]>
Commit | Line | Data |
---|---|---|
b75a7d8f A |
1 | /* |
2 | ******************************************************************************* | |
3 | * | |
2ca993e8 | 4 | * Copyright (C) 1999-2016, International Business Machines |
b75a7d8f A |
5 | * Corporation and others. All Rights Reserved. |
6 | * | |
7 | ******************************************************************************* | |
8 | * file name: scrptrun.cpp | |
9 | * | |
10 | * created on: 10/17/2001 | |
11 | * created by: Eric R. Mader | |
12 | */ | |
13 | ||
14 | #include "unicode/utypes.h" | |
15 | #include "unicode/uscript.h" | |
16 | ||
2ca993e8 | 17 | #include "cmemory.h" |
b75a7d8f A |
18 | #include "scrptrun.h" |
19 | ||
b75a7d8f A |
20 | const char ScriptRun::fgClassID=0; |
21 | ||
22 | UChar32 ScriptRun::pairedChars[] = { | |
23 | 0x0028, 0x0029, // ascii paired punctuation | |
24 | 0x003c, 0x003e, | |
25 | 0x005b, 0x005d, | |
26 | 0x007b, 0x007d, | |
27 | 0x00ab, 0x00bb, // guillemets | |
28 | 0x2018, 0x2019, // general punctuation | |
29 | 0x201c, 0x201d, | |
30 | 0x2039, 0x203a, | |
31 | 0x3008, 0x3009, // chinese paired punctuation | |
32 | 0x300a, 0x300b, | |
33 | 0x300c, 0x300d, | |
34 | 0x300e, 0x300f, | |
35 | 0x3010, 0x3011, | |
36 | 0x3014, 0x3015, | |
37 | 0x3016, 0x3017, | |
38 | 0x3018, 0x3019, | |
39 | 0x301a, 0x301b | |
40 | }; | |
41 | ||
2ca993e8 | 42 | const int32_t ScriptRun::pairedCharCount = UPRV_LENGTHOF(pairedChars); |
b75a7d8f A |
43 | const int32_t ScriptRun::pairedCharPower = 1 << highBit(pairedCharCount); |
44 | const int32_t ScriptRun::pairedCharExtra = pairedCharCount - pairedCharPower; | |
45 | ||
46 | int8_t ScriptRun::highBit(int32_t value) | |
47 | { | |
48 | if (value <= 0) { | |
49 | return -32; | |
50 | } | |
51 | ||
52 | int8_t bit = 0; | |
53 | ||
54 | if (value >= 1 << 16) { | |
55 | value >>= 16; | |
56 | bit += 16; | |
57 | } | |
58 | ||
59 | if (value >= 1 << 8) { | |
60 | value >>= 8; | |
61 | bit += 8; | |
62 | } | |
63 | ||
64 | if (value >= 1 << 4) { | |
65 | value >>= 4; | |
66 | bit += 4; | |
67 | } | |
68 | ||
69 | if (value >= 1 << 2) { | |
70 | value >>= 2; | |
71 | bit += 2; | |
72 | } | |
73 | ||
74 | if (value >= 1 << 1) { | |
75 | value >>= 1; | |
76 | bit += 1; | |
77 | } | |
78 | ||
79 | return bit; | |
80 | } | |
81 | ||
82 | int32_t ScriptRun::getPairIndex(UChar32 ch) | |
83 | { | |
84 | int32_t probe = pairedCharPower; | |
85 | int32_t index = 0; | |
86 | ||
87 | if (ch >= pairedChars[pairedCharExtra]) { | |
88 | index = pairedCharExtra; | |
89 | } | |
90 | ||
91 | while (probe > (1 << 0)) { | |
92 | probe >>= 1; | |
93 | ||
94 | if (ch >= pairedChars[index + probe]) { | |
95 | index += probe; | |
96 | } | |
97 | } | |
98 | ||
99 | if (pairedChars[index] != ch) { | |
100 | index = -1; | |
101 | } | |
102 | ||
103 | return index; | |
104 | } | |
105 | ||
106 | UBool ScriptRun::sameScript(int32_t scriptOne, int32_t scriptTwo) | |
107 | { | |
108 | return scriptOne <= USCRIPT_INHERITED || scriptTwo <= USCRIPT_INHERITED || scriptOne == scriptTwo; | |
109 | } | |
110 | ||
111 | UBool ScriptRun::next() | |
112 | { | |
113 | int32_t startSP = parenSP; // used to find the first new open character | |
114 | UErrorCode error = U_ZERO_ERROR; | |
115 | ||
116 | // if we've fallen off the end of the text, we're done | |
117 | if (scriptEnd >= charLimit) { | |
118 | return false; | |
119 | } | |
120 | ||
121 | scriptCode = USCRIPT_COMMON; | |
122 | ||
123 | for (scriptStart = scriptEnd; scriptEnd < charLimit; scriptEnd += 1) { | |
124 | UChar high = charArray[scriptEnd]; | |
125 | UChar32 ch = high; | |
126 | ||
127 | // if the character is a high surrogate and it's not the last one | |
128 | // in the text, see if it's followed by a low surrogate | |
129 | if (high >= 0xD800 && high <= 0xDBFF && scriptEnd < charLimit - 1) | |
130 | { | |
131 | UChar low = charArray[scriptEnd + 1]; | |
132 | ||
133 | // if it is followed by a low surrogate, | |
134 | // consume it and form the full character | |
135 | if (low >= 0xDC00 && low <= 0xDFFF) { | |
136 | ch = (high - 0xD800) * 0x0400 + low - 0xDC00 + 0x10000; | |
137 | scriptEnd += 1; | |
138 | } | |
139 | } | |
140 | ||
141 | UScriptCode sc = uscript_getScript(ch, &error); | |
142 | int32_t pairIndex = getPairIndex(ch); | |
143 | ||
144 | // Paired character handling: | |
145 | // | |
146 | // if it's an open character, push it onto the stack. | |
147 | // if it's a close character, find the matching open on the | |
148 | // stack, and use that script code. Any non-matching open | |
149 | // characters above it on the stack will be poped. | |
150 | if (pairIndex >= 0) { | |
151 | if ((pairIndex & 1) == 0) { | |
152 | parenStack[++parenSP].pairIndex = pairIndex; | |
153 | parenStack[parenSP].scriptCode = scriptCode; | |
154 | } else if (parenSP >= 0) { | |
155 | int32_t pi = pairIndex & ~1; | |
156 | ||
157 | while (parenSP >= 0 && parenStack[parenSP].pairIndex != pi) { | |
158 | parenSP -= 1; | |
159 | } | |
160 | ||
161 | if (parenSP < startSP) { | |
162 | startSP = parenSP; | |
163 | } | |
164 | ||
165 | if (parenSP >= 0) { | |
166 | sc = parenStack[parenSP].scriptCode; | |
167 | } | |
168 | } | |
169 | } | |
170 | ||
171 | if (sameScript(scriptCode, sc)) { | |
172 | if (scriptCode <= USCRIPT_INHERITED && sc > USCRIPT_INHERITED) { | |
173 | scriptCode = sc; | |
174 | ||
175 | // now that we have a final script code, fix any open | |
176 | // characters we pushed before we knew the script code. | |
177 | while (startSP < parenSP) { | |
178 | parenStack[++startSP].scriptCode = scriptCode; | |
179 | } | |
180 | } | |
181 | ||
182 | // if this character is a close paired character, | |
183 | // pop it from the stack | |
184 | if (pairIndex >= 0 && (pairIndex & 1) != 0 && parenSP >= 0) { | |
185 | parenSP -= 1; | |
186 | startSP -= 1; | |
187 | } | |
188 | } else { | |
189 | // if the run broke on a surrogate pair, | |
190 | // end it before the high surrogate | |
191 | if (ch >= 0x10000) { | |
192 | scriptEnd -= 1; | |
193 | } | |
194 | ||
195 | break; | |
196 | } | |
197 | } | |
198 | ||
199 | return true; | |
200 | } | |
201 |