1 // Copyright (C) 2016 and later: Unicode, Inc. and others. 2 // License & terms of use: http://www.unicode.org/copyright.html 3 /* 4 ******************************************************************************* 5 * Copyright (C) 2000-2011, International Business Machines 6 * Corporation and others. All Rights Reserved. 7 ******************************************************************************* 8 * file name: ucol_data.h 9 * encoding: US-ASCII 10 * tab size: 8 (not used) 11 * indentation:4 12 * 13 * created on: 2011jul02 14 * created by: Markus Scherer 15 * 16 * Private implementation header for C/C++ collation. 17 * Some file data structure definitions were moved here from i18n/ucol_imp.h 18 * so that the common library (via ucol_swp.cpp) need not depend on the i18n library at all. 19 * 20 * We do not want to move the collation swapper to the i18n library because 21 * a) the resource bundle swapper depends on it and would have to move too, and 22 * b) we might want to eventually implement runtime data swapping, 23 * which might (or might not) be easier if all swappers are in the common library. 24 */ 25 26 #ifndef __UCOL_DATA_H__ 27 #define __UCOL_DATA_H__ 28 29 #include "unicode/utypes.h" 30 31 #if !UCONFIG_NO_COLLATION 32 33 /* let us know whether reserved fields are reset to zero or junked */ 34 #define UCOL_HEADER_MAGIC 0x20030618 35 36 typedef struct { 37 int32_t size; 38 /* all the offsets are in bytes */ 39 /* to get the address add to the header address and cast properly */ 40 uint32_t options; /* these are the default options for the collator */ 41 uint32_t UCAConsts; /* structure which holds values for indirect positioning and implicit ranges */ 42 uint32_t contractionUCACombos; /* this one is needed only for UCA, to copy the appropriate contractions */ 43 uint32_t magic; /* magic number - lets us know whether reserved data is reset or junked */ 44 uint32_t mappingPosition; /* const uint8_t *mappingPosition; */ 45 uint32_t expansion; /* uint32_t *expansion; */ 46 uint32_t contractionIndex; /* UChar *contractionIndex; */ 47 uint32_t contractionCEs; /* uint32_t *contractionCEs; */ 48 uint32_t contractionSize; /* needed for various closures */ 49 /*int32_t latinOneMapping;*/ /* this is now handled in the trie itself *//* fast track to latin1 chars */ 50 51 uint32_t endExpansionCE; /* array of last collation element in 52 expansion */ 53 uint32_t expansionCESize; /* array of maximum expansion size 54 corresponding to the expansion 55 collation elements with last element 56 in endExpansionCE*/ 57 int32_t endExpansionCECount; /* size of endExpansionCE */ 58 uint32_t unsafeCP; /* hash table of unsafe code points */ 59 uint32_t contrEndCP; /* hash table of final code points */ 60 /* in contractions. */ 61 62 int32_t contractionUCACombosSize; /* number of UCA contraction items. */ 63 /*Length is contractionUCACombosSize*contractionUCACombosWidth*sizeof(UChar) */ 64 UBool jamoSpecial; /* is jamoSpecial */ 65 UBool isBigEndian; /* is this data big endian? from the UDataInfo header*/ 66 uint8_t charSetFamily; /* what is the charset family of this data from the UDataInfo header*/ 67 uint8_t contractionUCACombosWidth; /* width of UCA combos field */ 68 UVersionInfo version; 69 UVersionInfo UCAVersion; /* version of the UCA, read from file */ 70 UVersionInfo UCDVersion; /* UCD version, obtained by u_getUnicodeVersion */ 71 UVersionInfo formatVersion; /* format version from the UDataInfo header */ 72 uint32_t scriptToLeadByte; /* offset to script to lead collation byte mapping data */ 73 uint32_t leadByteToScript; /* offset to lead collation byte to script mapping data */ 74 uint8_t reserved[76]; /* for future use */ 75 } UCATableHeader; 76 77 typedef struct { 78 uint32_t byteSize; 79 uint32_t tableSize; 80 uint32_t contsSize; 81 uint32_t table; 82 uint32_t conts; 83 UVersionInfo UCAVersion; /* version of the UCA, read from file */ 84 uint8_t padding[8]; 85 } InverseUCATableHeader; 86 87 #endif /* !UCONFIG_NO_COLLATION */ 88 89 #endif /* __UCOL_DATA_H__ */ 90
