1.1 --- /dev/null Thu Jan 01 00:00:00 1970 +0000
1.2 +++ b/intl/icu/source/extra/scrptrun/scrptrun.cpp Wed Dec 31 06:09:35 2014 +0100
1.3 @@ -0,0 +1,202 @@
1.4 +/*
1.5 + *******************************************************************************
1.6 + *
1.7 + * Copyright (C) 1999-2001, International Business Machines
1.8 + * Corporation and others. All Rights Reserved.
1.9 + *
1.10 + *******************************************************************************
1.11 + * file name: scrptrun.cpp
1.12 + *
1.13 + * created on: 10/17/2001
1.14 + * created by: Eric R. Mader
1.15 + */
1.16 +
1.17 +#include "unicode/utypes.h"
1.18 +#include "unicode/uscript.h"
1.19 +
1.20 +#include "scrptrun.h"
1.21 +
1.22 +#define ARRAY_SIZE(array) (sizeof array / sizeof array[0])
1.23 +
1.24 +const char ScriptRun::fgClassID=0;
1.25 +
1.26 +UChar32 ScriptRun::pairedChars[] = {
1.27 + 0x0028, 0x0029, // ascii paired punctuation
1.28 + 0x003c, 0x003e,
1.29 + 0x005b, 0x005d,
1.30 + 0x007b, 0x007d,
1.31 + 0x00ab, 0x00bb, // guillemets
1.32 + 0x2018, 0x2019, // general punctuation
1.33 + 0x201c, 0x201d,
1.34 + 0x2039, 0x203a,
1.35 + 0x3008, 0x3009, // chinese paired punctuation
1.36 + 0x300a, 0x300b,
1.37 + 0x300c, 0x300d,
1.38 + 0x300e, 0x300f,
1.39 + 0x3010, 0x3011,
1.40 + 0x3014, 0x3015,
1.41 + 0x3016, 0x3017,
1.42 + 0x3018, 0x3019,
1.43 + 0x301a, 0x301b
1.44 +};
1.45 +
1.46 +const int32_t ScriptRun::pairedCharCount = ARRAY_SIZE(pairedChars);
1.47 +const int32_t ScriptRun::pairedCharPower = 1 << highBit(pairedCharCount);
1.48 +const int32_t ScriptRun::pairedCharExtra = pairedCharCount - pairedCharPower;
1.49 +
1.50 +int8_t ScriptRun::highBit(int32_t value)
1.51 +{
1.52 + if (value <= 0) {
1.53 + return -32;
1.54 + }
1.55 +
1.56 + int8_t bit = 0;
1.57 +
1.58 + if (value >= 1 << 16) {
1.59 + value >>= 16;
1.60 + bit += 16;
1.61 + }
1.62 +
1.63 + if (value >= 1 << 8) {
1.64 + value >>= 8;
1.65 + bit += 8;
1.66 + }
1.67 +
1.68 + if (value >= 1 << 4) {
1.69 + value >>= 4;
1.70 + bit += 4;
1.71 + }
1.72 +
1.73 + if (value >= 1 << 2) {
1.74 + value >>= 2;
1.75 + bit += 2;
1.76 + }
1.77 +
1.78 + if (value >= 1 << 1) {
1.79 + value >>= 1;
1.80 + bit += 1;
1.81 + }
1.82 +
1.83 + return bit;
1.84 +}
1.85 +
1.86 +int32_t ScriptRun::getPairIndex(UChar32 ch)
1.87 +{
1.88 + int32_t probe = pairedCharPower;
1.89 + int32_t index = 0;
1.90 +
1.91 + if (ch >= pairedChars[pairedCharExtra]) {
1.92 + index = pairedCharExtra;
1.93 + }
1.94 +
1.95 + while (probe > (1 << 0)) {
1.96 + probe >>= 1;
1.97 +
1.98 + if (ch >= pairedChars[index + probe]) {
1.99 + index += probe;
1.100 + }
1.101 + }
1.102 +
1.103 + if (pairedChars[index] != ch) {
1.104 + index = -1;
1.105 + }
1.106 +
1.107 + return index;
1.108 +}
1.109 +
1.110 +UBool ScriptRun::sameScript(int32_t scriptOne, int32_t scriptTwo)
1.111 +{
1.112 + return scriptOne <= USCRIPT_INHERITED || scriptTwo <= USCRIPT_INHERITED || scriptOne == scriptTwo;
1.113 +}
1.114 +
1.115 +UBool ScriptRun::next()
1.116 +{
1.117 + int32_t startSP = parenSP; // used to find the first new open character
1.118 + UErrorCode error = U_ZERO_ERROR;
1.119 +
1.120 + // if we've fallen off the end of the text, we're done
1.121 + if (scriptEnd >= charLimit) {
1.122 + return false;
1.123 + }
1.124 +
1.125 + scriptCode = USCRIPT_COMMON;
1.126 +
1.127 + for (scriptStart = scriptEnd; scriptEnd < charLimit; scriptEnd += 1) {
1.128 + UChar high = charArray[scriptEnd];
1.129 + UChar32 ch = high;
1.130 +
1.131 + // if the character is a high surrogate and it's not the last one
1.132 + // in the text, see if it's followed by a low surrogate
1.133 + if (high >= 0xD800 && high <= 0xDBFF && scriptEnd < charLimit - 1)
1.134 + {
1.135 + UChar low = charArray[scriptEnd + 1];
1.136 +
1.137 + // if it is followed by a low surrogate,
1.138 + // consume it and form the full character
1.139 + if (low >= 0xDC00 && low <= 0xDFFF) {
1.140 + ch = (high - 0xD800) * 0x0400 + low - 0xDC00 + 0x10000;
1.141 + scriptEnd += 1;
1.142 + }
1.143 + }
1.144 +
1.145 + UScriptCode sc = uscript_getScript(ch, &error);
1.146 + int32_t pairIndex = getPairIndex(ch);
1.147 +
1.148 + // Paired character handling:
1.149 + //
1.150 + // if it's an open character, push it onto the stack.
1.151 + // if it's a close character, find the matching open on the
1.152 + // stack, and use that script code. Any non-matching open
1.153 + // characters above it on the stack will be poped.
1.154 + if (pairIndex >= 0) {
1.155 + if ((pairIndex & 1) == 0) {
1.156 + parenStack[++parenSP].pairIndex = pairIndex;
1.157 + parenStack[parenSP].scriptCode = scriptCode;
1.158 + } else if (parenSP >= 0) {
1.159 + int32_t pi = pairIndex & ~1;
1.160 +
1.161 + while (parenSP >= 0 && parenStack[parenSP].pairIndex != pi) {
1.162 + parenSP -= 1;
1.163 + }
1.164 +
1.165 + if (parenSP < startSP) {
1.166 + startSP = parenSP;
1.167 + }
1.168 +
1.169 + if (parenSP >= 0) {
1.170 + sc = parenStack[parenSP].scriptCode;
1.171 + }
1.172 + }
1.173 + }
1.174 +
1.175 + if (sameScript(scriptCode, sc)) {
1.176 + if (scriptCode <= USCRIPT_INHERITED && sc > USCRIPT_INHERITED) {
1.177 + scriptCode = sc;
1.178 +
1.179 + // now that we have a final script code, fix any open
1.180 + // characters we pushed before we knew the script code.
1.181 + while (startSP < parenSP) {
1.182 + parenStack[++startSP].scriptCode = scriptCode;
1.183 + }
1.184 + }
1.185 +
1.186 + // if this character is a close paired character,
1.187 + // pop it from the stack
1.188 + if (pairIndex >= 0 && (pairIndex & 1) != 0 && parenSP >= 0) {
1.189 + parenSP -= 1;
1.190 + startSP -= 1;
1.191 + }
1.192 + } else {
1.193 + // if the run broke on a surrogate pair,
1.194 + // end it before the high surrogate
1.195 + if (ch >= 0x10000) {
1.196 + scriptEnd -= 1;
1.197 + }
1.198 +
1.199 + break;
1.200 + }
1.201 + }
1.202 +
1.203 + return true;
1.204 +}
1.205 +
