u150x  place holder(resolved in the 2ndary tab.)
    kRangeCanadian,          //u158x  place holder(resolved in the 2ndary tab.)
    kRangeCanadian,          //u160x  
    kRangeOghamRunic         //u168x  this contains two scripts, Ogham & Runic
};

// A two level index is almost enough for locating a range, with the 
// exception of u03xx and u05xx. Since we don't really care about range for
// combining diacritical marks in our font application, they are 
// not discriminated further. But future adoption of this module for other use 
// should be aware of this limitation. The implementation can be extended if 
// there is such a need.
// For Indic, Southeast Asian scripts and some other scripts between
// U+0700 and U+16FF, it's extended to the third level.
uint32_t FindCharUnicodeRange(uint32_t ch)
{
  uint32_t range;
  
  // aggregate ranges for non-BMP codepoints
  if (ch > 0xFFFF) {
    uint32_t p = (ch >> 16);
    if (p == 1) {
        return kRangeSMP;
    } else if (p == 2) {
        return kRangeSetCJK;
    }
    return kRangeHigherPlanes;
  }

  // lookup explicit range for BMP codepoints
  // first general range
  range = gUnicodeSubrangeTable[0][ch >> 12];
  
  // if general range is good enough, return that
  if (range < kRangeTableBase)
    // we try to get a specific range 
    return range;

  // otherwise, use subrange tables
  range = gUnicodeSubrangeTable[range - kRangeTableBase][(ch & 0x0f00) >> 8];
  if (range < kRangeTableBase)
    return range;
  if (range < kRangeTertiaryTable)
    return gUnicodeSubrangeTable[range - kRangeTableBase][(ch & 0x00f0) >> 4];

  // Yet another table to look at : U+0700 - U+16FF : 128 code point blocks
  return gUnicodeTertiaryRangeTable[(ch - 0x0700) >> 7];
}

nsIAtom *LangGroupFromUnicodeRange(uint8_t unicodeRange)
{
  if (kRangeSpecificItemNum > unicodeRange) {
    nsIAtom **atom = gUnicodeRangeToLangGroupAtomTable[unicodeRange];
    return *atom;
  }
  return nullptr;
}