X-Git-Url: https://git.saurik.com/apple/icu.git/blobdiff_plain/374ca955a76ecab1204ca8bfa63ff9238d998416..HEAD:/icuSources/common/uresdata.h diff --git a/icuSources/common/uresdata.h b/icuSources/common/uresdata.h index 92a8667d..d1b67bab 100644 --- a/icuSources/common/uresdata.h +++ b/icuSources/common/uresdata.h @@ -1,12 +1,12 @@ +// © 2016 and later: Unicode, Inc. and others. +// License & terms of use: http://www.unicode.org/copyright.html /* ****************************************************************************** -* * -* Copyright (C) 1999-2003, International Business Machines * -* Corporation and others. All Rights Reserved. * -* * +* Copyright (C) 1999-2016, International Business Machines +* Corporation and others. All Rights Reserved. ****************************************************************************** * file name: uresdata.h -* encoding: US-ASCII +* encoding: UTF-8 * tab size: 8 (not used) * indentation:4 * @@ -20,8 +20,41 @@ #include "unicode/utypes.h" #include "unicode/udata.h" +#include "unicode/ures.h" +#include "putilimp.h" #include "udataswp.h" +/** + * Numeric constants for internal-only types of resource items. + * These must use different numeric values than UResType constants + * because they are used together. + * Internal types are never returned by ures_getType(). + */ +typedef enum { + /** Include a negative value so that the compiler uses the same int type as for UResType. */ + URES_INTERNAL_NONE=-1, + + /** Resource type constant for tables with 32-bit count, key offsets and values. */ + URES_TABLE32=4, + + /** + * Resource type constant for tables with 16-bit count, key offsets and values. + * All values are URES_STRING_V2 strings. + */ + URES_TABLE16=5, + + /** Resource type constant for 16-bit Unicode strings in formatVersion 2. */ + URES_STRING_V2=6, + + /** + * Resource type constant for arrays with 16-bit count and values. + * All values are URES_STRING_V2 strings. + */ + URES_ARRAY16=9 + + /* Resource type 15 is not defined but effectively used by RES_BOGUS=0xffffffff. */ +} UResInternalType; + /* * A Resource is a 32-bit value that has 2 bit fields: * 31..28 4-bit type, see enum below @@ -30,34 +63,171 @@ typedef uint32_t Resource; #define RES_BOGUS 0xffffffff +#define RES_MAX_OFFSET 0x0fffffff -#define RES_GET_TYPE(res) ((res)>>28UL) +#define RES_GET_TYPE(res) ((int32_t)((res)>>28UL)) #define RES_GET_OFFSET(res) ((res)&0x0fffffff) #define RES_GET_POINTER(pRoot, res) ((pRoot)+RES_GET_OFFSET(res)) -/* get signed and unsigned integer values directly from the Resource handle */ -#define RES_GET_INT(res) (((int32_t)((res)<<4L))>>4L) -#define RES_GET_UINT(res) ((res)&0x0fffffff) +/* get signed and unsigned integer values directly from the Resource handle + * NOTE: For proper logging, please use the res_getInt() constexpr + */ +#if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC +# define RES_GET_INT_NO_TRACE(res) (((int32_t)((res)<<4L))>>4L) +#else +# define RES_GET_INT_NO_TRACE(res) (int32_t)(((res)&0x08000000) ? (res)|0xf0000000 : (res)&0x07ffffff) +#endif + +#define RES_GET_UINT_NO_TRACE(res) ((res)&0x0fffffff) + +#define URES_IS_ARRAY(type) ((int32_t)(type)==URES_ARRAY || (int32_t)(type)==URES_ARRAY16) +#define URES_IS_TABLE(type) ((int32_t)(type)==URES_TABLE || (int32_t)(type)==URES_TABLE16 || (int32_t)(type)==URES_TABLE32) +#define URES_IS_CONTAINER(type) (URES_IS_TABLE(type) || URES_IS_ARRAY(type)) + +#define URES_MAKE_RESOURCE(type, offset) (((Resource)(type)<<28)|(Resource)(offset)) +#define URES_MAKE_EMPTY_RESOURCE(type) ((Resource)(type)<<28) /* indexes[] value names; indexes are generally 32-bit (Resource) indexes */ enum { - URES_INDEX_LENGTH, /* [0] contains URES_INDEX_TOP==the length of indexes[] */ - URES_INDEX_STRINGS_TOP, /* [1] contains the top of the strings, */ - /* same as the bottom of resources, rounded up */ - URES_INDEX_RESOURCES_TOP, /* [2] contains the top of all resources */ - URES_INDEX_BUNDLE_TOP, /* [3] contains the top of the bundle, */ - /* in case it were ever different from [2] */ - URES_INDEX_MAX_TABLE_LENGTH,/* [4] max. length of any table */ + /** + * [0] contains the length of indexes[] + * which is at most URES_INDEX_TOP of the latest format version + * + * formatVersion==1: all bits contain the length of indexes[] + * but the length is much less than 0xff; + * formatVersion>1: + * only bits 7..0 contain the length of indexes[], + * bits 31..8 are reserved and set to 0 + * formatVersion>=3: + * bits 31..8 poolStringIndexLimit bits 23..0 + */ + URES_INDEX_LENGTH, + /** + * [1] contains the top of the key strings, + * same as the bottom of resources or UTF-16 strings, rounded up + */ + URES_INDEX_KEYS_TOP, + /** [2] contains the top of all resources */ + URES_INDEX_RESOURCES_TOP, + /** + * [3] contains the top of the bundle, + * in case it were ever different from [2] + */ + URES_INDEX_BUNDLE_TOP, + /** [4] max. length of any table */ + URES_INDEX_MAX_TABLE_LENGTH, + /** + * [5] attributes bit set, see URES_ATT_* (new in formatVersion 1.2) + * + * formatVersion>=3: + * bits 31..16 poolStringIndex16Limit + * bits 15..12 poolStringIndexLimit bits 27..24 + */ + URES_INDEX_ATTRIBUTES, + /** + * [6] top of the 16-bit units (UTF-16 string v2 UChars, URES_TABLE16, URES_ARRAY16), + * rounded up (new in formatVersion 2.0, ICU 4.4) + */ + URES_INDEX_16BIT_TOP, + /** [7] checksum of the pool bundle (new in formatVersion 2.0, ICU 4.4) */ + URES_INDEX_POOL_CHECKSUM, URES_INDEX_TOP }; -/* number of bytes at the beginning of the bundle before the strings start */ -enum { - URES_STRINGS_BOTTOM=(1+URES_INDEX_TOP)*4 -}; +/* + * Nofallback attribute, attribute bit 0 in indexes[URES_INDEX_ATTRIBUTES]. + * New in formatVersion 1.2 (ICU 3.6). + * + * If set, then this resource bundle is a standalone bundle. + * If not set, then the bundle participates in locale fallback, eventually + * all the way to the root bundle. + * If indexes[] is missing or too short, then the attribute cannot be determined + * reliably. Dependency checking should ignore such bundles, and loading should + * use fallbacks. + */ +#define URES_ATT_NO_FALLBACK 1 /* - * File format for .res resource bundle files (formatVersion=1.1) + * Attributes for bundles that are, or use, a pool bundle. + * A pool bundle provides key strings that are shared among several other bundles + * to reduce their total size. + * New in formatVersion 2 (ICU 4.4). + */ +#define URES_ATT_IS_POOL_BUNDLE 2 +#define URES_ATT_USES_POOL_BUNDLE 4 + +/* + * File format for .res resource bundle files + * + * ICU 56: New in formatVersion 3 compared with 2: ------------- + * + * Resource bundles can optionally use shared string-v2 values + * stored in the pool bundle. + * If so, then the indexes[] contain two new values + * in previously-unused bits of existing indexes[] slots: + * - poolStringIndexLimit: + * String-v2 offsets (in 32-bit Resource words) below this limit + * point to pool bundle string-v2 values. + * - poolStringIndex16Limit: + * Resource16 string-v2 offsets below this limit + * point to pool bundle string-v2 values. + * Guarantee: poolStringIndex16Limit <= poolStringIndexLimit + * + * The local bundle's poolStringIndexLimit is greater than + * any pool bundle string index used in the local bundle. + * The poolStringIndexLimit should not be greater than + * the maximum possible pool bundle string index. + * + * The maximum possible pool bundle string index is the index to the last non-NUL + * pool string character, due to suffix sharing. + * + * In the pool bundle, there is no structure that lists the strings. + * (The root resource is an empty Table.) + * If the strings need to be enumerated (as genrb --usePoolBundle does), + * then iterate through the pool bundle's 16-bit-units array from the beginning. + * Stop at the end of the array, or when an explicit or implicit string length + * would lead beyond the end of the array, + * or when an apparent string is not NUL-terminated. + * (Future genrb version might terminate the strings with + * what looks like a large explicit string length.) + * + * ICU 4.4: New in formatVersion 2 compared with 1.3: ------------- + * + * Three new resource types -- String-v2, Table16 and Array16 -- have their + * values stored in a new array of 16-bit units between the table key strings + * and the start of the other resources. + * + * genrb eliminates duplicates among Unicode string-v2 values. + * Multiple Unicode strings may use the same offset and string data, + * or a short string may point to the suffix of a longer string. ("Suffix sharing") + * For example, one string "abc" may be reused for another string "bc" by pointing + * to the second character. (Short strings-v2 are NUL-terminated + * and not preceded by an explicit length value.) + * + * It is allowed for all resource types to share values. + * The swapper code (ures_swap()) has been modified so that it swaps each item + * exactly once. + * + * A resource bundle may use a special pool bundle. Some or all of the table key strings + * of the using-bundle are omitted, and the key string offsets for such key strings refer + * to offsets in the pool bundle. + * The using-bundle's and the pool-bundle's indexes[URES_INDEX_POOL_CHECKSUM] values + * must match. + * Two bits in indexes[URES_INDEX_ATTRIBUTES] indicate whether a resource bundle + * is or uses a pool bundle. + * + * Table key strings must be compared in ASCII order, even if they are not + * stored in ASCII. + * + * New in formatVersion 1.3 compared with 1.2: ------------- + * + * genrb eliminates duplicates among key strings. + * Multiple table items may share one key string, or one item may point + * to the suffix of another's key string. ("Suffix sharing") + * For example, one key "abc" may be reused for another key "bc" by pointing + * to the second character. (Key strings are NUL-terminated.) + * + * ------------- * * An ICU4C resource bundle file (.res) is a binary, memory-mappable file * with nested, hierarchical data structures. @@ -67,12 +237,18 @@ enum { * currently, the root item must be a table or table32 resource item * int32_t indexes[indexes[0]]; -- array of indexes for friendly * reading and swapping; see URES_INDEX_* above - * new in formatVersion 1.1 + * new in formatVersion 1.1 (ICU 2.8) * char keys[]; -- characters for key strings * (formatVersion 1.0: up to 65k of characters; 1.1: <2G) * (minus the space for root and indexes[]), * which consist of invariant characters (ASCII/EBCDIC) and are NUL-terminated; * padded to multiple of 4 bytes for 4-alignment of the following data + * uint16_t 16BitUnits[]; -- resources that are stored entirely as sequences of 16-bit units + * (new in formatVersion 2/ICU 4.4) + * data is indexed by the offset values in 16-bit resource types, + * with offset 0 pointing to the beginning of this array; + * there is a 0 at offset 0, for empty resources; + * padded to multiple of 4 bytes for 4-alignment of the following data * data; -- data directly and indirectly indexed by the root item; * the structure is determined by walking the tree * @@ -83,8 +259,17 @@ enum { * Leaves of the tree may be stored first or last or anywhere in between, * and it is in theory possible to have unreferenced holes in the file. * + * 16-bit-unit values: + * Starting with formatVersion 2/ICU 4.4, some resources are stored in a special + * array of 16-bit units. Each resource value is a sequence of 16-bit units, + * with no per-resource padding to a 4-byte boundary. + * 16-bit container types (Table16 and Array16) contain Resource16 values + * which are offsets to String-v2 resources in the same 16-bit-units array. + * * Direct values: * - Empty Unicode strings have an offset value of 0 in the Resource handle itself. + * - Starting with formatVersion 2/ICU 4.4, an offset value of 0 for + * _any_ resource type indicates an empty value. * - Integer values are 28-bit values stored in the Resource handle itself; * the interpretation of unsigned vs. signed integers is up to the application. * @@ -94,18 +279,38 @@ enum { * To get byte offsets, the offset is multiplied by 4 (or shifted left by 2 bits). * All resource item values are 4-aligned. * + * New in formatVersion 2/ICU 4.4: Some types use offsets into the 16-bit-units array, + * indexing 16-bit units in that array. + * * The structures (memory layouts) for the values for each item type are listed - * in the table above. + * in the table below. * * Nested, hierarchical structures: ------------- * - * Table items contain key-value pairs where the keys are 16-bit offsets to char * key strings. - * Key string offsets are also relative to the start of the resource data (of the root handle), - * i.e., the first string has an offset of 4 (after the 4-byte root handle). + * Table items contain key-value pairs where the keys are offsets to char * key strings. + * The values of these pairs are either Resource handles or + * offsets into the 16-bit-units array, depending on the table type. + * + * Array items are simple vectors of Resource handles, + * or of offsets into the 16-bit-units array, depending on the array type. + * + * Table key string offsets: ------- * - * The values of these pairs are Resource handles. + * Key string offsets are relative to the start of the resource data (of the root handle), + * i.e., the first string has an offset of 4+sizeof(indexes). + * (After the 4-byte root handle and after the indexes array.) * - * Array items are simple vectors of Resource handles. + * If the resource bundle uses a pool bundle, then some key strings are stored + * in the pool bundle rather than in the local bundle itself. + * - In a Table or Table16, the 16-bit key string offset is local if it is + * less than indexes[URES_INDEX_KEYS_TOP]<<2. + * Otherwise, subtract indexes[URES_INDEX_KEYS_TOP]<<2 to get the offset into + * the pool bundle key strings. + * - In a Table32, the 32-bit key string offset is local if it is non-negative. + * Otherwise, reset bit 31 to get the pool key string offset. + * + * Unlike the local offset, the pool key offset is relative to + * the start of the key strings, not to the start of the bundle. * * An alias item is special (and new in ICU 2.4): -------------- * @@ -124,27 +329,53 @@ enum { * Some resource values are stored directly in the offset field of the Resource itself. * See UResType in unicode/ures.h for enumeration constants for Resource types. * + * Some resources have their values stored as sequences of 16-bit units, + * at 2-byte offsets from the start of a contiguous 16-bit-unit array between + * the table key strings and the other resources. (new in formatVersion 2/ICU 4.4) + * At offset 0 of that array is a 16-bit zero value for empty 16-bit resources. + * + * Resource16 values in Table16 and Array16 are 16-bit offsets to String-v2 + * resources, with the offsets relative to the start of the 16-bit-units array. + * Starting with formatVersion 3/ICU 56, if offset(URES_NONE)), + fTraceInfo() {} + virtual ~ResourceDataValue(); + + void setData(const ResourceData *data) { + resData = *data; + } + + void setResource(Resource r, ResourceTracer&& traceInfo) { + res = r; + fTraceInfo = traceInfo; + } + + const ResourceData &getData() const { return resData; } + virtual UResType getType() const; + virtual const UChar *getString(int32_t &length, UErrorCode &errorCode) const; + virtual const UChar *getAliasString(int32_t &length, UErrorCode &errorCode) const; + virtual int32_t getInt(UErrorCode &errorCode) const; + virtual uint32_t getUInt(UErrorCode &errorCode) const; + virtual const int32_t *getIntVector(int32_t &length, UErrorCode &errorCode) const; + virtual const uint8_t *getBinary(int32_t &length, UErrorCode &errorCode) const; + virtual ResourceArray getArray(UErrorCode &errorCode) const; + virtual ResourceTable getTable(UErrorCode &errorCode) const; + virtual UBool isNoInheritanceMarker() const; + virtual int32_t getStringArray(UnicodeString *dest, int32_t capacity, + UErrorCode &errorCode) const; + virtual int32_t getStringArrayOrStringAsArray(UnicodeString *dest, int32_t capacity, + UErrorCode &errorCode) const; + virtual UnicodeString getStringOrFirstOfArray(UErrorCode &errorCode) const; + +private: + // TODO(ICU-20769): If UResourceBundle.fResData becomes a pointer, + // then remove this value field again and just store a pResData pointer. + ResourceData resData; + Resource res; + ResourceTracer fTraceInfo; +}; + +U_NAMESPACE_END + +#endif /* __cplusplus */ /** * Swap an ICU resource bundle. See udataswp.h.