intl/icu/source/common/unicode/unifilt.h@6474c204b198 (annotated)
intl/icu/source/common/unicode/unifilt.h
Wed, 31 Dec 2014 06:09:35 +0100
- author
- Michael Schloh von Bennewitz <michael@schloh.com>
- date
- Wed, 31 Dec 2014 06:09:35 +0100
- changeset 0
- 6474c204b198
- permissions
- -rw-r--r--
Cloned upstream origin tor-browser at tor-browser-31.3.0esr-4.5-1-build1
revision ID fc1c9ff7c1b2defdbc039f12214767608f46423f for hacking purpose.
| michael@0 | 1 | /* |
| michael@0 | 2 | ********************************************************************** |
| michael@0 | 3 | * Copyright (C) 1999-2010, International Business Machines Corporation and others. |
| michael@0 | 4 | * All Rights Reserved. |
| michael@0 | 5 | ********************************************************************** |
| michael@0 | 6 | * Date Name Description |
| michael@0 | 7 | * 11/17/99 aliu Creation. |
| michael@0 | 8 | ********************************************************************** |
| michael@0 | 9 | */ |
| michael@0 | 10 | #ifndef UNIFILT_H |
| michael@0 | 11 | #define UNIFILT_H |
| michael@0 | 12 | |
| michael@0 | 13 | #include "unicode/unifunct.h" |
| michael@0 | 14 | #include "unicode/unimatch.h" |
| michael@0 | 15 | |
| michael@0 | 16 | /** |
| michael@0 | 17 | * \file |
| michael@0 | 18 | * \brief C++ API: Unicode Filter |
| michael@0 | 19 | */ |
| michael@0 | 20 | |
| michael@0 | 21 | U_NAMESPACE_BEGIN |
| michael@0 | 22 | |
| michael@0 | 23 | /** |
| michael@0 | 24 | * U_ETHER is used to represent character values for positions outside |
| michael@0 | 25 | * a range. For example, transliterator uses this to represent |
| michael@0 | 26 | * characters outside the range contextStart..contextLimit-1. This |
| michael@0 | 27 | * allows explicit matching by rules and UnicodeSets of text outside a |
| michael@0 | 28 | * defined range. |
| michael@0 | 29 | * @stable ICU 3.0 |
| michael@0 | 30 | */ |
| michael@0 | 31 | #define U_ETHER ((UChar)0xFFFF) |
| michael@0 | 32 | |
| michael@0 | 33 | /** |
| michael@0 | 34 | * |
| michael@0 | 35 | * <code>UnicodeFilter</code> defines a protocol for selecting a |
| michael@0 | 36 | * subset of the full range (U+0000 to U+10FFFF) of Unicode characters. |
| michael@0 | 37 | * Currently, filters are used in conjunction with classes like {@link |
| michael@0 | 38 | * Transliterator} to only process selected characters through a |
| michael@0 | 39 | * transformation. |
| michael@0 | 40 | * |
| michael@0 | 41 | * <p>Note: UnicodeFilter currently stubs out two pure virtual methods |
| michael@0 | 42 | * of its base class, UnicodeMatcher. These methods are toPattern() |
| michael@0 | 43 | * and matchesIndexValue(). This is done so that filter classes that |
| michael@0 | 44 | * are not actually used as matchers -- specifically, those in the |
| michael@0 | 45 | * UnicodeFilterLogic component, and those in tests -- can continue to |
| michael@0 | 46 | * work without defining these methods. As long as a filter is not |
| michael@0 | 47 | * used in an RBT during real transliteration, these methods will not |
| michael@0 | 48 | * be called. However, this breaks the UnicodeMatcher base class |
| michael@0 | 49 | * protocol, and it is not a correct solution. |
| michael@0 | 50 | * |
| michael@0 | 51 | * <p>In the future we may revisit the UnicodeMatcher / UnicodeFilter |
| michael@0 | 52 | * hierarchy and either redesign it, or simply remove the stubs in |
| michael@0 | 53 | * UnicodeFilter and force subclasses to implement the full |
| michael@0 | 54 | * UnicodeMatcher protocol. |
| michael@0 | 55 | * |
| michael@0 | 56 | * @see UnicodeFilterLogic |
| michael@0 | 57 | * @stable ICU 2.0 |
| michael@0 | 58 | */ |
| michael@0 | 59 | class U_COMMON_API UnicodeFilter : public UnicodeFunctor, public UnicodeMatcher { |
| michael@0 | 60 | |
| michael@0 | 61 | public: |
| michael@0 | 62 | /** |
| michael@0 | 63 | * Destructor |
| michael@0 | 64 | * @stable ICU 2.0 |
| michael@0 | 65 | */ |
| michael@0 | 66 | virtual ~UnicodeFilter(); |
| michael@0 | 67 | |
| michael@0 | 68 | /** |
| michael@0 | 69 | * Returns <tt>true</tt> for characters that are in the selected |
| michael@0 | 70 | * subset. In other words, if a character is <b>to be |
| michael@0 | 71 | * filtered</b>, then <tt>contains()</tt> returns |
| michael@0 | 72 | * <b><tt>false</tt></b>. |
| michael@0 | 73 | * @stable ICU 2.0 |
| michael@0 | 74 | */ |
| michael@0 | 75 | virtual UBool contains(UChar32 c) const = 0; |
| michael@0 | 76 | |
| michael@0 | 77 | /** |
| michael@0 | 78 | * UnicodeFunctor API. Cast 'this' to a UnicodeMatcher* pointer |
| michael@0 | 79 | * and return the pointer. |
| michael@0 | 80 | * @stable ICU 2.4 |
| michael@0 | 81 | */ |
| michael@0 | 82 | virtual UnicodeMatcher* toMatcher() const; |
| michael@0 | 83 | |
| michael@0 | 84 | /** |
| michael@0 | 85 | * Implement UnicodeMatcher API. |
| michael@0 | 86 | * @stable ICU 2.4 |
| michael@0 | 87 | */ |
| michael@0 | 88 | virtual UMatchDegree matches(const Replaceable& text, |
| michael@0 | 89 | int32_t& offset, |
| michael@0 | 90 | int32_t limit, |
| michael@0 | 91 | UBool incremental); |
| michael@0 | 92 | |
| michael@0 | 93 | /** |
| michael@0 | 94 | * UnicodeFunctor API. Nothing to do. |
| michael@0 | 95 | * @stable ICU 2.4 |
| michael@0 | 96 | */ |
| michael@0 | 97 | virtual void setData(const TransliterationRuleData*); |
| michael@0 | 98 | |
| michael@0 | 99 | /** |
| michael@0 | 100 | * ICU "poor man's RTTI", returns a UClassID for this class. |
| michael@0 | 101 | * |
| michael@0 | 102 | * @stable ICU 2.2 |
| michael@0 | 103 | */ |
| michael@0 | 104 | static UClassID U_EXPORT2 getStaticClassID(); |
| michael@0 | 105 | |
| michael@0 | 106 | protected: |
| michael@0 | 107 | |
| michael@0 | 108 | /* |
| michael@0 | 109 | * Since this class has pure virtual functions, |
| michael@0 | 110 | * a constructor can't be used. |
| michael@0 | 111 | * @stable ICU 2.0 |
| michael@0 | 112 | */ |
| michael@0 | 113 | /* UnicodeFilter();*/ |
| michael@0 | 114 | }; |
| michael@0 | 115 | |
| michael@0 | 116 | /*inline UnicodeFilter::UnicodeFilter() {}*/ |
| michael@0 | 117 | |
| michael@0 | 118 | U_NAMESPACE_END |
| michael@0 | 119 | |
| michael@0 | 120 | #endif |
