The Tor Browser / file comparison
comparison: intl/icu/source/common/uresdata.h
intl/icu/source/common/uresdata.h
- changeset 0
- 6474c204b198
equal
deleted
inserted
replaced
| -1:000000000000 | 0:438cb6f9cccd |
|---|---|
| 1 /* | |
| 2 ****************************************************************************** | |
| 3 * * | |
| 4 * Copyright (C) 1999-2012, International Business Machines * | |
| 5 * Corporation and others. All Rights Reserved. * | |
| 6 * * | |
| 7 ****************************************************************************** | |
| 8 * file name: uresdata.h | |
| 9 * encoding: US-ASCII | |
| 10 * tab size: 8 (not used) | |
| 11 * indentation:4 | |
| 12 * | |
| 13 * created on: 1999dec08 | |
| 14 * created by: Markus W. Scherer | |
| 15 * 06/24/02 weiv Added support for resource sharing | |
| 16 */ | |
| 17 | |
| 18 #ifndef __RESDATA_H__ | |
| 19 #define __RESDATA_H__ | |
| 20 | |
| 21 #include "unicode/utypes.h" | |
| 22 #include "unicode/udata.h" | |
| 23 #include "unicode/ures.h" | |
| 24 #include "putilimp.h" | |
| 25 #include "udataswp.h" | |
| 26 | |
| 27 /** | |
| 28 * Numeric constants for internal-only types of resource items. | |
| 29 * These must use different numeric values than UResType constants | |
| 30 * because they are used together. | |
| 31 * Internal types are never returned by ures_getType(). | |
| 32 */ | |
| 33 typedef enum { | |
| 34 /** Include a negative value so that the compiler uses the same int type as for UResType. */ | |
| 35 URES_INTERNAL_NONE=-1, | |
| 36 | |
| 37 /** Resource type constant for tables with 32-bit count, key offsets and values. */ | |
| 38 URES_TABLE32=4, | |
| 39 | |
| 40 /** | |
| 41 * Resource type constant for tables with 16-bit count, key offsets and values. | |
| 42 * All values are URES_STRING_V2 strings. | |
| 43 */ | |
| 44 URES_TABLE16=5, | |
| 45 | |
| 46 /** Resource type constant for 16-bit Unicode strings in formatVersion 2. */ | |
| 47 URES_STRING_V2=6, | |
| 48 | |
| 49 /** | |
| 50 * Resource type constant for arrays with 16-bit count and values. | |
| 51 * All values are URES_STRING_V2 strings. | |
| 52 */ | |
| 53 URES_ARRAY16=9 | |
| 54 } UResInternalType; | |
| 55 | |
| 56 /* | |
| 57 * A Resource is a 32-bit value that has 2 bit fields: | |
| 58 * 31..28 4-bit type, see enum below | |
| 59 * 27..0 28-bit four-byte-offset or value according to the type | |
| 60 */ | |
| 61 typedef uint32_t Resource; | |
| 62 | |
| 63 #define RES_BOGUS 0xffffffff | |
| 64 | |
| 65 #define RES_GET_TYPE(res) ((int32_t)((res)>>28UL)) | |
| 66 #define RES_GET_OFFSET(res) ((res)&0x0fffffff) | |
| 67 #define RES_GET_POINTER(pRoot, res) ((pRoot)+RES_GET_OFFSET(res)) | |
| 68 | |
| 69 /* get signed and unsigned integer values directly from the Resource handle */ | |
| 70 #if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC | |
| 71 # define RES_GET_INT(res) (((int32_t)((res)<<4L))>>4L) | |
| 72 #else | |
| 73 # define RES_GET_INT(res) (int32_t)(((res)&0x08000000) ? (res)|0xf0000000 : (res)&0x07ffffff) | |
| 74 #endif | |
| 75 | |
| 76 #define RES_GET_UINT(res) ((res)&0x0fffffff) | |
| 77 | |
| 78 #define URES_IS_ARRAY(type) ((int32_t)(type)==URES_ARRAY || (int32_t)(type)==URES_ARRAY16) | |
| 79 #define URES_IS_TABLE(type) ((int32_t)(type)==URES_TABLE || (int32_t)(type)==URES_TABLE16 || (int32_t)(type)==URES_TABLE32) | |
| 80 #define URES_IS_CONTAINER(type) (URES_IS_TABLE(type) || URES_IS_ARRAY(type)) | |
| 81 | |
| 82 #define URES_MAKE_RESOURCE(type, offset) (((Resource)(type)<<28)|(Resource)(offset)) | |
| 83 #define URES_MAKE_EMPTY_RESOURCE(type) ((Resource)(type)<<28) | |
| 84 | |
| 85 /* indexes[] value names; indexes are generally 32-bit (Resource) indexes */ | |
| 86 enum { | |
| 87 URES_INDEX_LENGTH, /* [0] contains URES_INDEX_TOP==the length of indexes[]; | |
| 88 * formatVersion==1: all bits contain the length of indexes[] | |
| 89 * but the length is much less than 0xff; | |
| 90 * formatVersion>1: | |
| 91 * only bits 7..0 contain the length of indexes[], | |
| 92 * bits 31..8 are reserved and set to 0 */ | |
| 93 URES_INDEX_KEYS_TOP, /* [1] contains the top of the key strings, */ | |
| 94 /* same as the bottom of resources or UTF-16 strings, rounded up */ | |
| 95 URES_INDEX_RESOURCES_TOP, /* [2] contains the top of all resources */ | |
| 96 URES_INDEX_BUNDLE_TOP, /* [3] contains the top of the bundle, */ | |
| 97 /* in case it were ever different from [2] */ | |
| 98 URES_INDEX_MAX_TABLE_LENGTH,/* [4] max. length of any table */ | |
| 99 URES_INDEX_ATTRIBUTES, /* [5] attributes bit set, see URES_ATT_* (new in formatVersion 1.2) */ | |
| 100 URES_INDEX_16BIT_TOP, /* [6] top of the 16-bit units (UTF-16 string v2 UChars, URES_TABLE16, URES_ARRAY16), | |
| 101 * rounded up (new in formatVersion 2.0, ICU 4.4) */ | |
| 102 URES_INDEX_POOL_CHECKSUM, /* [7] checksum of the pool bundle (new in formatVersion 2.0, ICU 4.4) */ | |
| 103 URES_INDEX_TOP | |
| 104 }; | |
| 105 | |
| 106 /* | |
| 107 * Nofallback attribute, attribute bit 0 in indexes[URES_INDEX_ATTRIBUTES]. | |
| 108 * New in formatVersion 1.2 (ICU 3.6). | |
| 109 * | |
| 110 * If set, then this resource bundle is a standalone bundle. | |
| 111 * If not set, then the bundle participates in locale fallback, eventually | |
| 112 * all the way to the root bundle. | |
| 113 * If indexes[] is missing or too short, then the attribute cannot be determined | |
| 114 * reliably. Dependency checking should ignore such bundles, and loading should | |
| 115 * use fallbacks. | |
| 116 */ | |
| 117 #define URES_ATT_NO_FALLBACK 1 | |
| 118 | |
| 119 /* | |
| 120 * Attributes for bundles that are, or use, a pool bundle. | |
| 121 * A pool bundle provides key strings that are shared among several other bundles | |
| 122 * to reduce their total size. | |
| 123 * New in formatVersion 2 (ICU 4.4). | |
| 124 */ | |
| 125 #define URES_ATT_IS_POOL_BUNDLE 2 | |
| 126 #define URES_ATT_USES_POOL_BUNDLE 4 | |
| 127 | |
| 128 /* | |
| 129 * File format for .res resource bundle files (formatVersion=2, ICU 4.4) | |
| 130 * | |
| 131 * New in formatVersion 2 compared with 1.3: ------------- | |
| 132 * | |
| 133 * Three new resource types -- String-v2, Table16 and Array16 -- have their | |
| 134 * values stored in a new array of 16-bit units between the table key strings | |
| 135 * and the start of the other resources. | |
| 136 * | |
| 137 * genrb eliminates duplicates among Unicode string-v2 values. | |
| 138 * Multiple Unicode strings may use the same offset and string data, | |
| 139 * or a short string may point to the suffix of a longer string. ("Suffix sharing") | |
| 140 * For example, one string "abc" may be reused for another string "bc" by pointing | |
| 141 * to the second character. (Short strings-v2 are NUL-terminated | |
| 142 * and not preceded by an explicit length value.) | |
| 143 * | |
| 144 * It is allowed for all resource types to share values. | |
| 145 * The swapper code (ures_swap()) has been modified so that it swaps each item | |
| 146 * exactly once. | |
| 147 * | |
| 148 * A resource bundle may use a special pool bundle. Some or all of the table key strings | |
| 149 * of the using-bundle are omitted, and the key string offsets for such key strings refer | |
| 150 * to offsets in the pool bundle. | |
| 151 * The using-bundle's and the pool-bundle's indexes[URES_INDEX_POOL_CHECKSUM] values | |
| 152 * must match. | |
| 153 * Two bits in indexes[URES_INDEX_ATTRIBUTES] indicate whether a resource bundle | |
| 154 * is or uses a pool bundle. | |
| 155 * | |
| 156 * Table key strings must be compared in ASCII order, even if they are not | |
| 157 * stored in ASCII. | |
| 158 * | |
| 159 * New in formatVersion 1.3 compared with 1.2: ------------- | |
| 160 * | |
| 161 * genrb eliminates duplicates among key strings. | |
| 162 * Multiple table items may share one key string, or one item may point | |
| 163 * to the suffix of another's key string. ("Suffix sharing") | |
| 164 * For example, one key "abc" may be reused for another key "bc" by pointing | |
| 165 * to the second character. (Key strings are NUL-terminated.) | |
| 166 * | |
| 167 * ------------- | |
| 168 * | |
| 169 * An ICU4C resource bundle file (.res) is a binary, memory-mappable file | |
| 170 * with nested, hierarchical data structures. | |
| 171 * It physically contains the following: | |
| 172 * | |
| 173 * Resource root; -- 32-bit Resource item, root item for this bundle's tree; | |
| 174 * currently, the root item must be a table or table32 resource item | |
| 175 * int32_t indexes[indexes[0]]; -- array of indexes for friendly | |
| 176 * reading and swapping; see URES_INDEX_* above | |
| 177 * new in formatVersion 1.1 (ICU 2.8) | |
| 178 * char keys[]; -- characters for key strings | |
| 179 * (formatVersion 1.0: up to 65k of characters; 1.1: <2G) | |
| 180 * (minus the space for root and indexes[]), | |
| 181 * which consist of invariant characters (ASCII/EBCDIC) and are NUL-terminated; | |
| 182 * padded to multiple of 4 bytes for 4-alignment of the following data | |
| 183 * uint16_t 16BitUnits[]; -- resources that are stored entirely as sequences of 16-bit units | |
| 184 * (new in formatVersion 2/ICU 4.4) | |
| 185 * data is indexed by the offset values in 16-bit resource types, | |
| 186 * with offset 0 pointing to the beginning of this array; | |
| 187 * there is a 0 at offset 0, for empty resources; | |
| 188 * padded to multiple of 4 bytes for 4-alignment of the following data | |
| 189 * data; -- data directly and indirectly indexed by the root item; | |
| 190 * the structure is determined by walking the tree | |
| 191 * | |
| 192 * Each resource bundle item has a 32-bit Resource handle (see typedef above) | |
| 193 * which contains the item type number in its upper 4 bits (31..28) and either | |
| 194 * an offset or a direct value in its lower 28 bits (27..0). | |
| 195 * The order of items is undefined and only determined by walking the tree. | |
| 196 * Leaves of the tree may be stored first or last or anywhere in between, | |
| 197 * and it is in theory possible to have unreferenced holes in the file. | |
| 198 * | |
| 199 * 16-bit-unit values: | |
| 200 * Starting with formatVersion 2/ICU 4.4, some resources are stored in a special | |
| 201 * array of 16-bit units. Each resource value is a sequence of 16-bit units, | |
| 202 * with no per-resource padding to a 4-byte boundary. | |
| 203 * 16-bit container types (Table16 and Array16) contain Resource16 values | |
| 204 * which are offsets to String-v2 resources in the same 16-bit-units array. | |
| 205 * | |
| 206 * Direct values: | |
| 207 * - Empty Unicode strings have an offset value of 0 in the Resource handle itself. | |
| 208 * - Starting with formatVersion 2/ICU 4.4, an offset value of 0 for | |
| 209 * _any_ resource type indicates an empty value. | |
| 210 * - Integer values are 28-bit values stored in the Resource handle itself; | |
| 211 * the interpretation of unsigned vs. signed integers is up to the application. | |
| 212 * | |
| 213 * All other types and values use 28-bit offsets to point to the item's data. | |
| 214 * The offset is an index to the first 32-bit word of the value, relative to the | |
| 215 * start of the resource data (i.e., the root item handle is at offset 0). | |
| 216 * To get byte offsets, the offset is multiplied by 4 (or shifted left by 2 bits). | |
| 217 * All resource item values are 4-aligned. | |
| 218 * | |
| 219 * New in formatVersion 2/ICU 4.4: Some types use offsets into the 16-bit-units array, | |
| 220 * indexing 16-bit units in that array. | |
| 221 * | |
| 222 * The structures (memory layouts) for the values for each item type are listed | |
| 223 * in the table below. | |
| 224 * | |
| 225 * Nested, hierarchical structures: ------------- | |
| 226 * | |
| 227 * Table items contain key-value pairs where the keys are offsets to char * key strings. | |
| 228 * The values of these pairs are either Resource handles or | |
| 229 * offsets into the 16-bit-units array, depending on the table type. | |
| 230 * | |
| 231 * Array items are simple vectors of Resource handles, | |
| 232 * or of offsets into the 16-bit-units array, depending on the array type. | |
| 233 * | |
| 234 * Table key string offsets: ------- | |
| 235 * | |
| 236 * Key string offsets are relative to the start of the resource data (of the root handle), | |
| 237 * i.e., the first string has an offset of 4+sizeof(indexes). | |
| 238 * (After the 4-byte root handle and after the indexes array.) | |
| 239 * | |
| 240 * If the resource bundle uses a pool bundle, then some key strings are stored | |
| 241 * in the pool bundle rather than in the local bundle itself. | |
| 242 * - In a Table or Table16, the 16-bit key string offset is local if it is | |
| 243 * less than indexes[URES_INDEX_KEYS_TOP]<<2. | |
| 244 * Otherwise, subtract indexes[URES_INDEX_KEYS_TOP]<<2 to get the offset into | |
| 245 * the pool bundle key strings. | |
| 246 * - In a Table32, the 32-bit key string offset is local if it is non-negative. | |
| 247 * Otherwise, reset bit 31 to get the pool key string offset. | |
| 248 * | |
| 249 * Unlike the local offset, the pool key offset is relative to | |
| 250 * the start of the key strings, not to the start of the bundle. | |
| 251 * | |
| 252 * An alias item is special (and new in ICU 2.4): -------------- | |
| 253 * | |
| 254 * Its memory layout is just like for a UnicodeString, but at runtime it resolves to | |
| 255 * another resource bundle's item according to the path in the string. | |
| 256 * This is used to share items across bundles that are in different lookup/fallback | |
| 257 * chains (e.g., large collation data among zh_TW and zh_HK). | |
| 258 * This saves space (for large items) and maintenance effort (less duplication of data). | |
| 259 * | |
| 260 * -------------------------------------------------------------------------- | |
| 261 * | |
| 262 * Resource types: | |
| 263 * | |
| 264 * Most resources have their values stored at four-byte offsets from the start | |
| 265 * of the resource data. These values are at least 4-aligned. | |
| 266 * Some resource values are stored directly in the offset field of the Resource itself. | |
| 267 * See UResType in unicode/ures.h for enumeration constants for Resource types. | |
| 268 * | |
| 269 * Some resources have their values stored as sequences of 16-bit units, | |
| 270 * at 2-byte offsets from the start of a contiguous 16-bit-unit array between | |
| 271 * the table key strings and the other resources. (new in formatVersion 2/ICU 4.4) | |
| 272 * At offset 0 of that array is a 16-bit zero value for empty 16-bit resources. | |
| 273 * Resource16 values in Table16 and Array16 are 16-bit offsets to String-v2 | |
| 274 * resources, with the offsets relative to the start of the 16-bit-units array. | |
| 275 * | |
| 276 * Type Name Memory layout of values | |
| 277 * (in parentheses: scalar, non-offset values) | |
| 278 * | |
| 279 * 0 Unicode String: int32_t length, UChar[length], (UChar)0, (padding) | |
| 280 * or (empty string ("") if offset==0) | |
| 281 * 1 Binary: int32_t length, uint8_t[length], (padding) | |
| 282 * - the start of the bytes is 16-aligned - | |
| 283 * 2 Table: uint16_t count, uint16_t keyStringOffsets[count], (uint16_t padding), Resource[count] | |
| 284 * 3 Alias: (physically same value layout as string, new in ICU 2.4) | |
| 285 * 4 Table32: int32_t count, int32_t keyStringOffsets[count], Resource[count] | |
| 286 * (new in formatVersion 1.1/ICU 2.8) | |
| 287 * 5 Table16: uint16_t count, uint16_t keyStringOffsets[count], Resource16[count] | |
| 288 * (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4) | |
| 289 * 6 Unicode String-v2:UChar[length], (UChar)0; length determined by the first UChar: | |
| 290 * - if first is not a trail surrogate, then the length is implicit | |
| 291 * and u_strlen() needs to be called | |
| 292 * - if first<0xdfef then length=first&0x3ff (and skip first) | |
| 293 * - if first<0xdfff then length=((first-0xdfef)<<16) | second UChar | |
| 294 * - if first==0xdfff then length=((second UChar)<<16) | third UChar | |
| 295 * (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4) | |
| 296 * 7 Integer: (28-bit offset is integer value) | |
| 297 * 8 Array: int32_t count, Resource[count] | |
| 298 * 9 Array16: uint16_t count, Resource16[count] | |
| 299 * (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4) | |
| 300 * 14 Integer Vector: int32_t length, int32_t[length] | |
| 301 * 15 Reserved: This value denotes special purpose resources and is for internal use. | |
| 302 * | |
| 303 * Note that there are 3 types with data vector values: | |
| 304 * - Vectors of 8-bit bytes stored as type Binary. | |
| 305 * - Vectors of 16-bit words stored as type Unicode String or Unicode String-v2 | |
| 306 * (no value restrictions, all values 0..ffff allowed!). | |
| 307 * - Vectors of 32-bit words stored as type Integer Vector. | |
| 308 */ | |
| 309 | |
| 310 /* | |
| 311 * Structure for a single, memory-mapped ResourceBundle. | |
| 312 */ | |
| 313 typedef struct { | |
| 314 UDataMemory *data; | |
| 315 const int32_t *pRoot; | |
| 316 const uint16_t *p16BitUnits; | |
| 317 const char *poolBundleKeys; | |
| 318 Resource rootRes; | |
| 319 int32_t localKeyLimit; | |
| 320 UBool noFallback; /* see URES_ATT_NO_FALLBACK */ | |
| 321 UBool isPoolBundle; | |
| 322 UBool usesPoolBundle; | |
| 323 UBool useNativeStrcmp; | |
| 324 } ResourceData; | |
| 325 | |
| 326 /* | |
| 327 * Read a resource bundle from memory. | |
| 328 */ | |
| 329 U_INTERNAL void U_EXPORT2 | |
| 330 res_read(ResourceData *pResData, | |
| 331 const UDataInfo *pInfo, const void *inBytes, int32_t length, | |
| 332 UErrorCode *errorCode); | |
| 333 | |
| 334 /* | |
| 335 * Load a resource bundle file. | |
| 336 * The ResourceData structure must be allocated externally. | |
| 337 */ | |
| 338 U_CFUNC void | |
| 339 res_load(ResourceData *pResData, | |
| 340 const char *path, const char *name, UErrorCode *errorCode); | |
| 341 | |
| 342 /* | |
| 343 * Release a resource bundle file. | |
| 344 * This does not release the ResourceData structure itself. | |
| 345 */ | |
| 346 U_CFUNC void | |
| 347 res_unload(ResourceData *pResData); | |
| 348 | |
| 349 U_INTERNAL UResType U_EXPORT2 | |
| 350 res_getPublicType(Resource res); | |
| 351 | |
| 352 /* | |
| 353 * Return a pointer to a zero-terminated, const UChar* string | |
| 354 * and set its length in *pLength. | |
| 355 * Returns NULL if not found. | |
| 356 */ | |
| 357 U_INTERNAL const UChar * U_EXPORT2 | |
| 358 res_getString(const ResourceData *pResData, Resource res, int32_t *pLength); | |
| 359 | |
| 360 U_INTERNAL const UChar * U_EXPORT2 | |
| 361 res_getAlias(const ResourceData *pResData, Resource res, int32_t *pLength); | |
| 362 | |
| 363 U_INTERNAL const uint8_t * U_EXPORT2 | |
| 364 res_getBinary(const ResourceData *pResData, Resource res, int32_t *pLength); | |
| 365 | |
| 366 U_INTERNAL const int32_t * U_EXPORT2 | |
| 367 res_getIntVector(const ResourceData *pResData, Resource res, int32_t *pLength); | |
| 368 | |
| 369 U_INTERNAL Resource U_EXPORT2 | |
| 370 res_getResource(const ResourceData *pResData, const char *key); | |
| 371 | |
| 372 U_INTERNAL int32_t U_EXPORT2 | |
| 373 res_countArrayItems(const ResourceData *pResData, Resource res); | |
| 374 | |
| 375 U_INTERNAL Resource U_EXPORT2 | |
| 376 res_getArrayItem(const ResourceData *pResData, Resource array, int32_t indexS); | |
| 377 | |
| 378 U_INTERNAL Resource U_EXPORT2 | |
| 379 res_getTableItemByIndex(const ResourceData *pResData, Resource table, int32_t indexS, const char ** key); | |
| 380 | |
| 381 U_INTERNAL Resource U_EXPORT2 | |
| 382 res_getTableItemByKey(const ResourceData *pResData, Resource table, int32_t *indexS, const char* * key); | |
| 383 | |
| 384 /* | |
| 385 * Modifies the contents of *path (replacing separators with NULs), | |
| 386 * and also moves *path forward while it finds items. | |
| 387 */ | |
| 388 U_CFUNC Resource res_findResource(const ResourceData *pResData, Resource r, char** path, const char** key); | |
| 389 | |
| 390 /** | |
| 391 * Swap an ICU resource bundle. See udataswp.h. | |
| 392 * @internal | |
| 393 */ | |
| 394 U_CAPI int32_t U_EXPORT2 | |
| 395 ures_swap(const UDataSwapper *ds, | |
| 396 const void *inData, int32_t length, void *outData, | |
| 397 UErrorCode *pErrorCode); | |
| 398 | |
| 399 #endif |
