1 | // © 2016 and later: Unicode, Inc. and others. |
2 | // License & terms of use: http://www.unicode.org/copyright.html |
3 | /* |
4 | ****************************************************************************** |
5 | * |
6 | * Copyright (C) 2007, International Business Machines |
7 | * Corporation and others. All Rights Reserved. |
8 | * |
9 | ****************************************************************************** |
10 | * file name: unisetspan.h |
11 | * encoding: UTF-8 |
12 | * tab size: 8 (not used) |
13 | * indentation:4 |
14 | * |
15 | * created on: 2007mar01 |
16 | * created by: Markus W. Scherer |
17 | */ |
18 | |
19 | #ifndef __UNISETSPAN_H__ |
20 | #define __UNISETSPAN_H__ |
21 | |
22 | #include "unicode/utypes.h" |
23 | #include "unicode/uniset.h" |
24 | |
25 | U_NAMESPACE_BEGIN |
26 | |
27 | /* |
28 | * Implement span() etc. for a set with strings. |
29 | * Avoid recursion because of its exponential complexity. |
30 | * Instead, try multiple paths at once and track them with an IndexList. |
31 | */ |
32 | class UnicodeSetStringSpan : public UMemory { |
33 | public: |
34 | /* |
35 | * Which span() variant will be used? |
36 | * The object is either built for one variant and used once, |
37 | * or built for all and may be used many times. |
38 | */ |
39 | enum { |
40 | FWD = 0x20, |
41 | BACK = 0x10, |
42 | UTF16 = 8, |
43 | UTF8 = 4, |
44 | CONTAINED = 2, |
45 | NOT_CONTAINED = 1, |
46 | |
47 | ALL = 0x3f, |
48 | |
49 | FWD_UTF16_CONTAINED = FWD | UTF16 | CONTAINED, |
50 | FWD_UTF16_NOT_CONTAINED = FWD | UTF16 | NOT_CONTAINED, |
51 | FWD_UTF8_CONTAINED = FWD | UTF8 | CONTAINED, |
52 | FWD_UTF8_NOT_CONTAINED = FWD | UTF8 | NOT_CONTAINED, |
53 | BACK_UTF16_CONTAINED = BACK | UTF16 | CONTAINED, |
54 | BACK_UTF16_NOT_CONTAINED= BACK | UTF16 | NOT_CONTAINED, |
55 | BACK_UTF8_CONTAINED = BACK | UTF8 | CONTAINED, |
56 | BACK_UTF8_NOT_CONTAINED = BACK | UTF8 | NOT_CONTAINED |
57 | }; |
58 | |
59 | UnicodeSetStringSpan(const UnicodeSet &set, const UVector &setStrings, uint32_t which); |
60 | |
61 | // Copy constructor. Assumes which==ALL for a frozen set. |
62 | UnicodeSetStringSpan(const UnicodeSetStringSpan &otherStringSpan, const UVector &newParentSetStrings); |
63 | |
64 | ~UnicodeSetStringSpan(); |
65 | |
66 | /* |
67 | * Do the strings need to be checked in span() etc.? |
68 | * @return TRUE if strings need to be checked (call span() here), |
69 | * FALSE if not (use a BMPSet for best performance). |
70 | */ |
71 | inline UBool needsStringSpanUTF16(); |
72 | inline UBool needsStringSpanUTF8(); |
73 | |
74 | // For fast UnicodeSet::contains(c). |
75 | inline UBool contains(UChar32 c) const; |
76 | |
77 | int32_t span(const UChar *s, int32_t length, USetSpanCondition spanCondition) const; |
78 | |
79 | int32_t spanBack(const UChar *s, int32_t length, USetSpanCondition spanCondition) const; |
80 | |
81 | int32_t spanUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const; |
82 | |
83 | int32_t spanBackUTF8(const uint8_t *s, int32_t length, USetSpanCondition spanCondition) const; |
84 | |
85 | private: |
86 | // Special spanLength byte values. |
87 | enum { |
88 | // The spanLength is >=0xfe. |
89 | LONG_SPAN=0xfe, |
90 | // All code points in the string are contained in the parent set. |
91 | ALL_CP_CONTAINED=0xff |
92 | }; |
93 | |
94 | // Add a starting or ending string character to the spanNotSet |
95 | // so that a character span ends before any string. |
96 | void addToSpanNotSet(UChar32 c); |
97 | |
98 | int32_t spanNot(const UChar *s, int32_t length) const; |
99 | int32_t spanNotBack(const UChar *s, int32_t length) const; |
100 | int32_t spanNotUTF8(const uint8_t *s, int32_t length) const; |
101 | int32_t spanNotBackUTF8(const uint8_t *s, int32_t length) const; |
102 | |
103 | // Set for span(). Same as parent but without strings. |
104 | UnicodeSet spanSet; |
105 | |
106 | // Set for span(not contained). |
107 | // Same as spanSet, plus characters that start or end strings. |
108 | UnicodeSet *pSpanNotSet; |
109 | |
110 | // The strings of the parent set. |
111 | const UVector &strings; |
112 | |
113 | // Pointer to the UTF-8 string lengths. |
114 | // Also pointer to further allocated storage for meta data and |
115 | // UTF-8 string contents as necessary. |
116 | int32_t *utf8Lengths; |
117 | |
118 | // Pointer to the part of the (utf8Lengths) memory block that stores |
119 | // the lengths of span(), spanBack() etc. for each string. |
120 | uint8_t *spanLengths; |
121 | |
122 | // Pointer to the part of the (utf8Lengths) memory block that stores |
123 | // the UTF-8 versions of the parent set's strings. |
124 | uint8_t *utf8; |
125 | |
126 | // Number of bytes for all UTF-8 versions of strings together. |
127 | int32_t utf8Length; |
128 | |
129 | // Maximum lengths of relevant strings. |
130 | int32_t maxLength16; |
131 | int32_t maxLength8; |
132 | |
133 | // Set up for all variants of span()? |
134 | UBool all; |
135 | |
136 | // Memory for small numbers and lengths of strings. |
137 | // For example, for 8 strings: |
138 | // 8 UTF-8 lengths, 8*4 bytes span lengths, 8*2 3-byte UTF-8 characters |
139 | // = 112 bytes = int32_t[28]. |
140 | int32_t staticLengths[32]; |
141 | }; |
142 | |
143 | UBool UnicodeSetStringSpan::needsStringSpanUTF16() { |
144 | return (UBool)(maxLength16!=0); |
145 | } |
146 | |
147 | UBool UnicodeSetStringSpan::needsStringSpanUTF8() { |
148 | return (UBool)(maxLength8!=0); |
149 | } |
150 | |
151 | UBool UnicodeSetStringSpan::contains(UChar32 c) const { |
152 | return spanSet.contains(c); |
153 | } |
154 | |
155 | U_NAMESPACE_END |
156 | |
157 | #endif |
158 | |