ICU 56 update step 1

source/common/uresdata.h

 /*
 ******************************************************************************
-*                                                                            *
-* Copyright (C) 1999-2012, International Business Machines                   *
-*                Corporation and others. All Rights Reserved.                *
-*                                                                            *
+* Copyright (C) 1999-2015, International Business Machines
+*                Corporation and others. All Rights Reserved.
 ******************************************************************************
 *   file name:  uresdata.h
 *   encoding:   US-ASCII
 *   tab size:   8 (not used)
 *   indentation:4
 *
 *   created on: 1999dec08
 *   created by: Markus W. Scherer
 *   06/24/02    weiv        Added support for resource sharing
 */
@@ skipping 27 matching lines @@
     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
  * 27..0    28-bit four-byte-offset or value according to the type
  */
 typedef uint32_t Resource;
 
 #define RES_BOGUS 0xffffffff
+#define RES_MAX_OFFSET 0x0fffffff
 
 #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 */
 #if U_SIGNED_RIGHT_SHIFT_IS_ARITHMETIC
 #   define RES_GET_INT(res) (((int32_t)((res)<<4L))>>4L)
 #else
 #   define RES_GET_INT(res) (int32_t)(((res)&0x08000000) ? (res)|0xf0000000 : (res)&0x07ffffff)
 #endif
 
 #define RES_GET_UINT(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[];
-                                 *     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 */
-    URES_INDEX_KEYS_TOP,        /* [1] contains the top of the key strings, */
-                                /*     same as the bottom of resources or UTF-16 strings, 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 */
-    URES_INDEX_ATTRIBUTES,      /* [5] attributes bit set, see URES_ATT_* (new in formatVersion 1.2) */
-    URES_INDEX_16BIT_TOP,       /* [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_POOL_CHECKSUM,   /* [7] checksum of the pool bundle (new in formatVersion 2.0, ICU 4.4) */
+    /**
+     * [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
 };
 
 /*
  * 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
 
 /*
  * 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 (formatVersion=2, ICU 4.4)
+ * File format for .res resource bundle files
  *
- * New in formatVersion 2 compared with 1.3: -------------
+ * 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
@@ skipping 121 matching lines @@
  *
  * Most resources have their values stored at four-byte offsets from the start
  * of the resource data. These values are at least 4-aligned.
  * 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<poolStringIndex16Limit
+ * then use the pool bundle's 16-bit-units array,
+ * otherwise subtract that limit and use the local 16-bit-units array.
  *
  * Type Name            Memory layout of values
  *                      (in parentheses: scalar, non-offset values)
  *
  * 0  Unicode String:   int32_t length, UChar[length], (UChar)0, (padding)
  *                  or  (empty string ("") if offset==0)
  * 1  Binary:           int32_t length, uint8_t[length], (padding)
  *                      - the start of the bytes is 16-aligned -
  * 2  Table:            uint16_t count, uint16_t keyStringOffsets[count], (uint16_t padding), Resource[count]
  * 3  Alias:            (physically same value layout as string, new in ICU 2.4)
  * 4  Table32:          int32_t count, int32_t keyStringOffsets[count], Resource[count]
  *                      (new in formatVersion 1.1/ICU 2.8)
  * 5  Table16:          uint16_t count, uint16_t keyStringOffsets[count], Resource16[count]
  *                      (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4)
  * 6  Unicode String-v2:UChar[length], (UChar)0; length determined by the first UChar:
  *                      - if first is not a trail surrogate, then the length is implicit
  *                        and u_strlen() needs to be called
  *                      - if first<0xdfef then length=first&0x3ff (and skip first)
  *                      - if first<0xdfff then length=((first-0xdfef)<<16) | second UChar
  *                      - if first==0xdfff then length=((second UChar)<<16) | third UChar
  *                      (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4)
+ *
+ *                      Starting with formatVersion 3/ICU 56, if offset<poolStringIndexLimit
+ *                      then use the pool bundle's 16-bit-units array,
+ *                      otherwise subtract that limit and use the local 16-bit-units array.
+ *                      (Note different limits for Resource16 vs. Resource.)
+ *
  * 7  Integer:          (28-bit offset is integer value)
  * 8  Array:            int32_t count, Resource[count]
  * 9  Array16:          uint16_t count, Resource16[count]
  *                      (stored in the 16-bit-units array; new in formatVersion 2/ICU 4.4)
  * 14 Integer Vector:   int32_t length, int32_t[length]
  * 15 Reserved:         This value denotes special purpose resources and is for internal use.
  *
  * Note that there are 3 types with data vector values:
  * - Vectors of 8-bit bytes stored as type Binary.
  * - Vectors of 16-bit words stored as type Unicode String or Unicode String-v2
  *                     (no value restrictions, all values 0..ffff allowed!).
  * - Vectors of 32-bit words stored as type Integer Vector.
  */
 
 /*
  * Structure for a single, memory-mapped ResourceBundle.
  */
 typedef struct {
     UDataMemory *data;
     const int32_t *pRoot;
     const uint16_t *p16BitUnits;
     const char *poolBundleKeys;
     Resource rootRes;
     int32_t localKeyLimit;
+    const uint16_t *poolBundleStrings;
+    int32_t poolStringIndexLimit;
+    int32_t poolStringIndex16Limit;
     UBool noFallback; /* see URES_ATT_NO_FALLBACK */
     UBool isPoolBundle;
     UBool usesPoolBundle;
     UBool useNativeStrcmp;
 } ResourceData;
 
 /*
  * Read a resource bundle from memory.
  */
 U_INTERNAL void U_EXPORT2
@@ skipping 60 matching lines @@
 /**
  * Swap an ICU resource bundle. See udataswp.h.
  * @internal
  */
 U_CAPI int32_t U_EXPORT2
 ures_swap(const UDataSwapper *ds,
           const void *inData, int32_t length, void *outData,
           UErrorCode *pErrorCode);
 
 #endif