1 /*
2 * Copyright 2013 Google Inc.
3 *
4 * Use of this source code is governed by a BSD-style license that can be
5 * found in the LICENSE file.
6 */
7
8 #include "src/sfnt/SkOTTable_name.h"
9
10 #include "src/base/SkEndian.h"
11 #include "src/base/SkTSearch.h"
12 #include "src/base/SkUTF.h"
13 #include "src/core/SkStringUtils.h"
14
next_unichar_UTF16BE(const uint8_t ** srcPtr,size_t * length)15 static SkUnichar next_unichar_UTF16BE(const uint8_t** srcPtr, size_t* length) {
16 SkASSERT(srcPtr && *srcPtr && length);
17 SkASSERT(*length > 0);
18
19 uint16_t leading;
20 if (*length < sizeof(leading)) {
21 *length = 0;
22 return 0xFFFD;
23 }
24 memcpy(&leading, *srcPtr, sizeof(leading));
25 *srcPtr += sizeof(leading);
26 *length -= sizeof(leading);
27 SkUnichar c = SkEndian_SwapBE16(leading);
28
29 if (SkUTF::IsTrailingSurrogateUTF16(c)) {
30 return 0xFFFD;
31 }
32 if (SkUTF::IsLeadingSurrogateUTF16(c)) {
33 uint16_t trailing;
34 if (*length < sizeof(trailing)) {
35 *length = 0;
36 return 0xFFFD;
37 }
38 memcpy(&trailing, *srcPtr, sizeof(trailing));
39 SkUnichar c2 = SkEndian_SwapBE16(trailing);
40 if (!SkUTF::IsTrailingSurrogateUTF16(c2)) {
41 return 0xFFFD;
42 }
43 *srcPtr += sizeof(trailing);
44 *length -= sizeof(trailing);
45
46 c = (c << 10) + c2 + (0x10000 - (0xD800 << 10) - 0xDC00);
47 }
48 return c;
49 }
50
SkString_from_UTF16BE(const uint8_t * utf16be,size_t length,SkString & utf8)51 static void SkString_from_UTF16BE(const uint8_t* utf16be, size_t length, SkString& utf8) {
52 // Note that utf16be may not be 2-byte aligned.
53 SkASSERT(utf16be != nullptr);
54
55 utf8.reset();
56 while (length) {
57 utf8.appendUnichar(next_unichar_UTF16BE(&utf16be, &length));
58 }
59 }
60
61 /** UnicodeFromMacRoman[macRomanPoint - 0x80] -> unicodeCodePoint.
62 * Derived from http://www.unicode.org/Public/MAPPINGS/VENDORS/APPLE/ROMAN.TXT .
63 * In MacRoman the first 128 code points match ASCII code points.
64 * This maps the second 128 MacRoman code points to unicode code points.
65 */
66 static const uint16_t UnicodeFromMacRoman[0x80] = {
67 0x00C4, 0x00C5, 0x00C7, 0x00C9, 0x00D1, 0x00D6, 0x00DC, 0x00E1,
68 0x00E0, 0x00E2, 0x00E4, 0x00E3, 0x00E5, 0x00E7, 0x00E9, 0x00E8,
69 0x00EA, 0x00EB, 0x00ED, 0x00EC, 0x00EE, 0x00EF, 0x00F1, 0x00F3,
70 0x00F2, 0x00F4, 0x00F6, 0x00F5, 0x00FA, 0x00F9, 0x00FB, 0x00FC,
71 0x2020, 0x00B0, 0x00A2, 0x00A3, 0x00A7, 0x2022, 0x00B6, 0x00DF,
72 0x00AE, 0x00A9, 0x2122, 0x00B4, 0x00A8, 0x2260, 0x00C6, 0x00D8,
73 0x221E, 0x00B1, 0x2264, 0x2265, 0x00A5, 0x00B5, 0x2202, 0x2211,
74 0x220F, 0x03C0, 0x222B, 0x00AA, 0x00BA, 0x03A9, 0x00E6, 0x00F8,
75 0x00BF, 0x00A1, 0x00AC, 0x221A, 0x0192, 0x2248, 0x2206, 0x00AB,
76 0x00BB, 0x2026, 0x00A0, 0x00C0, 0x00C3, 0x00D5, 0x0152, 0x0153,
77 0x2013, 0x2014, 0x201C, 0x201D, 0x2018, 0x2019, 0x00F7, 0x25CA,
78 0x00FF, 0x0178, 0x2044, 0x20AC, 0x2039, 0x203A, 0xFB01, 0xFB02,
79 0x2021, 0x00B7, 0x201A, 0x201E, 0x2030, 0x00C2, 0x00CA, 0x00C1,
80 0x00CB, 0x00C8, 0x00CD, 0x00CE, 0x00CF, 0x00CC, 0x00D3, 0x00D4,
81 0xF8FF, 0x00D2, 0x00DA, 0x00DB, 0x00D9, 0x0131, 0x02C6, 0x02DC,
82 0x00AF, 0x02D8, 0x02D9, 0x02DA, 0x00B8, 0x02DD, 0x02DB, 0x02C7,
83 };
84
SkStringFromMacRoman(const uint8_t * macRoman,size_t length,SkString & utf8)85 static void SkStringFromMacRoman(const uint8_t* macRoman, size_t length, SkString& utf8) {
86 utf8.reset();
87 for (size_t i = 0; i < length; ++i) {
88 utf8.appendUnichar(macRoman[i] < 0x80 ? macRoman[i]
89 : UnicodeFromMacRoman[macRoman[i] - 0x80]);
90 }
91 }
92
93 static const struct BCP47FromLanguageId {
94 uint16_t languageID;
95 const char* bcp47;
96 }
97 /** The Mac and Windows values do not conflict, so this is currently one single table. */
98 BCP47FromLanguageID[] = {
99 /** A mapping from Mac Language Designators to BCP 47 codes.
100 * The following list was constructed more or less manually.
101 * Apple now uses BCP 47 (post OSX10.4), so there will be no new entries.
102 */
103 {0, "en"}, //English
104 {1, "fr"}, //French
105 {2, "de"}, //German
106 {3, "it"}, //Italian
107 {4, "nl"}, //Dutch
108 {5, "sv"}, //Swedish
109 {6, "es"}, //Spanish
110 {7, "da"}, //Danish
111 {8, "pt"}, //Portuguese
112 {9, "nb"}, //Norwegian
113 {10, "he"}, //Hebrew
114 {11, "ja"}, //Japanese
115 {12, "ar"}, //Arabic
116 {13, "fi"}, //Finnish
117 {14, "el"}, //Greek
118 {15, "is"}, //Icelandic
119 {16, "mt"}, //Maltese
120 {17, "tr"}, //Turkish
121 {18, "hr"}, //Croatian
122 {19, "zh-Hant"}, //Chinese (Traditional)
123 {20, "ur"}, //Urdu
124 {21, "hi"}, //Hindi
125 {22, "th"}, //Thai
126 {23, "ko"}, //Korean
127 {24, "lt"}, //Lithuanian
128 {25, "pl"}, //Polish
129 {26, "hu"}, //Hungarian
130 {27, "et"}, //Estonian
131 {28, "lv"}, //Latvian
132 {29, "se"}, //Sami
133 {30, "fo"}, //Faroese
134 {31, "fa"}, //Farsi (Persian)
135 {32, "ru"}, //Russian
136 {33, "zh-Hans"}, //Chinese (Simplified)
137 {34, "nl"}, //Dutch
138 {35, "ga"}, //Irish(Gaelic)
139 {36, "sq"}, //Albanian
140 {37, "ro"}, //Romanian
141 {38, "cs"}, //Czech
142 {39, "sk"}, //Slovak
143 {40, "sl"}, //Slovenian
144 {41, "yi"}, //Yiddish
145 {42, "sr"}, //Serbian
146 {43, "mk"}, //Macedonian
147 {44, "bg"}, //Bulgarian
148 {45, "uk"}, //Ukrainian
149 {46, "be"}, //Byelorussian
150 {47, "uz"}, //Uzbek
151 {48, "kk"}, //Kazakh
152 {49, "az-Cyrl"}, //Azerbaijani (Cyrillic)
153 {50, "az-Arab"}, //Azerbaijani (Arabic)
154 {51, "hy"}, //Armenian
155 {52, "ka"}, //Georgian
156 {53, "mo"}, //Moldavian
157 {54, "ky"}, //Kirghiz
158 {55, "tg"}, //Tajiki
159 {56, "tk"}, //Turkmen
160 {57, "mn-Mong"}, //Mongolian (Traditional)
161 {58, "mn-Cyrl"}, //Mongolian (Cyrillic)
162 {59, "ps"}, //Pashto
163 {60, "ku"}, //Kurdish
164 {61, "ks"}, //Kashmiri
165 {62, "sd"}, //Sindhi
166 {63, "bo"}, //Tibetan
167 {64, "ne"}, //Nepali
168 {65, "sa"}, //Sanskrit
169 {66, "mr"}, //Marathi
170 {67, "bn"}, //Bengali
171 {68, "as"}, //Assamese
172 {69, "gu"}, //Gujarati
173 {70, "pa"}, //Punjabi
174 {71, "or"}, //Oriya
175 {72, "ml"}, //Malayalam
176 {73, "kn"}, //Kannada
177 {74, "ta"}, //Tamil
178 {75, "te"}, //Telugu
179 {76, "si"}, //Sinhalese
180 {77, "my"}, //Burmese
181 {78, "km"}, //Khmer
182 {79, "lo"}, //Lao
183 {80, "vi"}, //Vietnamese
184 {81, "id"}, //Indonesian
185 {82, "tl"}, //Tagalog
186 {83, "ms-Latn"}, //Malay (Roman)
187 {84, "ms-Arab"}, //Malay (Arabic)
188 {85, "am"}, //Amharic
189 {86, "ti"}, //Tigrinya
190 {87, "om"}, //Oromo
191 {88, "so"}, //Somali
192 {89, "sw"}, //Swahili
193 {90, "rw"}, //Kinyarwanda/Ruanda
194 {91, "rn"}, //Rundi
195 {92, "ny"}, //Nyanja/Chewa
196 {93, "mg"}, //Malagasy
197 {94, "eo"}, //Esperanto
198 {128, "cy"}, //Welsh
199 {129, "eu"}, //Basque
200 {130, "ca"}, //Catalan
201 {131, "la"}, //Latin
202 {132, "qu"}, //Quechua
203 {133, "gn"}, //Guarani
204 {134, "ay"}, //Aymara
205 {135, "tt"}, //Tatar
206 {136, "ug"}, //Uighur
207 {137, "dz"}, //Dzongkha
208 {138, "jv-Latn"}, //Javanese (Roman)
209 {139, "su-Latn"}, //Sundanese (Roman)
210 {140, "gl"}, //Galician
211 {141, "af"}, //Afrikaans
212 {142, "br"}, //Breton
213 {143, "iu"}, //Inuktitut
214 {144, "gd"}, //Scottish (Gaelic)
215 {145, "gv"}, //Manx (Gaelic)
216 {146, "ga"}, //Irish (Gaelic with Lenition)
217 {147, "to"}, //Tongan
218 {148, "el"}, //Greek (Polytonic) Note: ISO 15924 does not have an equivalent script name.
219 {149, "kl"}, //Greenlandic
220 {150, "az-Latn"}, //Azerbaijani (Roman)
221 {151, "nn"}, //Nynorsk
222
223 /** A mapping from Windows LCID to BCP 47 codes.
224 * This list is the sorted, curated output of tools/win_lcid.cpp.
225 * Note that these are sorted by value for quick binary lookup, and not logically by lsb.
226 * The 'bare' language ids (e.g. 0x0001 for Arabic) are ommitted
227 * as they do not appear as valid language ids in the OpenType specification.
228 */
229 { 0x0401, "ar-SA" }, //Arabic
230 { 0x0402, "bg-BG" }, //Bulgarian
231 { 0x0403, "ca-ES" }, //Catalan
232 { 0x0404, "zh-TW" }, //Chinese (Traditional)
233 { 0x0405, "cs-CZ" }, //Czech
234 { 0x0406, "da-DK" }, //Danish
235 { 0x0407, "de-DE" }, //German
236 { 0x0408, "el-GR" }, //Greek
237 { 0x0409, "en-US" }, //English
238 { 0x040a, "es-ES_tradnl" }, //Spanish
239 { 0x040b, "fi-FI" }, //Finnish
240 { 0x040c, "fr-FR" }, //French
241 { 0x040d, "he-IL" }, //Hebrew
242 { 0x040d, "he" }, //Hebrew
243 { 0x040e, "hu-HU" }, //Hungarian
244 { 0x040e, "hu" }, //Hungarian
245 { 0x040f, "is-IS" }, //Icelandic
246 { 0x0410, "it-IT" }, //Italian
247 { 0x0411, "ja-JP" }, //Japanese
248 { 0x0412, "ko-KR" }, //Korean
249 { 0x0413, "nl-NL" }, //Dutch
250 { 0x0414, "nb-NO" }, //Norwegian (Bokmål)
251 { 0x0415, "pl-PL" }, //Polish
252 { 0x0416, "pt-BR" }, //Portuguese
253 { 0x0417, "rm-CH" }, //Romansh
254 { 0x0418, "ro-RO" }, //Romanian
255 { 0x0419, "ru-RU" }, //Russian
256 { 0x041a, "hr-HR" }, //Croatian
257 { 0x041b, "sk-SK" }, //Slovak
258 { 0x041c, "sq-AL" }, //Albanian
259 { 0x041d, "sv-SE" }, //Swedish
260 { 0x041e, "th-TH" }, //Thai
261 { 0x041f, "tr-TR" }, //Turkish
262 { 0x0420, "ur-PK" }, //Urdu
263 { 0x0421, "id-ID" }, //Indonesian
264 { 0x0422, "uk-UA" }, //Ukrainian
265 { 0x0423, "be-BY" }, //Belarusian
266 { 0x0424, "sl-SI" }, //Slovenian
267 { 0x0425, "et-EE" }, //Estonian
268 { 0x0426, "lv-LV" }, //Latvian
269 { 0x0427, "lt-LT" }, //Lithuanian
270 { 0x0428, "tg-Cyrl-TJ" }, //Tajik (Cyrillic)
271 { 0x0429, "fa-IR" }, //Persian
272 { 0x042a, "vi-VN" }, //Vietnamese
273 { 0x042b, "hy-AM" }, //Armenian
274 { 0x042c, "az-Latn-AZ" }, //Azeri (Latin)
275 { 0x042d, "eu-ES" }, //Basque
276 { 0x042e, "hsb-DE" }, //Upper Sorbian
277 { 0x042f, "mk-MK" }, //Macedonian (FYROM)
278 { 0x0432, "tn-ZA" }, //Setswana
279 { 0x0434, "xh-ZA" }, //isiXhosa
280 { 0x0435, "zu-ZA" }, //isiZulu
281 { 0x0436, "af-ZA" }, //Afrikaans
282 { 0x0437, "ka-GE" }, //Georgian
283 { 0x0438, "fo-FO" }, //Faroese
284 { 0x0439, "hi-IN" }, //Hindi
285 { 0x043a, "mt-MT" }, //Maltese
286 { 0x043b, "se-NO" }, //Sami (Northern)
287 { 0x043e, "ms-MY" }, //Malay
288 { 0x043f, "kk-KZ" }, //Kazakh
289 { 0x0440, "ky-KG" }, //Kyrgyz
290 { 0x0441, "sw-KE" }, //Kiswahili
291 { 0x0442, "tk-TM" }, //Turkmen
292 { 0x0443, "uz-Latn-UZ" }, //Uzbek (Latin)
293 { 0x0443, "uz" }, //Uzbek
294 { 0x0444, "tt-RU" }, //Tatar
295 { 0x0445, "bn-IN" }, //Bengali
296 { 0x0446, "pa-IN" }, //Punjabi
297 { 0x0447, "gu-IN" }, //Gujarati
298 { 0x0448, "or-IN" }, //Oriya
299 { 0x0449, "ta-IN" }, //Tamil
300 { 0x044a, "te-IN" }, //Telugu
301 { 0x044b, "kn-IN" }, //Kannada
302 { 0x044c, "ml-IN" }, //Malayalam
303 { 0x044d, "as-IN" }, //Assamese
304 { 0x044e, "mr-IN" }, //Marathi
305 { 0x044f, "sa-IN" }, //Sanskrit
306 { 0x0450, "mn-Cyrl" }, //Mongolian (Cyrillic)
307 { 0x0451, "bo-CN" }, //Tibetan
308 { 0x0452, "cy-GB" }, //Welsh
309 { 0x0453, "km-KH" }, //Khmer
310 { 0x0454, "lo-LA" }, //Lao
311 { 0x0456, "gl-ES" }, //Galician
312 { 0x0457, "kok-IN" }, //Konkani
313 { 0x045a, "syr-SY" }, //Syriac
314 { 0x045b, "si-LK" }, //Sinhala
315 { 0x045d, "iu-Cans-CA" }, //Inuktitut (Syllabics)
316 { 0x045e, "am-ET" }, //Amharic
317 { 0x0461, "ne-NP" }, //Nepali
318 { 0x0462, "fy-NL" }, //Frisian
319 { 0x0463, "ps-AF" }, //Pashto
320 { 0x0464, "fil-PH" }, //Filipino
321 { 0x0465, "dv-MV" }, //Divehi
322 { 0x0468, "ha-Latn-NG" }, //Hausa (Latin)
323 { 0x046a, "yo-NG" }, //Yoruba
324 { 0x046b, "quz-BO" }, //Quechua
325 { 0x046c, "nso-ZA" }, //Sesotho sa Leboa
326 { 0x046d, "ba-RU" }, //Bashkir
327 { 0x046e, "lb-LU" }, //Luxembourgish
328 { 0x046f, "kl-GL" }, //Greenlandic
329 { 0x0470, "ig-NG" }, //Igbo
330 { 0x0478, "ii-CN" }, //Yi
331 { 0x047a, "arn-CL" }, //Mapudungun
332 { 0x047c, "moh-CA" }, //Mohawk
333 { 0x047e, "br-FR" }, //Breton
334 { 0x0480, "ug-CN" }, //Uyghur
335 { 0x0481, "mi-NZ" }, //Maori
336 { 0x0482, "oc-FR" }, //Occitan
337 { 0x0483, "co-FR" }, //Corsican
338 { 0x0484, "gsw-FR" }, //Alsatian
339 { 0x0485, "sah-RU" }, //Yakut
340 { 0x0486, "qut-GT" }, //K'iche
341 { 0x0487, "rw-RW" }, //Kinyarwanda
342 { 0x0488, "wo-SN" }, //Wolof
343 { 0x048c, "prs-AF" }, //Dari
344 { 0x0491, "gd-GB" }, //Scottish Gaelic
345 { 0x0801, "ar-IQ" }, //Arabic
346 { 0x0804, "zh-Hans" }, //Chinese (Simplified)
347 { 0x0807, "de-CH" }, //German
348 { 0x0809, "en-GB" }, //English
349 { 0x080a, "es-MX" }, //Spanish
350 { 0x080c, "fr-BE" }, //French
351 { 0x0810, "it-CH" }, //Italian
352 { 0x0813, "nl-BE" }, //Dutch
353 { 0x0814, "nn-NO" }, //Norwegian (Nynorsk)
354 { 0x0816, "pt-PT" }, //Portuguese
355 { 0x081a, "sr-Latn-CS" }, //Serbian (Latin)
356 { 0x081d, "sv-FI" }, //Swedish
357 { 0x082c, "az-Cyrl-AZ" }, //Azeri (Cyrillic)
358 { 0x082e, "dsb-DE" }, //Lower Sorbian
359 { 0x082e, "dsb" }, //Lower Sorbian
360 { 0x083b, "se-SE" }, //Sami (Northern)
361 { 0x083c, "ga-IE" }, //Irish
362 { 0x083e, "ms-BN" }, //Malay
363 { 0x0843, "uz-Cyrl-UZ" }, //Uzbek (Cyrillic)
364 { 0x0845, "bn-BD" }, //Bengali
365 { 0x0850, "mn-Mong-CN" }, //Mongolian (Traditional Mongolian)
366 { 0x085d, "iu-Latn-CA" }, //Inuktitut (Latin)
367 { 0x085f, "tzm-Latn-DZ" }, //Tamazight (Latin)
368 { 0x086b, "quz-EC" }, //Quechua
369 { 0x0c01, "ar-EG" }, //Arabic
370 { 0x0c04, "zh-Hant" }, //Chinese (Traditional)
371 { 0x0c07, "de-AT" }, //German
372 { 0x0c09, "en-AU" }, //English
373 { 0x0c0a, "es-ES" }, //Spanish
374 { 0x0c0c, "fr-CA" }, //French
375 { 0x0c1a, "sr-Cyrl-CS" }, //Serbian (Cyrillic)
376 { 0x0c3b, "se-FI" }, //Sami (Northern)
377 { 0x0c6b, "quz-PE" }, //Quechua
378 { 0x1001, "ar-LY" }, //Arabic
379 { 0x1004, "zh-SG" }, //Chinese (Simplified)
380 { 0x1007, "de-LU" }, //German
381 { 0x1009, "en-CA" }, //English
382 { 0x100a, "es-GT" }, //Spanish
383 { 0x100c, "fr-CH" }, //French
384 { 0x101a, "hr-BA" }, //Croatian (Latin)
385 { 0x103b, "smj-NO" }, //Sami (Lule)
386 { 0x1401, "ar-DZ" }, //Arabic
387 { 0x1404, "zh-MO" }, //Chinese (Traditional)
388 { 0x1407, "de-LI" }, //German
389 { 0x1409, "en-NZ" }, //English
390 { 0x140a, "es-CR" }, //Spanish
391 { 0x140c, "fr-LU" }, //French
392 { 0x141a, "bs-Latn-BA" }, //Bosnian (Latin)
393 { 0x141a, "bs" }, //Bosnian
394 { 0x143b, "smj-SE" }, //Sami (Lule)
395 { 0x143b, "smj" }, //Sami (Lule)
396 { 0x1801, "ar-MA" }, //Arabic
397 { 0x1809, "en-IE" }, //English
398 { 0x180a, "es-PA" }, //Spanish
399 { 0x180c, "fr-MC" }, //French
400 { 0x181a, "sr-Latn-BA" }, //Serbian (Latin)
401 { 0x183b, "sma-NO" }, //Sami (Southern)
402 { 0x1c01, "ar-TN" }, //Arabic
403 { 0x1c09, "en-ZA" }, //English
404 { 0x1c0a, "es-DO" }, //Spanish
405 { 0x1c1a, "sr-Cyrl-BA" }, //Serbian (Cyrillic)
406 { 0x1c3b, "sma-SE" }, //Sami (Southern)
407 { 0x1c3b, "sma" }, //Sami (Southern)
408 { 0x2001, "ar-OM" }, //Arabic
409 { 0x2009, "en-JM" }, //English
410 { 0x200a, "es-VE" }, //Spanish
411 { 0x201a, "bs-Cyrl-BA" }, //Bosnian (Cyrillic)
412 { 0x201a, "bs-Cyrl" }, //Bosnian (Cyrillic)
413 { 0x203b, "sms-FI" }, //Sami (Skolt)
414 { 0x203b, "sms" }, //Sami (Skolt)
415 { 0x2401, "ar-YE" }, //Arabic
416 { 0x2409, "en-029" }, //English
417 { 0x240a, "es-CO" }, //Spanish
418 { 0x241a, "sr-Latn-RS" }, //Serbian (Latin)
419 { 0x243b, "smn-FI" }, //Sami (Inari)
420 { 0x2801, "ar-SY" }, //Arabic
421 { 0x2809, "en-BZ" }, //English
422 { 0x280a, "es-PE" }, //Spanish
423 { 0x281a, "sr-Cyrl-RS" }, //Serbian (Cyrillic)
424 { 0x2c01, "ar-JO" }, //Arabic
425 { 0x2c09, "en-TT" }, //English
426 { 0x2c0a, "es-AR" }, //Spanish
427 { 0x2c1a, "sr-Latn-ME" }, //Serbian (Latin)
428 { 0x3001, "ar-LB" }, //Arabic
429 { 0x3009, "en-ZW" }, //English
430 { 0x300a, "es-EC" }, //Spanish
431 { 0x301a, "sr-Cyrl-ME" }, //Serbian (Cyrillic)
432 { 0x3401, "ar-KW" }, //Arabic
433 { 0x3409, "en-PH" }, //English
434 { 0x340a, "es-CL" }, //Spanish
435 { 0x3801, "ar-AE" }, //Arabic
436 { 0x380a, "es-UY" }, //Spanish
437 { 0x3c01, "ar-BH" }, //Arabic
438 { 0x3c0a, "es-PY" }, //Spanish
439 { 0x4001, "ar-QA" }, //Arabic
440 { 0x4009, "en-IN" }, //English
441 { 0x400a, "es-BO" }, //Spanish
442 { 0x4409, "en-MY" }, //English
443 { 0x440a, "es-SV" }, //Spanish
444 { 0x4809, "en-SG" }, //English
445 { 0x480a, "es-HN" }, //Spanish
446 { 0x4c0a, "es-NI" }, //Spanish
447 { 0x500a, "es-PR" }, //Spanish
448 { 0x540a, "es-US" }, //Spanish
449 };
450
451 namespace {
BCP47FromLanguageIdLess(const BCP47FromLanguageId & a,const BCP47FromLanguageId & b)452 bool BCP47FromLanguageIdLess(const BCP47FromLanguageId& a, const BCP47FromLanguageId& b) {
453 return a.languageID < b.languageID;
454 }
455 } // namespace
456
next(SkOTTableName::Iterator::Record & record)457 bool SkOTTableName::Iterator::next(SkOTTableName::Iterator::Record& record) {
458 SkOTTableName nameTable;
459 if (fNameTableSize < sizeof(nameTable)) {
460 return false;
461 }
462 memcpy(&nameTable, fNameTable, sizeof(nameTable));
463
464 const uint8_t* nameRecords = fNameTable + sizeof(nameTable);
465 const size_t nameRecordsSize = fNameTableSize - sizeof(nameTable);
466
467 const size_t stringTableOffset = SkEndian_SwapBE16(nameTable.stringOffset);
468 if (fNameTableSize < stringTableOffset) {
469 return false;
470 }
471 const uint8_t* stringTable = fNameTable + stringTableOffset;
472 const size_t stringTableSize = fNameTableSize - stringTableOffset;
473
474 // Find the next record which matches the requested type.
475 SkOTTableName::Record nameRecord;
476 const size_t nameRecordsCount = SkEndian_SwapBE16(nameTable.count);
477 const size_t nameRecordsMax = std::min(nameRecordsCount, nameRecordsSize / sizeof(nameRecord));
478 do {
479 if (fIndex >= nameRecordsMax) {
480 return false;
481 }
482
483 memcpy(&nameRecord, nameRecords + sizeof(nameRecord)*fIndex, sizeof(nameRecord));
484 ++fIndex;
485 } while (fType != -1 && nameRecord.nameID.fontSpecific != fType);
486
487 record.type = nameRecord.nameID.fontSpecific;
488
489 // Decode the name into UTF-8.
490 const size_t nameOffset = SkEndian_SwapBE16(nameRecord.offset);
491 const size_t nameLength = SkEndian_SwapBE16(nameRecord.length);
492 if (stringTableSize < nameOffset + nameLength) {
493 return false; // continue?
494 }
495 const uint8_t* nameString = stringTable + nameOffset;
496 switch (nameRecord.platformID.value) {
497 case SkOTTableName::Record::PlatformID::Windows:
498 if (SkOTTableName::Record::EncodingID::Windows::UnicodeBMPUCS2
499 != nameRecord.encodingID.windows.value
500 && SkOTTableName::Record::EncodingID::Windows::UnicodeUCS4
501 != nameRecord.encodingID.windows.value
502 && SkOTTableName::Record::EncodingID::Windows::Symbol
503 != nameRecord.encodingID.windows.value)
504 {
505 record.name.reset();
506 break; // continue?
507 }
508 [[fallthrough]];
509 case SkOTTableName::Record::PlatformID::Unicode:
510 case SkOTTableName::Record::PlatformID::ISO:
511 SkString_from_UTF16BE(nameString, nameLength, record.name);
512 break;
513
514 case SkOTTableName::Record::PlatformID::Macintosh:
515 // TODO: need better decoding, especially on Mac.
516 if (SkOTTableName::Record::EncodingID::Macintosh::Roman
517 != nameRecord.encodingID.macintosh.value)
518 {
519 record.name.reset();
520 break; // continue?
521 }
522 SkStringFromMacRoman(nameString, nameLength, record.name);
523 break;
524
525 case SkOTTableName::Record::PlatformID::Custom:
526 // These should never appear in a 'name' table.
527 default:
528 SkASSERT(false);
529 record.name.reset();
530 break; // continue?
531 }
532
533 // Determine the language.
534 const uint16_t languageID = SkEndian_SwapBE16(nameRecord.languageID.languageTagID);
535
536 // Handle format 1 languages.
537 if (SkOTTableName::format_1 == nameTable.format && languageID >= 0x8000) {
538 const uint16_t languageTagRecordIndex = languageID - 0x8000;
539
540 if (nameRecordsSize < sizeof(nameRecord)*nameRecordsCount) {
541 return false; //"und" or break?
542 }
543 const uint8_t* format1extData = nameRecords + sizeof(nameRecord)*nameRecordsCount;
544 size_t format1extSize = nameRecordsSize - sizeof(nameRecord)*nameRecordsCount;
545 SkOTTableName::Format1Ext format1ext;
546 if (format1extSize < sizeof(format1ext)) {
547 return false; // "und" or break?
548 }
549 memcpy(&format1ext, format1extData, sizeof(format1ext));
550
551 const uint8_t* languageTagRecords = format1extData + sizeof(format1ext);
552 size_t languageTagRecordsSize = format1extSize - sizeof(format1ext);
553 if (languageTagRecordIndex < SkEndian_SwapBE16(format1ext.langTagCount)) {
554 SkOTTableName::Format1Ext::LangTagRecord languageTagRecord;
555 if (languageTagRecordsSize < sizeof(languageTagRecord)*(languageTagRecordIndex+1)) {
556 return false; // "und"?
557 }
558 const uint8_t* languageTagData = languageTagRecords
559 + sizeof(languageTagRecord)*languageTagRecordIndex;
560 memcpy(&languageTagRecord, languageTagData, sizeof(languageTagRecord));
561
562 uint16_t languageOffset = SkEndian_SwapBE16(languageTagRecord.offset);
563 uint16_t languageLength = SkEndian_SwapBE16(languageTagRecord.length);
564
565 if (fNameTableSize < stringTableOffset + languageOffset + languageLength) {
566 return false; // "und"?
567 }
568 const uint8_t* languageString = stringTable + languageOffset;
569 SkString_from_UTF16BE(languageString, languageLength, record.language);
570 return true;
571 }
572 }
573
574 // Handle format 0 languages, translating them into BCP 47.
575 const BCP47FromLanguageId target = { languageID, "" };
576 int languageIndex = SkTSearch<BCP47FromLanguageId, BCP47FromLanguageIdLess>(
577 BCP47FromLanguageID, std::size(BCP47FromLanguageID), target, sizeof(target));
578 if (languageIndex >= 0) {
579 record.language = BCP47FromLanguageID[languageIndex].bcp47;
580 return true;
581 }
582
583 // Unknown language, return the BCP 47 code 'und' for 'undetermined'.
584 record.language = "und";
585 return true;
586 }
587