intl/icu/source/i18n/usrchimp.h@fc2d59ddac77 (annotated)
intl/icu/source/i18n/usrchimp.h
Wed, 31 Dec 2014 07:22:50 +0100
- author
- Michael Schloh von Bennewitz <michael@schloh.com>
- date
- Wed, 31 Dec 2014 07:22:50 +0100
- branch
- TOR_BUG_3246
- changeset 4
- fc2d59ddac77
- permissions
- -rw-r--r--
Correct previous dual key logic pending first delivery installment.
| michael@0 | 1 | /* |
| michael@0 | 2 | ********************************************************************** |
| michael@0 | 3 | * Copyright (C) 2001-2011 IBM and others. All rights reserved. |
| michael@0 | 4 | ********************************************************************** |
| michael@0 | 5 | * Date Name Description |
| michael@0 | 6 | * 08/13/2001 synwee Creation. |
| michael@0 | 7 | ********************************************************************** |
| michael@0 | 8 | */ |
| michael@0 | 9 | #ifndef USRCHIMP_H |
| michael@0 | 10 | #define USRCHIMP_H |
| michael@0 | 11 | |
| michael@0 | 12 | #include "unicode/utypes.h" |
| michael@0 | 13 | |
| michael@0 | 14 | #if !UCONFIG_NO_COLLATION |
| michael@0 | 15 | |
| michael@0 | 16 | #include "unicode/normalizer2.h" |
| michael@0 | 17 | #include "unicode/ucol.h" |
| michael@0 | 18 | #include "unicode/ucoleitr.h" |
| michael@0 | 19 | #include "unicode/ubrk.h" |
| michael@0 | 20 | |
| michael@0 | 21 | #define INITIAL_ARRAY_SIZE_ 256 |
| michael@0 | 22 | #define MAX_TABLE_SIZE_ 257 |
| michael@0 | 23 | |
| michael@0 | 24 | struct USearch { |
| michael@0 | 25 | // required since collation element iterator does not have a getText API |
| michael@0 | 26 | const UChar *text; |
| michael@0 | 27 | int32_t textLength; // exact length |
| michael@0 | 28 | UBool isOverlap; |
| michael@0 | 29 | UBool isCanonicalMatch; |
| michael@0 | 30 | int16_t elementComparisonType; |
| michael@0 | 31 | UBreakIterator *internalBreakIter; //internal character breakiterator |
| michael@0 | 32 | UBreakIterator *breakIter; |
| michael@0 | 33 | // value USEARCH_DONE is the default value |
| michael@0 | 34 | // if we are not at the start of the text or the end of the text, |
| michael@0 | 35 | // depending on the iteration direction and matchedIndex is USEARCH_DONE |
| michael@0 | 36 | // it means that we can't find any more matches in that particular direction |
| michael@0 | 37 | int32_t matchedIndex; |
| michael@0 | 38 | int32_t matchedLength; |
| michael@0 | 39 | UBool isForwardSearching; |
| michael@0 | 40 | UBool reset; |
| michael@0 | 41 | }; |
| michael@0 | 42 | |
| michael@0 | 43 | struct UPattern { |
| michael@0 | 44 | const UChar *text; |
| michael@0 | 45 | int32_t textLength; // exact length |
| michael@0 | 46 | // length required for backwards ce comparison |
| michael@0 | 47 | int32_t CELength; |
| michael@0 | 48 | int32_t *CE; |
| michael@0 | 49 | int32_t CEBuffer[INITIAL_ARRAY_SIZE_]; |
| michael@0 | 50 | int32_t PCELength; |
| michael@0 | 51 | int64_t *PCE; |
| michael@0 | 52 | int64_t PCEBuffer[INITIAL_ARRAY_SIZE_]; |
| michael@0 | 53 | UBool hasPrefixAccents; |
| michael@0 | 54 | UBool hasSuffixAccents; |
| michael@0 | 55 | int16_t defaultShiftSize; |
| michael@0 | 56 | int16_t shift[MAX_TABLE_SIZE_]; |
| michael@0 | 57 | int16_t backShift[MAX_TABLE_SIZE_]; |
| michael@0 | 58 | }; |
| michael@0 | 59 | |
| michael@0 | 60 | struct UStringSearch { |
| michael@0 | 61 | struct USearch *search; |
| michael@0 | 62 | struct UPattern pattern; |
| michael@0 | 63 | const UCollator *collator; |
| michael@0 | 64 | const icu::Normalizer2 *nfd; |
| michael@0 | 65 | // positions within the collation element iterator is used to determine |
| michael@0 | 66 | // if we are at the start of the text. |
| michael@0 | 67 | UCollationElements *textIter; |
| michael@0 | 68 | // utility collation element, used throughout program for temporary |
| michael@0 | 69 | // iteration. |
| michael@0 | 70 | UCollationElements *utilIter; |
| michael@0 | 71 | UBool ownCollator; |
| michael@0 | 72 | UCollationStrength strength; |
| michael@0 | 73 | uint32_t ceMask; |
| michael@0 | 74 | uint32_t variableTop; |
| michael@0 | 75 | UBool toShift; |
| michael@0 | 76 | UChar canonicalPrefixAccents[INITIAL_ARRAY_SIZE_]; |
| michael@0 | 77 | UChar canonicalSuffixAccents[INITIAL_ARRAY_SIZE_]; |
| michael@0 | 78 | }; |
| michael@0 | 79 | |
| michael@0 | 80 | /** |
| michael@0 | 81 | * Exact matches without checking for the ends for extra accents. |
| michael@0 | 82 | * The match after the position within the collation element iterator is to be |
| michael@0 | 83 | * found. |
| michael@0 | 84 | * After a match is found the offset in the collation element iterator will be |
| michael@0 | 85 | * shifted to the start of the match. |
| michael@0 | 86 | * Implementation note: |
| michael@0 | 87 | * For tertiary we can't use the collator->tertiaryMask, that is a |
| michael@0 | 88 | * preprocessed mask that takes into account case options. since we are only |
| michael@0 | 89 | * concerned with exact matches, we don't need that. |
| michael@0 | 90 | * Alternate handling - since only the 16 most significant digits is only used, |
| michael@0 | 91 | * we can safely do a compare without masking if the ce is a variable, we mask |
| michael@0 | 92 | * and get only the primary values no shifting to quartenary is required since |
| michael@0 | 93 | * all primary values less than variabletop will need to be masked off anyway. |
| michael@0 | 94 | * If the end character is composite and the pattern ce does not match the text |
| michael@0 | 95 | * ce, we skip it until we find a match in the end composite character or when |
| michael@0 | 96 | * it has passed the character. This is so that we can match pattern "a" with |
| michael@0 | 97 | * the text "\u00e6" |
| michael@0 | 98 | * @param strsrch string search data |
| michael@0 | 99 | * @param status error status if any |
| michael@0 | 100 | * @return TRUE if an exact match is found, FALSE otherwise |
| michael@0 | 101 | */ |
| michael@0 | 102 | U_CFUNC |
| michael@0 | 103 | UBool usearch_handleNextExact(UStringSearch *strsrch, UErrorCode *status); |
| michael@0 | 104 | |
| michael@0 | 105 | /** |
| michael@0 | 106 | * Canonical matches. |
| michael@0 | 107 | * According to the definition, matches found here will include the whole span |
| michael@0 | 108 | * of beginning and ending accents if it overlaps that region. |
| michael@0 | 109 | * @param strsrch string search data |
| michael@0 | 110 | * @param status error status if any |
| michael@0 | 111 | * @return TRUE if a canonical match is found, FALSE otherwise |
| michael@0 | 112 | */ |
| michael@0 | 113 | U_CFUNC |
| michael@0 | 114 | UBool usearch_handleNextCanonical(UStringSearch *strsrch, UErrorCode *status); |
| michael@0 | 115 | |
| michael@0 | 116 | /** |
| michael@0 | 117 | * Gets the previous match. |
| michael@0 | 118 | * Comments follows from handleNextExact |
| michael@0 | 119 | * @param strsrch string search data |
| michael@0 | 120 | * @param status error status if any |
| michael@0 | 121 | * @return True if a exact math is found, FALSE otherwise. |
| michael@0 | 122 | */ |
| michael@0 | 123 | U_CFUNC |
| michael@0 | 124 | UBool usearch_handlePreviousExact(UStringSearch *strsrch, UErrorCode *status); |
| michael@0 | 125 | |
| michael@0 | 126 | /** |
| michael@0 | 127 | * Canonical matches. |
| michael@0 | 128 | * According to the definition, matches found here will include the whole span |
| michael@0 | 129 | * of beginning and ending accents if it overlaps that region. |
| michael@0 | 130 | * @param strsrch string search data |
| michael@0 | 131 | * @param status error status if any |
| michael@0 | 132 | * @return TRUE if a canonical match is found, FALSE otherwise |
| michael@0 | 133 | */ |
| michael@0 | 134 | U_CFUNC |
| michael@0 | 135 | UBool usearch_handlePreviousCanonical(UStringSearch *strsrch, |
| michael@0 | 136 | UErrorCode *status); |
| michael@0 | 137 | |
| michael@0 | 138 | #endif /* #if !UCONFIG_NO_COLLATION */ |
| michael@0 | 139 | |
| michael@0 | 140 | #endif |
