1 /*
2 **********************************************************************
3 * Copyright (C) 2000-2009, International Business Machines
4 * Corporation and others. All Rights Reserved.
5 **********************************************************************
6 * file name: ucnvisci.c
7 * encoding: US-ASCII
8 * tab size: 8 (not used)
9 * indentation:4
10 *
11 * created on: 2001JUN26
12 * created by: Ram Viswanadha
13 *
14 * Date Name Description
15 * 24/7/2001 Ram Added support for EXT character handling
16 */
17
18 #include "unicode/utypes.h"
19
20 #if !UCONFIG_NO_CONVERSION && !UCONFIG_NO_LEGACY_CONVERSION
21
22 #include "cmemory.h"
23 #include "ucnv_bld.h"
24 #include "unicode/ucnv.h"
25 #include "ucnv_cnv.h"
26 #include "unicode/ucnv_cb.h"
27 #include "unicode/uset.h"
28 #include "cstring.h"
29
30 #define UCNV_OPTIONS_VERSION_MASK 0xf
31 #define NUKTA 0x093c
32 #define HALANT 0x094d
33 #define ZWNJ 0x200c /* Zero Width Non Joiner */
34 #define ZWJ 0x200d /* Zero width Joiner */
35 #define INVALID_CHAR 0xffff
36 #define ATR 0xEF /* Attribute code */
37 #define EXT 0xF0 /* Extension code */
38 #define DANDA 0x0964
39 #define DOUBLE_DANDA 0x0965
40 #define ISCII_NUKTA 0xE9
41 #define ISCII_HALANT 0xE8
42 #define ISCII_DANDA 0xEA
43 #define ISCII_INV 0xD9
44 #define ISCII_VOWEL_SIGN_E 0xE0
45 #define INDIC_BLOCK_BEGIN 0x0900
46 #define INDIC_BLOCK_END 0x0D7F
47 #define INDIC_RANGE (INDIC_BLOCK_END - INDIC_BLOCK_BEGIN)
48 #define VOCALLIC_RR 0x0931
49 #define LF 0x0A
50 #define ASCII_END 0xA0
51 #define NO_CHAR_MARKER 0xFFFE
52 #define TELUGU_DELTA DELTA * TELUGU
53 #define DEV_ABBR_SIGN 0x0970
54 #define DEV_ANUDATTA 0x0952
55 #define EXT_RANGE_BEGIN 0xA1
56 #define EXT_RANGE_END 0xEE
57
58 #define PNJ_DELTA 0x0100
59 #define PNJ_BINDI 0x0A02
60 #define PNJ_TIPPI 0x0A70
61 #define PNJ_SIGN_VIRAMA 0x0A4D
62 #define PNJ_ADHAK 0x0A71
63 #define PNJ_HA 0x0A39
64 #define PNJ_RRA 0x0A5C
65
66 static USet* PNJ_BINDI_TIPPI_SET= NULL;
67 static USet* PNJ_CONSONANT_SET= NULL;
68
69 typedef enum {
70 DEVANAGARI =0,
71 BENGALI,
72 GURMUKHI,
73 GUJARATI,
74 ORIYA,
75 TAMIL,
76 TELUGU,
77 KANNADA,
78 MALAYALAM,
79 DELTA=0x80
80 }UniLang;
81
82 /**
83 * Enumeration for switching code pages if <ATR>+<one of below values>
84 * is encountered
85 */
86 typedef enum {
87 DEF = 0x40,
88 RMN = 0x41,
89 DEV = 0x42,
90 BNG = 0x43,
91 TML = 0x44,
92 TLG = 0x45,
93 ASM = 0x46,
94 ORI = 0x47,
95 KND = 0x48,
96 MLM = 0x49,
97 GJR = 0x4A,
98 PNJ = 0x4B,
99 ARB = 0x71,
100 PES = 0x72,
101 URD = 0x73,
102 SND = 0x74,
103 KSM = 0x75,
104 PST = 0x76
105 }ISCIILang;
106
107 typedef enum {
108 DEV_MASK =0x80,
109 PNJ_MASK =0x40,
110 GJR_MASK =0x20,
111 ORI_MASK =0x10,
112 BNG_MASK =0x08,
113 KND_MASK =0x04,
114 MLM_MASK =0x02,
115 TML_MASK =0x01,
116 ZERO =0x00
117 }MaskEnum;
118
119 #define ISCII_CNV_PREFIX "ISCII,version="
120
121 typedef struct {
122 UChar contextCharToUnicode; /* previous Unicode codepoint for contextual analysis */
123 UChar contextCharFromUnicode; /* previous Unicode codepoint for contextual analysis */
124 uint16_t defDeltaToUnicode; /* delta for switching to default state when DEF is encountered */
125 uint16_t currentDeltaFromUnicode; /* current delta in Indic block */
126 uint16_t currentDeltaToUnicode; /* current delta in Indic block */
127 MaskEnum currentMaskFromUnicode; /* mask for current state in toUnicode */
128 MaskEnum currentMaskToUnicode; /* mask for current state in toUnicode */
129 MaskEnum defMaskToUnicode; /* mask for default state in toUnicode */
130 UBool isFirstBuffer; /* boolean for fromUnicode to see if we need to announce the first script */
131 UBool resetToDefaultToUnicode; /* boolean for reseting to default delta and mask when a newline is encountered*/
132 char name[sizeof(ISCII_CNV_PREFIX) + 1];
133 UChar32 prevToUnicodeStatus; /* Hold the previous toUnicodeStatus. This is necessary because we may need to know the last two code points. */
134 } UConverterDataISCII;
135
136 typedef struct LookupDataStruct {
137 UniLang uniLang;
138 MaskEnum maskEnum;
139 ISCIILang isciiLang;
140 } LookupDataStruct;
141
142 static const LookupDataStruct lookupInitialData[]={
143 { DEVANAGARI, DEV_MASK, DEV },
144 { BENGALI, BNG_MASK, BNG },
145 { GURMUKHI, PNJ_MASK, PNJ },
146 { GUJARATI, GJR_MASK, GJR },
147 { ORIYA, ORI_MASK, ORI },
148 { TAMIL, TML_MASK, TML },
149 { TELUGU, KND_MASK, TLG },
150 { KANNADA, KND_MASK, KND },
151 { MALAYALAM, MLM_MASK, MLM }
152 };
153
initializeSets()154 static void initializeSets() {
155 /* TODO: Replace the following two lines with PNJ_CONSONANT_SET = uset_openEmpty(); */
156 PNJ_CONSONANT_SET = uset_open(0,0);
157 uset_clear(PNJ_CONSONANT_SET);
158
159 uset_addRange(PNJ_CONSONANT_SET, 0x0A15, 0x0A28);
160 uset_addRange(PNJ_CONSONANT_SET, 0x0A2A, 0x0A30);
161 uset_addRange(PNJ_CONSONANT_SET, 0x0A35, 0x0A36);
162 uset_addRange(PNJ_CONSONANT_SET, 0x0A38, 0x0A39);
163
164 PNJ_BINDI_TIPPI_SET = uset_clone(PNJ_CONSONANT_SET);
165 uset_add(PNJ_BINDI_TIPPI_SET, 0x0A05);
166 uset_add(PNJ_BINDI_TIPPI_SET, 0x0A07);
167 uset_add(PNJ_BINDI_TIPPI_SET, 0x0A3F);
168 uset_addRange(PNJ_BINDI_TIPPI_SET, 0x0A41, 0x0A42);
169
170 uset_compact(PNJ_CONSONANT_SET);
171 uset_compact(PNJ_BINDI_TIPPI_SET);
172 }
173
_ISCIIOpen(UConverter * cnv,UConverterLoadArgs * pArgs,UErrorCode * errorCode)174 static void _ISCIIOpen(UConverter *cnv, UConverterLoadArgs *pArgs, UErrorCode *errorCode) {
175 if(pArgs->onlyTestIsLoadable) {
176 return;
177 }
178
179 /* Ensure that the sets used in special handling of certain Gurmukhi characters are initialized. */
180 initializeSets();
181
182 cnv->extraInfo = uprv_malloc(sizeof(UConverterDataISCII));
183
184 if (cnv->extraInfo != NULL) {
185 int32_t len=0;
186 UConverterDataISCII *converterData=
187 (UConverterDataISCII *) cnv->extraInfo;
188 converterData->contextCharToUnicode=NO_CHAR_MARKER;
189 cnv->toUnicodeStatus = missingCharMarker;
190 converterData->contextCharFromUnicode=0x0000;
191 converterData->resetToDefaultToUnicode=FALSE;
192 /* check if the version requested is supported */
193 if ((pArgs->options & UCNV_OPTIONS_VERSION_MASK) < 9) {
194 /* initialize state variables */
195 converterData->currentDeltaFromUnicode
196 = converterData->currentDeltaToUnicode
197 = converterData->defDeltaToUnicode = (uint16_t)(lookupInitialData[pArgs->options & UCNV_OPTIONS_VERSION_MASK].uniLang * DELTA);
198
199 converterData->currentMaskFromUnicode
200 = converterData->currentMaskToUnicode
201 = converterData->defMaskToUnicode = lookupInitialData[pArgs->options & UCNV_OPTIONS_VERSION_MASK].maskEnum;
202
203 converterData->isFirstBuffer=TRUE;
204 (void)uprv_strcpy(converterData->name, ISCII_CNV_PREFIX);
205 len = (int32_t)uprv_strlen(converterData->name);
206 converterData->name[len]= (char)((pArgs->options & UCNV_OPTIONS_VERSION_MASK) + '0');
207 converterData->name[len+1]=0;
208
209 converterData->prevToUnicodeStatus = 0x0000;
210 } else {
211 uprv_free(cnv->extraInfo);
212 cnv->extraInfo = NULL;
213 *errorCode = U_ILLEGAL_ARGUMENT_ERROR;
214 }
215
216 } else {
217 *errorCode =U_MEMORY_ALLOCATION_ERROR;
218 }
219 }
220
_ISCIIClose(UConverter * cnv)221 static void _ISCIIClose(UConverter *cnv) {
222 if (cnv->extraInfo!=NULL) {
223 if (!cnv->isExtraLocal) {
224 uprv_free(cnv->extraInfo);
225 }
226 cnv->extraInfo=NULL;
227 }
228 if (PNJ_CONSONANT_SET != NULL) {
229 uset_close(PNJ_CONSONANT_SET);
230 PNJ_CONSONANT_SET = NULL;
231 }
232 if (PNJ_BINDI_TIPPI_SET != NULL) {
233 uset_close(PNJ_BINDI_TIPPI_SET);
234 PNJ_BINDI_TIPPI_SET = NULL;
235 }
236 }
237
_ISCIIgetName(const UConverter * cnv)238 static const char* _ISCIIgetName(const UConverter* cnv) {
239 if (cnv->extraInfo) {
240 UConverterDataISCII* myData= (UConverterDataISCII*)cnv->extraInfo;
241 return myData->name;
242 }
243 return NULL;
244 }
245
_ISCIIReset(UConverter * cnv,UConverterResetChoice choice)246 static void _ISCIIReset(UConverter *cnv, UConverterResetChoice choice) {
247 UConverterDataISCII* data =(UConverterDataISCII *) (cnv->extraInfo);
248 if (choice<=UCNV_RESET_TO_UNICODE) {
249 cnv->toUnicodeStatus = missingCharMarker;
250 cnv->mode=0;
251 data->currentDeltaToUnicode=data->defDeltaToUnicode;
252 data->currentMaskToUnicode = data->defMaskToUnicode;
253 data->contextCharToUnicode=NO_CHAR_MARKER;
254 data->prevToUnicodeStatus = 0x0000;
255 }
256 if (choice!=UCNV_RESET_TO_UNICODE) {
257 cnv->fromUChar32=0x0000;
258 data->contextCharFromUnicode=0x00;
259 data->currentMaskFromUnicode=data->defMaskToUnicode;
260 data->currentDeltaFromUnicode=data->defDeltaToUnicode;
261 data->isFirstBuffer=TRUE;
262 data->resetToDefaultToUnicode=FALSE;
263 }
264 }
265
266 /**
267 * The values in validity table are indexed by the lower bits of Unicode
268 * range 0x0900 - 0x09ff. The values have a structure like:
269 * ---------------------------------------------------------------
270 * | DEV | PNJ | GJR | ORI | BNG | TLG | MLM | TML |
271 * | | | | | ASM | KND | | |
272 * ---------------------------------------------------------------
273 * If a code point is valid in a particular script
274 * then that bit is turned on
275 *
276 * Unicode does not distinguish between Bengali and Assamese so we use 1 bit for
277 * to represent these languages
278 *
279 * Telugu and Kannada have same codepoints except for Vocallic_RR which we special case
280 * and combine and use 1 bit to represent these languages.
281 *
282 * TODO: It is probably easier to understand and maintain to change this
283 * to use uint16_t and give each of the 9 Unicode/script blocks its own bit.
284 */
285
286 static const uint8_t validityTable[128] = {
287 /* This state table is tool generated please do not edit unless you know exactly what you are doing */
288 /* Note: This table was edited to mirror the Windows XP implementation */
289 /*ISCII:Valid:Unicode */
290 /*0xa0 : 0x00: 0x900 */ ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
291 /*0xa1 : 0xb8: 0x901 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + ZERO + ZERO + ZERO ,
292 /*0xa2 : 0xfe: 0x902 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
293 /*0xa3 : 0xbf: 0x903 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
294 /*0x00 : 0x00: 0x904 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
295 /*0xa4 : 0xff: 0x905 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
296 /*0xa5 : 0xff: 0x906 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
297 /*0xa6 : 0xff: 0x907 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
298 /*0xa7 : 0xff: 0x908 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
299 /*0xa8 : 0xff: 0x909 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
300 /*0xa9 : 0xff: 0x90a */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
301 /*0xaa : 0xfe: 0x90b */ DEV_MASK + ZERO + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
302 /*0x00 : 0x00: 0x90c */ DEV_MASK + ZERO + ZERO + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
303 /*0xae : 0x80: 0x90d */ DEV_MASK + ZERO + GJR_MASK + ZERO + ZERO + ZERO + ZERO + ZERO ,
304 /*0xab : 0x87: 0x90e */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + KND_MASK + MLM_MASK + TML_MASK ,
305 /*0xac : 0xff: 0x90f */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
306 /*0xad : 0xff: 0x910 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
307 /*0xb2 : 0x80: 0x911 */ DEV_MASK + ZERO + GJR_MASK + ZERO + ZERO + ZERO + ZERO + ZERO ,
308 /*0xaf : 0x87: 0x912 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + KND_MASK + MLM_MASK + TML_MASK ,
309 /*0xb0 : 0xff: 0x913 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
310 /*0xb1 : 0xff: 0x914 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
311 /*0xb3 : 0xff: 0x915 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
312 /*0xb4 : 0xfe: 0x916 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
313 /*0xb5 : 0xfe: 0x917 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
314 /*0xb6 : 0xfe: 0x918 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
315 /*0xb7 : 0xff: 0x919 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
316 /*0xb8 : 0xff: 0x91a */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
317 /*0xb9 : 0xfe: 0x91b */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
318 /*0xba : 0xff: 0x91c */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
319 /*0xbb : 0xfe: 0x91d */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
320 /*0xbc : 0xff: 0x91e */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
321 /*0xbd : 0xff: 0x91f */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
322 /*0xbe : 0xfe: 0x920 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
323 /*0xbf : 0xfe: 0x921 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
324 /*0xc0 : 0xfe: 0x922 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
325 /*0xc1 : 0xff: 0x923 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
326 /*0xc2 : 0xff: 0x924 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
327 /*0xc3 : 0xfe: 0x925 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
328 /*0xc4 : 0xfe: 0x926 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
329 /*0xc5 : 0xfe: 0x927 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
330 /*0xc6 : 0xff: 0x928 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
331 /*0xc7 : 0x81: 0x929 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + TML_MASK ,
332 /*0xc8 : 0xff: 0x92a */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
333 /*0xc9 : 0xfe: 0x92b */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
334 /*0xca : 0xfe: 0x92c */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
335 /*0xcb : 0xfe: 0x92d */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
336 /*0xcc : 0xfe: 0x92e */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
337 /*0xcd : 0xff: 0x92f */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
338 /*0xcf : 0xff: 0x930 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
339 /*0xd0 : 0x87: 0x931 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + MLM_MASK + TML_MASK ,
340 /*0xd1 : 0xff: 0x932 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
341 /*0xd2 : 0xb7: 0x933 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + ZERO + KND_MASK + MLM_MASK + TML_MASK ,
342 /*0xd3 : 0x83: 0x934 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + MLM_MASK + TML_MASK ,
343 /*0xd4 : 0xff: 0x935 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + ZERO + KND_MASK + MLM_MASK + TML_MASK ,
344 /*0xd5 : 0xfe: 0x936 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
345 /*0xd6 : 0xbf: 0x937 */ DEV_MASK + ZERO + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
346 /*0xd7 : 0xff: 0x938 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
347 /*0xd8 : 0xff: 0x939 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
348 /*0x00 : 0x00: 0x93A */ ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
349 /*0x00 : 0x00: 0x93B */ ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
350 /*0xe9 : 0xda: 0x93c */ DEV_MASK + PNJ_MASK + ZERO + ORI_MASK + BNG_MASK + ZERO + ZERO + ZERO ,
351 /*0x00 : 0x00: 0x93d */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
352 /*0xda : 0xff: 0x93e */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
353 /*0xdb : 0xff: 0x93f */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
354 /*0xdc : 0xff: 0x940 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
355 /*0xdd : 0xff: 0x941 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
356 /*0xde : 0xff: 0x942 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
357 /*0xdf : 0xbe: 0x943 */ DEV_MASK + ZERO + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
358 /*0x00 : 0x00: 0x944 */ DEV_MASK + ZERO + GJR_MASK + ZERO + BNG_MASK + KND_MASK + ZERO + ZERO ,
359 /*0xe3 : 0x80: 0x945 */ DEV_MASK + ZERO + GJR_MASK + ZERO + ZERO + ZERO + ZERO + ZERO ,
360 /*0xe0 : 0x87: 0x946 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + KND_MASK + MLM_MASK + TML_MASK ,
361 /*0xe1 : 0xff: 0x947 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
362 /*0xe2 : 0xff: 0x948 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
363 /*0xe7 : 0x80: 0x949 */ DEV_MASK + ZERO + GJR_MASK + ZERO + ZERO + ZERO + ZERO + ZERO ,
364 /*0xe4 : 0x87: 0x94a */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + KND_MASK + MLM_MASK + TML_MASK ,
365 /*0xe5 : 0xff: 0x94b */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
366 /*0xe6 : 0xff: 0x94c */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
367 /*0xe8 : 0xff: 0x94d */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
368 /*0xec : 0x00: 0x94e */ ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
369 /*0xed : 0x00: 0x94f */ ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
370 /*0x00 : 0x00: 0x950 */ DEV_MASK + ZERO + GJR_MASK + ZERO + ZERO + ZERO + ZERO + ZERO ,
371 /*0x00 : 0x00: 0x951 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
372 /*0x00 : 0x00: 0x952 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
373 /*0x00 : 0x00: 0x953 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
374 /*0x00 : 0x00: 0x954 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
375 /*0x00 : 0x00: 0x955 */ ZERO + ZERO + ZERO + ZERO + ZERO + KND_MASK + ZERO + ZERO ,
376 /*0x00 : 0x00: 0x956 */ ZERO + ZERO + ZERO + ORI_MASK + ZERO + KND_MASK + ZERO + ZERO ,
377 /*0x00 : 0x00: 0x957 */ ZERO + ZERO + ZERO + ORI_MASK + BNG_MASK + ZERO + MLM_MASK + ZERO ,
378 /*0x00 : 0x00: 0x958 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
379 /*0x00 : 0x00: 0x959 */ DEV_MASK + PNJ_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
380 /*0x00 : 0x00: 0x95a */ DEV_MASK + PNJ_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
381 /*0x00 : 0x00: 0x95b */ DEV_MASK + PNJ_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
382 /*0x00 : 0x00: 0x95c */ DEV_MASK + PNJ_MASK + ZERO + ZERO + BNG_MASK + ZERO + ZERO + ZERO ,
383 /*0x00 : 0x00: 0x95d */ DEV_MASK + ZERO + ZERO + ORI_MASK + BNG_MASK + ZERO + ZERO + ZERO ,
384 /*0x00 : 0x00: 0x95e */ DEV_MASK + PNJ_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
385 /*0xce : 0x98: 0x95f */ DEV_MASK + ZERO + ZERO + ORI_MASK + BNG_MASK + ZERO + ZERO + ZERO ,
386 /*0x00 : 0x00: 0x960 */ DEV_MASK + ZERO + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
387 /*0x00 : 0x00: 0x961 */ DEV_MASK + ZERO + ZERO + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + ZERO ,
388 /*0x00 : 0x00: 0x962 */ DEV_MASK + ZERO + ZERO + ZERO + BNG_MASK + ZERO + ZERO + ZERO ,
389 /*0x00 : 0x00: 0x963 */ DEV_MASK + ZERO + ZERO + ZERO + BNG_MASK + ZERO + ZERO + ZERO ,
390 /*0xea : 0xf8: 0x964 */ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
391 /*0xeaea : 0x00: 0x965*/ DEV_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
392 /*0xf1 : 0xff: 0x966 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
393 /*0xf2 : 0xff: 0x967 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
394 /*0xf3 : 0xff: 0x968 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
395 /*0xf4 : 0xff: 0x969 */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
396 /*0xf5 : 0xff: 0x96a */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
397 /*0xf6 : 0xff: 0x96b */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
398 /*0xf7 : 0xff: 0x96c */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
399 /*0xf8 : 0xff: 0x96d */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
400 /*0xf9 : 0xff: 0x96e */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
401 /*0xfa : 0xff: 0x96f */ DEV_MASK + PNJ_MASK + GJR_MASK + ORI_MASK + BNG_MASK + KND_MASK + MLM_MASK + TML_MASK ,
402 /*0x00 : 0x80: 0x970 */ DEV_MASK + PNJ_MASK + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO ,
403 /*
404 * The length of the array is 128 to provide values for 0x900..0x97f.
405 * The last 15 entries for 0x971..0x97f of the validity table are all zero
406 * because no Indic script uses such Unicode code points.
407 */
408 /*0x00 : 0x00: 0x9yz */ ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO + ZERO
409 };
410
411 static const uint16_t fromUnicodeTable[128]={
412 0x00a0 ,/* 0x0900 */
413 0x00a1 ,/* 0x0901 */
414 0x00a2 ,/* 0x0902 */
415 0x00a3 ,/* 0x0903 */
416 0xa4e0 ,/* 0x0904 */
417 0x00a4 ,/* 0x0905 */
418 0x00a5 ,/* 0x0906 */
419 0x00a6 ,/* 0x0907 */
420 0x00a7 ,/* 0x0908 */
421 0x00a8 ,/* 0x0909 */
422 0x00a9 ,/* 0x090a */
423 0x00aa ,/* 0x090b */
424 0xA6E9 ,/* 0x090c */
425 0x00ae ,/* 0x090d */
426 0x00ab ,/* 0x090e */
427 0x00ac ,/* 0x090f */
428 0x00ad ,/* 0x0910 */
429 0x00b2 ,/* 0x0911 */
430 0x00af ,/* 0x0912 */
431 0x00b0 ,/* 0x0913 */
432 0x00b1 ,/* 0x0914 */
433 0x00b3 ,/* 0x0915 */
434 0x00b4 ,/* 0x0916 */
435 0x00b5 ,/* 0x0917 */
436 0x00b6 ,/* 0x0918 */
437 0x00b7 ,/* 0x0919 */
438 0x00b8 ,/* 0x091a */
439 0x00b9 ,/* 0x091b */
440 0x00ba ,/* 0x091c */
441 0x00bb ,/* 0x091d */
442 0x00bc ,/* 0x091e */
443 0x00bd ,/* 0x091f */
444 0x00be ,/* 0x0920 */
445 0x00bf ,/* 0x0921 */
446 0x00c0 ,/* 0x0922 */
447 0x00c1 ,/* 0x0923 */
448 0x00c2 ,/* 0x0924 */
449 0x00c3 ,/* 0x0925 */
450 0x00c4 ,/* 0x0926 */
451 0x00c5 ,/* 0x0927 */
452 0x00c6 ,/* 0x0928 */
453 0x00c7 ,/* 0x0929 */
454 0x00c8 ,/* 0x092a */
455 0x00c9 ,/* 0x092b */
456 0x00ca ,/* 0x092c */
457 0x00cb ,/* 0x092d */
458 0x00cc ,/* 0x092e */
459 0x00cd ,/* 0x092f */
460 0x00cf ,/* 0x0930 */
461 0x00d0 ,/* 0x0931 */
462 0x00d1 ,/* 0x0932 */
463 0x00d2 ,/* 0x0933 */
464 0x00d3 ,/* 0x0934 */
465 0x00d4 ,/* 0x0935 */
466 0x00d5 ,/* 0x0936 */
467 0x00d6 ,/* 0x0937 */
468 0x00d7 ,/* 0x0938 */
469 0x00d8 ,/* 0x0939 */
470 0xFFFF ,/* 0x093A */
471 0xFFFF ,/* 0x093B */
472 0x00e9 ,/* 0x093c */
473 0xEAE9 ,/* 0x093d */
474 0x00da ,/* 0x093e */
475 0x00db ,/* 0x093f */
476 0x00dc ,/* 0x0940 */
477 0x00dd ,/* 0x0941 */
478 0x00de ,/* 0x0942 */
479 0x00df ,/* 0x0943 */
480 0xDFE9 ,/* 0x0944 */
481 0x00e3 ,/* 0x0945 */
482 0x00e0 ,/* 0x0946 */
483 0x00e1 ,/* 0x0947 */
484 0x00e2 ,/* 0x0948 */
485 0x00e7 ,/* 0x0949 */
486 0x00e4 ,/* 0x094a */
487 0x00e5 ,/* 0x094b */
488 0x00e6 ,/* 0x094c */
489 0x00e8 ,/* 0x094d */
490 0x00ec ,/* 0x094e */
491 0x00ed ,/* 0x094f */
492 0xA1E9 ,/* 0x0950 */ /* OM Symbol */
493 0xFFFF ,/* 0x0951 */
494 0xF0B8 ,/* 0x0952 */
495 0xFFFF ,/* 0x0953 */
496 0xFFFF ,/* 0x0954 */
497 0xFFFF ,/* 0x0955 */
498 0xFFFF ,/* 0x0956 */
499 0xFFFF ,/* 0x0957 */
500 0xb3e9 ,/* 0x0958 */
501 0xb4e9 ,/* 0x0959 */
502 0xb5e9 ,/* 0x095a */
503 0xbae9 ,/* 0x095b */
504 0xbfe9 ,/* 0x095c */
505 0xC0E9 ,/* 0x095d */
506 0xc9e9 ,/* 0x095e */
507 0x00ce ,/* 0x095f */
508 0xAAe9 ,/* 0x0960 */
509 0xA7E9 ,/* 0x0961 */
510 0xDBE9 ,/* 0x0962 */
511 0xDCE9 ,/* 0x0963 */
512 0x00ea ,/* 0x0964 */
513 0xeaea ,/* 0x0965 */
514 0x00f1 ,/* 0x0966 */
515 0x00f2 ,/* 0x0967 */
516 0x00f3 ,/* 0x0968 */
517 0x00f4 ,/* 0x0969 */
518 0x00f5 ,/* 0x096a */
519 0x00f6 ,/* 0x096b */
520 0x00f7 ,/* 0x096c */
521 0x00f8 ,/* 0x096d */
522 0x00f9 ,/* 0x096e */
523 0x00fa ,/* 0x096f */
524 0xF0BF ,/* 0x0970 */
525 0xFFFF ,/* 0x0971 */
526 0xFFFF ,/* 0x0972 */
527 0xFFFF ,/* 0x0973 */
528 0xFFFF ,/* 0x0974 */
529 0xFFFF ,/* 0x0975 */
530 0xFFFF ,/* 0x0976 */
531 0xFFFF ,/* 0x0977 */
532 0xFFFF ,/* 0x0978 */
533 0xFFFF ,/* 0x0979 */
534 0xFFFF ,/* 0x097a */
535 0xFFFF ,/* 0x097b */
536 0xFFFF ,/* 0x097c */
537 0xFFFF ,/* 0x097d */
538 0xFFFF ,/* 0x097e */
539 0xFFFF ,/* 0x097f */
540 };
541 static const uint16_t toUnicodeTable[256]={
542 0x0000,/* 0x00 */
543 0x0001,/* 0x01 */
544 0x0002,/* 0x02 */
545 0x0003,/* 0x03 */
546 0x0004,/* 0x04 */
547 0x0005,/* 0x05 */
548 0x0006,/* 0x06 */
549 0x0007,/* 0x07 */
550 0x0008,/* 0x08 */
551 0x0009,/* 0x09 */
552 0x000a,/* 0x0a */
553 0x000b,/* 0x0b */
554 0x000c,/* 0x0c */
555 0x000d,/* 0x0d */
556 0x000e,/* 0x0e */
557 0x000f,/* 0x0f */
558 0x0010,/* 0x10 */
559 0x0011,/* 0x11 */
560 0x0012,/* 0x12 */
561 0x0013,/* 0x13 */
562 0x0014,/* 0x14 */
563 0x0015,/* 0x15 */
564 0x0016,/* 0x16 */
565 0x0017,/* 0x17 */
566 0x0018,/* 0x18 */
567 0x0019,/* 0x19 */
568 0x001a,/* 0x1a */
569 0x001b,/* 0x1b */
570 0x001c,/* 0x1c */
571 0x001d,/* 0x1d */
572 0x001e,/* 0x1e */
573 0x001f,/* 0x1f */
574 0x0020,/* 0x20 */
575 0x0021,/* 0x21 */
576 0x0022,/* 0x22 */
577 0x0023,/* 0x23 */
578 0x0024,/* 0x24 */
579 0x0025,/* 0x25 */
580 0x0026,/* 0x26 */
581 0x0027,/* 0x27 */
582 0x0028,/* 0x28 */
583 0x0029,/* 0x29 */
584 0x002a,/* 0x2a */
585 0x002b,/* 0x2b */
586 0x002c,/* 0x2c */
587 0x002d,/* 0x2d */
588 0x002e,/* 0x2e */
589 0x002f,/* 0x2f */
590 0x0030,/* 0x30 */
591 0x0031,/* 0x31 */
592 0x0032,/* 0x32 */
593 0x0033,/* 0x33 */
594 0x0034,/* 0x34 */
595 0x0035,/* 0x35 */
596 0x0036,/* 0x36 */
597 0x0037,/* 0x37 */
598 0x0038,/* 0x38 */
599 0x0039,/* 0x39 */
600 0x003A,/* 0x3A */
601 0x003B,/* 0x3B */
602 0x003c,/* 0x3c */
603 0x003d,/* 0x3d */
604 0x003e,/* 0x3e */
605 0x003f,/* 0x3f */
606 0x0040,/* 0x40 */
607 0x0041,/* 0x41 */
608 0x0042,/* 0x42 */
609 0x0043,/* 0x43 */
610 0x0044,/* 0x44 */
611 0x0045,/* 0x45 */
612 0x0046,/* 0x46 */
613 0x0047,/* 0x47 */
614 0x0048,/* 0x48 */
615 0x0049,/* 0x49 */
616 0x004a,/* 0x4a */
617 0x004b,/* 0x4b */
618 0x004c,/* 0x4c */
619 0x004d,/* 0x4d */
620 0x004e,/* 0x4e */
621 0x004f,/* 0x4f */
622 0x0050,/* 0x50 */
623 0x0051,/* 0x51 */
624 0x0052,/* 0x52 */
625 0x0053,/* 0x53 */
626 0x0054,/* 0x54 */
627 0x0055,/* 0x55 */
628 0x0056,/* 0x56 */
629 0x0057,/* 0x57 */
630 0x0058,/* 0x58 */
631 0x0059,/* 0x59 */
632 0x005a,/* 0x5a */
633 0x005b,/* 0x5b */
634 0x005c,/* 0x5c */
635 0x005d,/* 0x5d */
636 0x005e,/* 0x5e */
637 0x005f,/* 0x5f */
638 0x0060,/* 0x60 */
639 0x0061,/* 0x61 */
640 0x0062,/* 0x62 */
641 0x0063,/* 0x63 */
642 0x0064,/* 0x64 */
643 0x0065,/* 0x65 */
644 0x0066,/* 0x66 */
645 0x0067,/* 0x67 */
646 0x0068,/* 0x68 */
647 0x0069,/* 0x69 */
648 0x006a,/* 0x6a */
649 0x006b,/* 0x6b */
650 0x006c,/* 0x6c */
651 0x006d,/* 0x6d */
652 0x006e,/* 0x6e */
653 0x006f,/* 0x6f */
654 0x0070,/* 0x70 */
655 0x0071,/* 0x71 */
656 0x0072,/* 0x72 */
657 0x0073,/* 0x73 */
658 0x0074,/* 0x74 */
659 0x0075,/* 0x75 */
660 0x0076,/* 0x76 */
661 0x0077,/* 0x77 */
662 0x0078,/* 0x78 */
663 0x0079,/* 0x79 */
664 0x007a,/* 0x7a */
665 0x007b,/* 0x7b */
666 0x007c,/* 0x7c */
667 0x007d,/* 0x7d */
668 0x007e,/* 0x7e */
669 0x007f,/* 0x7f */
670 0x0080,/* 0x80 */
671 0x0081,/* 0x81 */
672 0x0082,/* 0x82 */
673 0x0083,/* 0x83 */
674 0x0084,/* 0x84 */
675 0x0085,/* 0x85 */
676 0x0086,/* 0x86 */
677 0x0087,/* 0x87 */
678 0x0088,/* 0x88 */
679 0x0089,/* 0x89 */
680 0x008a,/* 0x8a */
681 0x008b,/* 0x8b */
682 0x008c,/* 0x8c */
683 0x008d,/* 0x8d */
684 0x008e,/* 0x8e */
685 0x008f,/* 0x8f */
686 0x0090,/* 0x90 */
687 0x0091,/* 0x91 */
688 0x0092,/* 0x92 */
689 0x0093,/* 0x93 */
690 0x0094,/* 0x94 */
691 0x0095,/* 0x95 */
692 0x0096,/* 0x96 */
693 0x0097,/* 0x97 */
694 0x0098,/* 0x98 */
695 0x0099,/* 0x99 */
696 0x009a,/* 0x9a */
697 0x009b,/* 0x9b */
698 0x009c,/* 0x9c */
699 0x009d,/* 0x9d */
700 0x009e,/* 0x9e */
701 0x009f,/* 0x9f */
702 0x00A0,/* 0xa0 */
703 0x0901,/* 0xa1 */
704 0x0902,/* 0xa2 */
705 0x0903,/* 0xa3 */
706 0x0905,/* 0xa4 */
707 0x0906,/* 0xa5 */
708 0x0907,/* 0xa6 */
709 0x0908,/* 0xa7 */
710 0x0909,/* 0xa8 */
711 0x090a,/* 0xa9 */
712 0x090b,/* 0xaa */
713 0x090e,/* 0xab */
714 0x090f,/* 0xac */
715 0x0910,/* 0xad */
716 0x090d,/* 0xae */
717 0x0912,/* 0xaf */
718 0x0913,/* 0xb0 */
719 0x0914,/* 0xb1 */
720 0x0911,/* 0xb2 */
721 0x0915,/* 0xb3 */
722 0x0916,/* 0xb4 */
723 0x0917,/* 0xb5 */
724 0x0918,/* 0xb6 */
725 0x0919,/* 0xb7 */
726 0x091a,/* 0xb8 */
727 0x091b,/* 0xb9 */
728 0x091c,/* 0xba */
729 0x091d,/* 0xbb */
730 0x091e,/* 0xbc */
731 0x091f,/* 0xbd */
732 0x0920,/* 0xbe */
733 0x0921,/* 0xbf */
734 0x0922,/* 0xc0 */
735 0x0923,/* 0xc1 */
736 0x0924,/* 0xc2 */
737 0x0925,/* 0xc3 */
738 0x0926,/* 0xc4 */
739 0x0927,/* 0xc5 */
740 0x0928,/* 0xc6 */
741 0x0929,/* 0xc7 */
742 0x092a,/* 0xc8 */
743 0x092b,/* 0xc9 */
744 0x092c,/* 0xca */
745 0x092d,/* 0xcb */
746 0x092e,/* 0xcc */
747 0x092f,/* 0xcd */
748 0x095f,/* 0xce */
749 0x0930,/* 0xcf */
750 0x0931,/* 0xd0 */
751 0x0932,/* 0xd1 */
752 0x0933,/* 0xd2 */
753 0x0934,/* 0xd3 */
754 0x0935,/* 0xd4 */
755 0x0936,/* 0xd5 */
756 0x0937,/* 0xd6 */
757 0x0938,/* 0xd7 */
758 0x0939,/* 0xd8 */
759 0x200D,/* 0xd9 */
760 0x093e,/* 0xda */
761 0x093f,/* 0xdb */
762 0x0940,/* 0xdc */
763 0x0941,/* 0xdd */
764 0x0942,/* 0xde */
765 0x0943,/* 0xdf */
766 0x0946,/* 0xe0 */
767 0x0947,/* 0xe1 */
768 0x0948,/* 0xe2 */
769 0x0945,/* 0xe3 */
770 0x094a,/* 0xe4 */
771 0x094b,/* 0xe5 */
772 0x094c,/* 0xe6 */
773 0x0949,/* 0xe7 */
774 0x094d,/* 0xe8 */
775 0x093c,/* 0xe9 */
776 0x0964,/* 0xea */
777 0xFFFF,/* 0xeb */
778 0xFFFF,/* 0xec */
779 0xFFFF,/* 0xed */
780 0xFFFF,/* 0xee */
781 0xFFFF,/* 0xef */
782 0xFFFF,/* 0xf0 */
783 0x0966,/* 0xf1 */
784 0x0967,/* 0xf2 */
785 0x0968,/* 0xf3 */
786 0x0969,/* 0xf4 */
787 0x096a,/* 0xf5 */
788 0x096b,/* 0xf6 */
789 0x096c,/* 0xf7 */
790 0x096d,/* 0xf8 */
791 0x096e,/* 0xf9 */
792 0x096f,/* 0xfa */
793 0xFFFF,/* 0xfb */
794 0xFFFF,/* 0xfc */
795 0xFFFF,/* 0xfd */
796 0xFFFF,/* 0xfe */
797 0xFFFF /* 0xff */
798 };
799
800 static const uint16_t vowelSignESpecialCases[][2]={
801 { 2 /*length of array*/ , 0 },
802 { 0xA4 , 0x0904 },
803 };
804
805 static const uint16_t nuktaSpecialCases[][2]={
806 { 16 /*length of array*/ , 0 },
807 { 0xA6 , 0x090c },
808 { 0xEA , 0x093D },
809 { 0xDF , 0x0944 },
810 { 0xA1 , 0x0950 },
811 { 0xb3 , 0x0958 },
812 { 0xb4 , 0x0959 },
813 { 0xb5 , 0x095a },
814 { 0xba , 0x095b },
815 { 0xbf , 0x095c },
816 { 0xC0 , 0x095d },
817 { 0xc9 , 0x095e },
818 { 0xAA , 0x0960 },
819 { 0xA7 , 0x0961 },
820 { 0xDB , 0x0962 },
821 { 0xDC , 0x0963 },
822 };
823
824
825 #define WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,targetByteUnit,err){ \
826 int32_t offset = (int32_t)(source - args->source-1); \
827 /* write the targetUniChar to target */ \
828 if(target < targetLimit){ \
829 if(targetByteUnit <= 0xFF){ \
830 *(target)++ = (uint8_t)(targetByteUnit); \
831 if(offsets){ \
832 *(offsets++) = offset; \
833 } \
834 }else{ \
835 if (targetByteUnit > 0xFFFF) { \
836 *(target)++ = (uint8_t)(targetByteUnit>>16); \
837 if (offsets) { \
838 --offset; \
839 *(offsets++) = offset; \
840 } \
841 } \
842 if (!(target < targetLimit)) { \
843 args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = \
844 (uint8_t)(targetByteUnit >> 8); \
845 args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = \
846 (uint8_t)targetByteUnit; \
847 *err = U_BUFFER_OVERFLOW_ERROR; \
848 } else { \
849 *(target)++ = (uint8_t)(targetByteUnit>>8); \
850 if(offsets){ \
851 *(offsets++) = offset; \
852 } \
853 if(target < targetLimit){ \
854 *(target)++ = (uint8_t) targetByteUnit; \
855 if(offsets){ \
856 *(offsets++) = offset ; \
857 } \
858 }else{ \
859 args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] =\
860 (uint8_t) (targetByteUnit); \
861 *err = U_BUFFER_OVERFLOW_ERROR; \
862 } \
863 } \
864 } \
865 }else{ \
866 if (targetByteUnit & 0xFF0000) { \
867 args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = \
868 (uint8_t) (targetByteUnit >>16); \
869 } \
870 if(targetByteUnit & 0xFF00){ \
871 args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = \
872 (uint8_t) (targetByteUnit >>8); \
873 } \
874 args->converter->charErrorBuffer[args->converter->charErrorBufferLength++] = \
875 (uint8_t) (targetByteUnit); \
876 *err = U_BUFFER_OVERFLOW_ERROR; \
877 } \
878 }
879
880 /* Rules:
881 * Explicit Halant :
882 * <HALANT> + <ZWNJ>
883 * Soft Halant :
884 * <HALANT> + <ZWJ>
885 */
886
UConverter_fromUnicode_ISCII_OFFSETS_LOGIC(UConverterFromUnicodeArgs * args,UErrorCode * err)887 static void UConverter_fromUnicode_ISCII_OFFSETS_LOGIC(
888 UConverterFromUnicodeArgs * args, UErrorCode * err) {
889 const UChar *source = args->source;
890 const UChar *sourceLimit = args->sourceLimit;
891 unsigned char *target = (unsigned char *) args->target;
892 unsigned char *targetLimit = (unsigned char *) args->targetLimit;
893 int32_t* offsets = args->offsets;
894 uint32_t targetByteUnit = 0x0000;
895 UChar32 sourceChar = 0x0000;
896 UChar32 tempContextFromUnicode = 0x0000; /* For special handling of the Gurmukhi script. */
897 UConverterDataISCII *converterData;
898 uint16_t newDelta=0;
899 uint16_t range = 0;
900 UBool deltaChanged = FALSE;
901
902 if ((args->converter == NULL) || (args->targetLimit < args->target) || (args->sourceLimit < args->source)) {
903 *err = U_ILLEGAL_ARGUMENT_ERROR;
904 return;
905 }
906 /* initialize data */
907 converterData=(UConverterDataISCII*)args->converter->extraInfo;
908 newDelta=converterData->currentDeltaFromUnicode;
909 range = (uint16_t)(newDelta/DELTA);
910
911 if ((sourceChar = args->converter->fromUChar32)!=0) {
912 goto getTrail;
913 }
914
915 /*writing the char to the output stream */
916 while (source < sourceLimit) {
917 /* Write the language code following LF only if LF is not the last character. */
918 if (args->converter->fromUnicodeStatus == LF) {
919 targetByteUnit = ATR<<8;
920 targetByteUnit += (uint8_t) lookupInitialData[range].isciiLang;
921 args->converter->fromUnicodeStatus = 0x0000;
922 /* now append ATR and language code */
923 WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,targetByteUnit,err);
924 if (U_FAILURE(*err)) {
925 break;
926 }
927 }
928
929 sourceChar = *source++;
930 tempContextFromUnicode = converterData->contextCharFromUnicode;
931
932 targetByteUnit = missingCharMarker;
933
934 /*check if input is in ASCII and C0 control codes range*/
935 if (sourceChar <= ASCII_END) {
936 args->converter->fromUnicodeStatus = sourceChar;
937 WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,sourceChar,err);
938 if (U_FAILURE(*err)) {
939 break;
940 }
941 continue;
942 }
943 switch (sourceChar) {
944 case ZWNJ:
945 /* contextChar has HALANT */
946 if (converterData->contextCharFromUnicode) {
947 converterData->contextCharFromUnicode = 0x00;
948 targetByteUnit = ISCII_HALANT;
949 } else {
950 /* consume ZWNJ and continue */
951 converterData->contextCharFromUnicode = 0x00;
952 continue;
953 }
954 break;
955 case ZWJ:
956 /* contextChar has HALANT */
957 if (converterData->contextCharFromUnicode) {
958 targetByteUnit = ISCII_NUKTA;
959 } else {
960 targetByteUnit =ISCII_INV;
961 }
962 converterData->contextCharFromUnicode = 0x00;
963 break;
964 default:
965 /* is the sourceChar in the INDIC_RANGE? */
966 if ((uint16_t)(INDIC_BLOCK_END-sourceChar) <= INDIC_RANGE) {
967 /* Danda and Double Danda are valid in Northern scripts.. since Unicode
968 * does not include these codepoints in all Northern scrips we need to
969 * filter them out
970 */
971 if (sourceChar!= DANDA && sourceChar != DOUBLE_DANDA) {
972 /* find out to which block the souceChar belongs*/
973 range =(uint16_t)((sourceChar-INDIC_BLOCK_BEGIN)/DELTA);
974 newDelta =(uint16_t)(range*DELTA);
975
976 /* Now are we in the same block as the previous? */
977 if (newDelta!= converterData->currentDeltaFromUnicode || converterData->isFirstBuffer) {
978 converterData->currentDeltaFromUnicode = newDelta;
979 converterData->currentMaskFromUnicode = lookupInitialData[range].maskEnum;
980 deltaChanged =TRUE;
981 converterData->isFirstBuffer=FALSE;
982 }
983
984 if (converterData->currentDeltaFromUnicode == PNJ_DELTA) {
985 if (sourceChar == PNJ_TIPPI) {
986 /* Make sure Tippi is converterd to Bindi. */
987 sourceChar = PNJ_BINDI;
988 } else if (sourceChar == PNJ_ADHAK) {
989 /* This is for consonant cluster handling. */
990 converterData->contextCharFromUnicode = PNJ_ADHAK;
991 }
992
993 }
994 /* Normalize all Indic codepoints to Devanagari and map them to ISCII */
995 /* now subtract the new delta from sourceChar*/
996 sourceChar -= converterData->currentDeltaFromUnicode;
997 }
998
999 /* get the target byte unit */
1000 targetByteUnit=fromUnicodeTable[(uint8_t)sourceChar];
1001
1002 /* is the code point valid in current script? */
1003 if ((validityTable[(uint8_t)sourceChar] & converterData->currentMaskFromUnicode)==0) {
1004 /* Vocallic RR is assigned in ISCII Telugu and Unicode */
1005 if (converterData->currentDeltaFromUnicode!=(TELUGU_DELTA) || sourceChar!=VOCALLIC_RR) {
1006 targetByteUnit=missingCharMarker;
1007 }
1008 }
1009
1010 if (deltaChanged) {
1011 /* we are in a script block which is different than
1012 * previous sourceChar's script block write ATR and language codes
1013 */
1014 uint32_t temp=0;
1015 temp =(uint16_t)(ATR<<8);
1016 temp += (uint16_t)((uint8_t) lookupInitialData[range].isciiLang);
1017 /* reset */
1018 deltaChanged=FALSE;
1019 /* now append ATR and language code */
1020 WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,temp,err);
1021 if (U_FAILURE(*err)) {
1022 break;
1023 }
1024 }
1025
1026 if (converterData->currentDeltaFromUnicode == PNJ_DELTA && (sourceChar + PNJ_DELTA) == PNJ_ADHAK) {
1027 continue;
1028 }
1029 }
1030 /* reset context char */
1031 converterData->contextCharFromUnicode = 0x00;
1032 break;
1033 }
1034 if (converterData->currentDeltaFromUnicode == PNJ_DELTA && tempContextFromUnicode == PNJ_ADHAK && uset_contains(PNJ_CONSONANT_SET, (sourceChar + PNJ_DELTA))) {
1035 /* If the previous codepoint is Adhak and the current codepoint is a consonant, the targetByteUnit should be C + Halant + C. */
1036 /* reset context char */
1037 converterData->contextCharFromUnicode = 0x0000;
1038 targetByteUnit = targetByteUnit << 16 | ISCII_HALANT << 8 | targetByteUnit;
1039 /* write targetByteUnit to target */
1040 WRITE_TO_TARGET_FROM_U(args, offsets, source, target, targetLimit, targetByteUnit,err);
1041 if (U_FAILURE(*err)) {
1042 break;
1043 }
1044 } else if (targetByteUnit != missingCharMarker) {
1045 if (targetByteUnit==ISCII_HALANT) {
1046 converterData->contextCharFromUnicode = (UChar)targetByteUnit;
1047 }
1048 /* write targetByteUnit to target*/
1049 WRITE_TO_TARGET_FROM_U(args,offsets,source,target,targetLimit,targetByteUnit,err);
1050 if (U_FAILURE(*err)) {
1051 break;
1052 }
1053 } else {
1054 /* oops.. the code point is unassigned */
1055 /*check if the char is a First surrogate*/
1056 if (UTF_IS_SURROGATE(sourceChar)) {
1057 if (UTF_IS_SURROGATE_FIRST(sourceChar)) {
1058 getTrail:
1059 /*look ahead to find the trail surrogate*/
1060 if (source < sourceLimit) {
1061 /* test the following code unit */
1062 UChar trail= (*source);
1063 if (UTF_IS_SECOND_SURROGATE(trail)) {
1064 source++;
1065 sourceChar=UTF16_GET_PAIR_VALUE(sourceChar, trail);
1066 *err =U_INVALID_CHAR_FOUND;
1067 /* convert this surrogate code point */
1068 /* exit this condition tree */
1069 } else {
1070 /* this is an unmatched lead code unit (1st surrogate) */
1071 /* callback(illegal) */
1072 *err=U_ILLEGAL_CHAR_FOUND;
1073 }
1074 } else {
1075 /* no more input */
1076 *err = U_ZERO_ERROR;
1077 }
1078 } else {
1079 /* this is an unmatched trail code unit (2nd surrogate) */
1080 /* callback(illegal) */
1081 *err=U_ILLEGAL_CHAR_FOUND;
1082 }
1083 } else {
1084 /* callback(unassigned) for a BMP code point */
1085 *err = U_INVALID_CHAR_FOUND;
1086 }
1087
1088 args->converter->fromUChar32=sourceChar;
1089 break;
1090 }
1091 }/* end while(mySourceIndex<mySourceLength) */
1092
1093 /*save the state and return */
1094 args->source = source;
1095 args->target = (char*)target;
1096 }
1097
1098 static const uint16_t lookupTable[][2]={
1099 { ZERO, ZERO }, /*DEFALT*/
1100 { ZERO, ZERO }, /*ROMAN*/
1101 { DEVANAGARI, DEV_MASK },
1102 { BENGALI, BNG_MASK },
1103 { TAMIL, TML_MASK },
1104 { TELUGU, KND_MASK },
1105 { BENGALI, BNG_MASK },
1106 { ORIYA, ORI_MASK },
1107 { KANNADA, KND_MASK },
1108 { MALAYALAM, MLM_MASK },
1109 { GUJARATI, GJR_MASK },
1110 { GURMUKHI, PNJ_MASK }
1111 };
1112
1113 #define WRITE_TO_TARGET_TO_U(args,source,target,offsets,offset,targetUniChar,delta, err){\
1114 /* add offset to current Indic Block */ \
1115 if(targetUniChar>ASCII_END && \
1116 targetUniChar != ZWJ && \
1117 targetUniChar != ZWNJ && \
1118 targetUniChar != DANDA && \
1119 targetUniChar != DOUBLE_DANDA){ \
1120 \
1121 targetUniChar+=(uint16_t)(delta); \
1122 } \
1123 /* now write the targetUniChar */ \
1124 if(target<args->targetLimit){ \
1125 *(target)++ = (UChar)targetUniChar; \
1126 if(offsets){ \
1127 *(offsets)++ = (int32_t)(offset); \
1128 } \
1129 }else{ \
1130 args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++] = \
1131 (UChar)targetUniChar; \
1132 *err = U_BUFFER_OVERFLOW_ERROR; \
1133 } \
1134 }
1135
1136 #define GET_MAPPING(sourceChar,targetUniChar,data){ \
1137 targetUniChar = toUnicodeTable[(sourceChar)] ; \
1138 /* is the code point valid in current script? */ \
1139 if(sourceChar> ASCII_END && \
1140 (validityTable[(uint8_t)targetUniChar] & data->currentMaskToUnicode)==0){ \
1141 /* Vocallic RR is assigne in ISCII Telugu and Unicode */ \
1142 if(data->currentDeltaToUnicode!=(TELUGU_DELTA) || \
1143 targetUniChar!=VOCALLIC_RR){ \
1144 targetUniChar=missingCharMarker; \
1145 } \
1146 } \
1147 }
1148
1149 /***********
1150 * Rules for ISCII to Unicode converter
1151 * ISCII is stateful encoding. To convert ISCII bytes to Unicode,
1152 * which has both precomposed and decomposed forms characters
1153 * pre-context and post-context need to be considered.
1154 *
1155 * Post context
1156 * i) ATR : Attribute code is used to declare the font and script switching.
1157 * Currently we only switch scripts and font codes consumed without generating an error
1158 * ii) EXT : Extention code is used to declare switching to Sanskrit and for obscure,
1159 * obsolete characters
1160 * Pre context
1161 * i) Halant: if preceeded by a halant then it is a explicit halant
1162 * ii) Nukta :
1163 * a) if preceeded by a halant then it is a soft halant
1164 * b) if preceeded by specific consonants and the ligatures have pre-composed
1165 * characters in Unicode then convert to pre-composed characters
1166 * iii) Danda: If Danda is preceeded by a Danda then convert to Double Danda
1167 *
1168 */
1169
UConverter_toUnicode_ISCII_OFFSETS_LOGIC(UConverterToUnicodeArgs * args,UErrorCode * err)1170 static void UConverter_toUnicode_ISCII_OFFSETS_LOGIC(UConverterToUnicodeArgs *args, UErrorCode* err) {
1171 const char *source = ( char *) args->source;
1172 UChar *target = args->target;
1173 const char *sourceLimit = args->sourceLimit;
1174 const UChar* targetLimit = args->targetLimit;
1175 uint32_t targetUniChar = 0x0000;
1176 uint8_t sourceChar = 0x0000;
1177 UConverterDataISCII* data;
1178 UChar32* toUnicodeStatus=NULL;
1179 UChar32 tempTargetUniChar = 0x0000;
1180 UChar* contextCharToUnicode= NULL;
1181 UBool found;
1182 int i;
1183 int offset = 0;
1184
1185 if ((args->converter == NULL) || (target < args->target) || (source < args->source)) {
1186 *err = U_ILLEGAL_ARGUMENT_ERROR;
1187 return;
1188 }
1189
1190 data = (UConverterDataISCII*)(args->converter->extraInfo);
1191 contextCharToUnicode = &data->contextCharToUnicode; /* contains previous ISCII codepoint visited */
1192 toUnicodeStatus = (UChar32*)&args->converter->toUnicodeStatus;/* contains the mapping to Unicode of the above codepoint*/
1193
1194 while (U_SUCCESS(*err) && source<sourceLimit) {
1195
1196 targetUniChar = missingCharMarker;
1197
1198 if (target < targetLimit) {
1199 sourceChar = (unsigned char)*(source)++;
1200
1201 /* look at the post-context preform special processing */
1202 if (*contextCharToUnicode==ATR) {
1203
1204 /* If we have ATR in *contextCharToUnicode then we need to change our
1205 * state to the Indic Script specified by sourceChar
1206 */
1207
1208 /* check if the sourceChar is supported script range*/
1209 if ((uint8_t)(PNJ-sourceChar)<=PNJ-DEV) {
1210 data->currentDeltaToUnicode = (uint16_t)(lookupTable[sourceChar & 0x0F][0] * DELTA);
1211 data->currentMaskToUnicode = (MaskEnum)lookupTable[sourceChar & 0x0F][1];
1212 } else if (sourceChar==DEF) {
1213 /* switch back to default */
1214 data->currentDeltaToUnicode = data->defDeltaToUnicode;
1215 data->currentMaskToUnicode = data->defMaskToUnicode;
1216 } else {
1217 if ((sourceChar >= 0x21 && sourceChar <= 0x3F)) {
1218 /* these are display codes consume and continue */
1219 } else {
1220 *err =U_ILLEGAL_CHAR_FOUND;
1221 /* reset */
1222 *contextCharToUnicode=NO_CHAR_MARKER;
1223 goto CALLBACK;
1224 }
1225 }
1226
1227 /* reset */
1228 *contextCharToUnicode=NO_CHAR_MARKER;
1229
1230 continue;
1231
1232 } else if (*contextCharToUnicode==EXT) {
1233 /* check if sourceChar is in 0xA1-0xEE range */
1234 if ((uint8_t) (EXT_RANGE_END - sourceChar) <= (EXT_RANGE_END - EXT_RANGE_BEGIN)) {
1235 /* We currently support only Anudatta and Devanagari abbreviation sign */
1236 if (sourceChar==0xBF || sourceChar == 0xB8) {
1237 targetUniChar = (sourceChar==0xBF) ? DEV_ABBR_SIGN : DEV_ANUDATTA;
1238
1239 /* find out if the mapping is valid in this state */
1240 if (validityTable[(uint8_t)targetUniChar] & data->currentMaskToUnicode) {
1241 *contextCharToUnicode= NO_CHAR_MARKER;
1242
1243 /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
1244 if (data->prevToUnicodeStatus) {
1245 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
1246 data->prevToUnicodeStatus = 0x0000;
1247 }
1248 /* write to target */
1249 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),targetUniChar,data->currentDeltaToUnicode,err);
1250
1251 continue;
1252 }
1253 }
1254 /* byte unit is unassigned */
1255 targetUniChar = missingCharMarker;
1256 *err= U_INVALID_CHAR_FOUND;
1257 } else {
1258 /* only 0xA1 - 0xEE are legal after EXT char */
1259 *contextCharToUnicode= NO_CHAR_MARKER;
1260 *err = U_ILLEGAL_CHAR_FOUND;
1261 }
1262 goto CALLBACK;
1263 } else if (*contextCharToUnicode==ISCII_INV) {
1264 if (sourceChar==ISCII_HALANT) {
1265 targetUniChar = 0x0020; /* replace with space accoding to Indic FAQ */
1266 } else {
1267 targetUniChar = ZWJ;
1268 }
1269
1270 /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
1271 if (data->prevToUnicodeStatus) {
1272 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
1273 data->prevToUnicodeStatus = 0x0000;
1274 }
1275 /* write to target */
1276 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),targetUniChar,data->currentDeltaToUnicode,err);
1277 /* reset */
1278 *contextCharToUnicode=NO_CHAR_MARKER;
1279 }
1280
1281 /* look at the pre-context and perform special processing */
1282 switch (sourceChar) {
1283 case ISCII_INV:
1284 case EXT: /*falls through*/
1285 case ATR:
1286 *contextCharToUnicode = (UChar)sourceChar;
1287
1288 if (*toUnicodeStatus != missingCharMarker) {
1289 /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
1290 if (data->prevToUnicodeStatus) {
1291 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
1292 data->prevToUnicodeStatus = 0x0000;
1293 }
1294 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),*toUnicodeStatus,data->currentDeltaToUnicode,err);
1295 *toUnicodeStatus = missingCharMarker;
1296 }
1297 continue;
1298 case ISCII_DANDA:
1299 /* handle double danda*/
1300 if (*contextCharToUnicode== ISCII_DANDA) {
1301 targetUniChar = DOUBLE_DANDA;
1302 /* clear the context */
1303 *contextCharToUnicode = NO_CHAR_MARKER;
1304 *toUnicodeStatus = missingCharMarker;
1305 } else {
1306 GET_MAPPING(sourceChar,targetUniChar,data);
1307 *contextCharToUnicode = sourceChar;
1308 }
1309 break;
1310 case ISCII_HALANT:
1311 /* handle explicit halant */
1312 if (*contextCharToUnicode == ISCII_HALANT) {
1313 targetUniChar = ZWNJ;
1314 /* clear the context */
1315 *contextCharToUnicode = NO_CHAR_MARKER;
1316 } else {
1317 GET_MAPPING(sourceChar,targetUniChar,data);
1318 *contextCharToUnicode = sourceChar;
1319 }
1320 break;
1321 case 0x0A:
1322 /* fall through */
1323 case 0x0D:
1324 data->resetToDefaultToUnicode = TRUE;
1325 GET_MAPPING(sourceChar,targetUniChar,data)
1326 ;
1327 *contextCharToUnicode = sourceChar;
1328 break;
1329
1330 case ISCII_VOWEL_SIGN_E:
1331 i=1;
1332 found=FALSE;
1333 for (; i<vowelSignESpecialCases[0][0]; i++) {
1334 if (vowelSignESpecialCases[i][0]==(uint8_t)*contextCharToUnicode) {
1335 targetUniChar=vowelSignESpecialCases[i][1];
1336 found=TRUE;
1337 break;
1338 }
1339 }
1340 if (found) {
1341 /* find out if the mapping is valid in this state */
1342 if (validityTable[(uint8_t)targetUniChar] & data->currentMaskToUnicode) {
1343 /*targetUniChar += data->currentDeltaToUnicode ;*/
1344 *contextCharToUnicode= NO_CHAR_MARKER;
1345 *toUnicodeStatus = missingCharMarker;
1346 break;
1347 }
1348 }
1349 GET_MAPPING(sourceChar,targetUniChar,data);
1350 *contextCharToUnicode = sourceChar;
1351 break;
1352
1353 case ISCII_NUKTA:
1354 /* handle soft halant */
1355 if (*contextCharToUnicode == ISCII_HALANT) {
1356 targetUniChar = ZWJ;
1357 /* clear the context */
1358 *contextCharToUnicode = NO_CHAR_MARKER;
1359 break;
1360 } else if (data->currentDeltaToUnicode == PNJ_DELTA && data->contextCharToUnicode == 0xc0) {
1361 /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
1362 if (data->prevToUnicodeStatus) {
1363 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
1364 data->prevToUnicodeStatus = 0x0000;
1365 }
1366 /* We got here because ISCII_NUKTA was preceded by 0xc0 and we are converting Gurmukhi.
1367 * In that case we must convert (0xc0 0xe9) to (\u0a5c\u0a4d\u0a39).
1368 */
1369 targetUniChar = PNJ_RRA;
1370 WRITE_TO_TARGET_TO_U(args, source, target, args->offsets, (source-args->source)-2, targetUniChar, 0, err);
1371 if (U_SUCCESS(*err)) {
1372 targetUniChar = PNJ_SIGN_VIRAMA;
1373 WRITE_TO_TARGET_TO_U(args, source, target, args->offsets, (source-args->source)-2, targetUniChar, 0, err);
1374 if (U_SUCCESS(*err)) {
1375 targetUniChar = PNJ_HA;
1376 WRITE_TO_TARGET_TO_U(args, source, target, args->offsets, (source-args->source)-2, targetUniChar, 0, err);
1377 } else {
1378 args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]= PNJ_HA;
1379 }
1380 } else {
1381 args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]= PNJ_SIGN_VIRAMA;
1382 args->converter->UCharErrorBuffer[args->converter->UCharErrorBufferLength++]= PNJ_HA;
1383 }
1384 *toUnicodeStatus = missingCharMarker;
1385 data->contextCharToUnicode = NO_CHAR_MARKER;
1386 continue;
1387 } else {
1388 /* try to handle <CHAR> + ISCII_NUKTA special mappings */
1389 i=1;
1390 found =FALSE;
1391 for (; i<nuktaSpecialCases[0][0]; i++) {
1392 if (nuktaSpecialCases[i][0]==(uint8_t)
1393 *contextCharToUnicode) {
1394 targetUniChar=nuktaSpecialCases[i][1];
1395 found =TRUE;
1396 break;
1397 }
1398 }
1399 if (found) {
1400 /* find out if the mapping is valid in this state */
1401 if (validityTable[(uint8_t)targetUniChar] & data->currentMaskToUnicode) {
1402 /*targetUniChar += data->currentDeltaToUnicode ;*/
1403 *contextCharToUnicode= NO_CHAR_MARKER;
1404 *toUnicodeStatus = missingCharMarker;
1405 if (data->currentDeltaToUnicode == PNJ_DELTA) {
1406 /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
1407 if (data->prevToUnicodeStatus) {
1408 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
1409 data->prevToUnicodeStatus = 0x0000;
1410 }
1411 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),targetUniChar,data->currentDeltaToUnicode,err);
1412 continue;
1413 }
1414 break;
1415 }
1416 /* else fall through to default */
1417 }
1418 /* else fall through to default */
1419 }
1420 default:GET_MAPPING(sourceChar,targetUniChar,data)
1421 ;
1422 *contextCharToUnicode = sourceChar;
1423 break;
1424 }
1425
1426 if (*toUnicodeStatus != missingCharMarker) {
1427 /* Check to make sure that consonant clusters are handled correct for Gurmukhi script. */
1428 if (data->currentDeltaToUnicode == PNJ_DELTA && data->prevToUnicodeStatus != 0 && uset_contains(PNJ_CONSONANT_SET, data->prevToUnicodeStatus) &&
1429 (*toUnicodeStatus + PNJ_DELTA) == PNJ_SIGN_VIRAMA && (targetUniChar + PNJ_DELTA) == data->prevToUnicodeStatus) {
1430 /* Consonant clusters C + HALANT + C should be encoded as ADHAK + C */
1431 offset = (int)(source-args->source - 3);
1432 tempTargetUniChar = PNJ_ADHAK; /* This is necessary to avoid some compiler warnings. */
1433 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,offset,tempTargetUniChar,0,err);
1434 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,offset,data->prevToUnicodeStatus,0,err);
1435 data->prevToUnicodeStatus = 0x0000; /* reset the previous unicode code point */
1436 *toUnicodeStatus = missingCharMarker;
1437 continue;
1438 } else {
1439 /* Write the previous toUnicodeStatus, this was delayed to handle consonant clustering for Gurmukhi script. */
1440 if (data->prevToUnicodeStatus) {
1441 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -1),data->prevToUnicodeStatus,0,err);
1442 data->prevToUnicodeStatus = 0x0000;
1443 }
1444 /* Check to make sure that Bindi and Tippi are handled correctly for Gurmukhi script.
1445 * If 0xA2 is preceded by a codepoint in the PNJ_BINDI_TIPPI_SET then the target codepoint should be Tippi instead of Bindi.
1446 */
1447 if (data->currentDeltaToUnicode == PNJ_DELTA && (targetUniChar + PNJ_DELTA) == PNJ_BINDI && uset_contains(PNJ_BINDI_TIPPI_SET, (*toUnicodeStatus + PNJ_DELTA))) {
1448 targetUniChar = PNJ_TIPPI - PNJ_DELTA;
1449 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),*toUnicodeStatus,PNJ_DELTA,err);
1450 } else if (data->currentDeltaToUnicode == PNJ_DELTA && (targetUniChar + PNJ_DELTA) == PNJ_SIGN_VIRAMA && uset_contains(PNJ_CONSONANT_SET, (*toUnicodeStatus + PNJ_DELTA))) {
1451 /* Store the current toUnicodeStatus code point for later handling of consonant cluster in Gurmukhi. */
1452 data->prevToUnicodeStatus = *toUnicodeStatus + PNJ_DELTA;
1453 } else {
1454 /* write the previously mapped codepoint */
1455 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source-args->source -2),*toUnicodeStatus,data->currentDeltaToUnicode,err);
1456 }
1457 }
1458 *toUnicodeStatus = missingCharMarker;
1459 }
1460
1461 if (targetUniChar != missingCharMarker) {
1462 /* now save the targetUniChar for delayed write */
1463 *toUnicodeStatus = (UChar) targetUniChar;
1464 if (data->resetToDefaultToUnicode==TRUE) {
1465 data->currentDeltaToUnicode = data->defDeltaToUnicode;
1466 data->currentMaskToUnicode = data->defMaskToUnicode;
1467 data->resetToDefaultToUnicode=FALSE;
1468 }
1469 } else {
1470
1471 /* we reach here only if targetUniChar == missingCharMarker
1472 * so assign codes to reason and err
1473 */
1474 *err = U_INVALID_CHAR_FOUND;
1475 CALLBACK:
1476 args->converter->toUBytes[0] = (uint8_t) sourceChar;
1477 args->converter->toULength = 1;
1478 break;
1479 }
1480
1481 } else {
1482 *err =U_BUFFER_OVERFLOW_ERROR;
1483 break;
1484 }
1485 }
1486
1487 if (U_SUCCESS(*err) && args->flush && source == sourceLimit) {
1488 /* end of the input stream */
1489 UConverter *cnv = args->converter;
1490
1491 if (*contextCharToUnicode==ATR || *contextCharToUnicode==EXT || *contextCharToUnicode==ISCII_INV) {
1492 /* set toUBytes[] */
1493 cnv->toUBytes[0] = (uint8_t)*contextCharToUnicode;
1494 cnv->toULength = 1;
1495
1496 /* avoid looping on truncated sequences */
1497 *contextCharToUnicode = NO_CHAR_MARKER;
1498 } else {
1499 cnv->toULength = 0;
1500 }
1501
1502 if (*toUnicodeStatus != missingCharMarker) {
1503 /* output a remaining target character */
1504 WRITE_TO_TARGET_TO_U(args,source,target,args->offsets,(source - args->source -1),*toUnicodeStatus,data->currentDeltaToUnicode,err);
1505 *toUnicodeStatus = missingCharMarker;
1506 }
1507 }
1508
1509 args->target = target;
1510 args->source = source;
1511 }
1512
1513 /* structure for SafeClone calculations */
1514 struct cloneISCIIStruct {
1515 UConverter cnv;
1516 UConverterDataISCII mydata;
1517 };
1518
1519 static UConverter *
_ISCII_SafeClone(const UConverter * cnv,void * stackBuffer,int32_t * pBufferSize,UErrorCode * status)1520 _ISCII_SafeClone(const UConverter *cnv,
1521 void *stackBuffer,
1522 int32_t *pBufferSize,
1523 UErrorCode *status)
1524 {
1525 struct cloneISCIIStruct * localClone;
1526 int32_t bufferSizeNeeded = sizeof(struct cloneISCIIStruct);
1527
1528 if (U_FAILURE(*status)) {
1529 return 0;
1530 }
1531
1532 if (*pBufferSize == 0) { /* 'preflighting' request - set needed size into *pBufferSize */
1533 *pBufferSize = bufferSizeNeeded;
1534 return 0;
1535 }
1536
1537 localClone = (struct cloneISCIIStruct *)stackBuffer;
1538 /* ucnv.c/ucnv_safeClone() copied the main UConverter already */
1539
1540 uprv_memcpy(&localClone->mydata, cnv->extraInfo, sizeof(UConverterDataISCII));
1541 localClone->cnv.extraInfo = &localClone->mydata;
1542 localClone->cnv.isExtraLocal = TRUE;
1543
1544 return &localClone->cnv;
1545 }
1546
1547 static void
_ISCIIGetUnicodeSet(const UConverter * cnv,const USetAdder * sa,UConverterUnicodeSet which,UErrorCode * pErrorCode)1548 _ISCIIGetUnicodeSet(const UConverter *cnv,
1549 const USetAdder *sa,
1550 UConverterUnicodeSet which,
1551 UErrorCode *pErrorCode)
1552 {
1553 int32_t idx, script;
1554 uint8_t mask;
1555
1556 /* Since all ISCII versions allow switching to other ISCII
1557 scripts, we add all roundtrippable characters to this set. */
1558 sa->addRange(sa->set, 0, ASCII_END);
1559 for (script = DEVANAGARI; script <= MALAYALAM; script++) {
1560 mask = (uint8_t)(lookupInitialData[script].maskEnum);
1561 for (idx = 0; idx < DELTA; idx++) {
1562 /* added check for TELUGU character */
1563 if ((validityTable[idx] & mask) || (script==TELUGU && idx==0x31)) {
1564 sa->add(sa->set, idx + (script * DELTA) + INDIC_BLOCK_BEGIN);
1565 }
1566 }
1567 }
1568 sa->add(sa->set, DANDA);
1569 sa->add(sa->set, DOUBLE_DANDA);
1570 sa->add(sa->set, ZWNJ);
1571 sa->add(sa->set, ZWJ);
1572 }
1573
1574 static const UConverterImpl _ISCIIImpl={
1575
1576 UCNV_ISCII,
1577
1578 NULL,
1579 NULL,
1580
1581 _ISCIIOpen,
1582 _ISCIIClose,
1583 _ISCIIReset,
1584
1585 UConverter_toUnicode_ISCII_OFFSETS_LOGIC,
1586 UConverter_toUnicode_ISCII_OFFSETS_LOGIC,
1587 UConverter_fromUnicode_ISCII_OFFSETS_LOGIC,
1588 UConverter_fromUnicode_ISCII_OFFSETS_LOGIC,
1589 NULL,
1590
1591 NULL,
1592 _ISCIIgetName,
1593 NULL,
1594 _ISCII_SafeClone,
1595 _ISCIIGetUnicodeSet
1596 };
1597
1598 static const UConverterStaticData _ISCIIStaticData={
1599 sizeof(UConverterStaticData),
1600 "ISCII",
1601 0,
1602 UCNV_IBM,
1603 UCNV_ISCII,
1604 1,
1605 4,
1606 { 0x1a, 0, 0, 0 },
1607 0x1,
1608 FALSE,
1609 FALSE,
1610 0x0,
1611 0x0,
1612 { 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0 }, /* reserved */
1613
1614 };
1615
1616 const UConverterSharedData _ISCIIData={
1617 sizeof(UConverterSharedData),
1618 ~((uint32_t) 0),
1619 NULL,
1620 NULL,
1621 &_ISCIIStaticData,
1622 FALSE,
1623 &_ISCIIImpl,
1624 0
1625 };
1626
1627 #endif /* #if !UCONFIG_NO_LEGACY_CONVERSION */
1628