]> git.saurik.com Git - wxWidgets.git/blob - include/wx/stringops.h
optimize FindItem(data) performance (closes #9870)
[wxWidgets.git] / include / wx / stringops.h
1 ///////////////////////////////////////////////////////////////////////////////
2 // Name: wx/stringops.h
3 // Purpose: implementation of wxString primitive operations
4 // Author: Vaclav Slavik
5 // Modified by:
6 // Created: 2007-04-16
7 // RCS-ID: $Id$
8 // Copyright: (c) 2007 REA Elektronik GmbH
9 // Licence: wxWindows licence
10 ///////////////////////////////////////////////////////////////////////////////
11
12 #ifndef _WX_WXSTRINGOPS_H__
13 #define _WX_WXSTRINGOPS_H__
14
15 #include "wx/chartype.h"
16 #include "wx/stringimpl.h"
17 #include "wx/unichar.h"
18 #include "wx/buffer.h"
19
20 // This header contains wxStringOperations "namespace" class that implements
21 // elementary operations on string data as static methods; wxString methods and
22 // iterators are implemented in terms of it. Two implementations are available,
23 // one for UTF-8 encoded char* string and one for "raw" wchar_t* strings (or
24 // char* in ANSI build).
25
26 // FIXME-UTF8: only wchar after we remove ANSI build
27 #if wxUSE_UNICODE_WCHAR || !wxUSE_UNICODE
28 struct WXDLLIMPEXP_BASE wxStringOperationsWchar
29 {
30 // moves the iterator to the next Unicode character
31 template <typename Iterator>
32 static void IncIter(Iterator& i) { ++i; }
33
34 // moves the iterator to the previous Unicode character
35 template <typename Iterator>
36 static void DecIter(Iterator& i) { --i; }
37
38 // moves the iterator by n Unicode characters
39 template <typename Iterator>
40 static Iterator AddToIter(const Iterator& i, ptrdiff_t n)
41 { return i + n; }
42
43 // returns distance of the two iterators in Unicode characters
44 template <typename Iterator>
45 static ptrdiff_t DiffIters(const Iterator& i1, const Iterator& i2)
46 { return i1 - i2; }
47
48 // encodes the character to a form used to represent it in internal
49 // representation (returns a string in UTF8 version)
50 static wxChar EncodeChar(const wxUniChar& ch) { return (wxChar)ch; }
51
52 static wxUniChar DecodeChar(const wxStringImpl::const_iterator& i)
53 { return *i; }
54 };
55 #endif // wxUSE_UNICODE_WCHAR || !wxUSE_UNICODE
56
57
58 #if wxUSE_UNICODE_UTF8
59 struct WXDLLIMPEXP_BASE wxStringOperationsUtf8
60 {
61 // checks correctness of UTF-8 sequence
62 static bool IsValidUtf8String(const char *c,
63 size_t len = wxStringImpl::npos);
64 static bool IsValidUtf8LeadByte(unsigned char c)
65 {
66 return (c <= 0x7F) || (c >= 0xC2 && c <= 0xF4);
67 }
68
69 // table of offsets to skip forward when iterating over UTF-8 sequence
70 static const unsigned char ms_utf8IterTable[256];
71
72
73 template<typename Iterator>
74 static void IncIter(Iterator& i)
75 {
76 wxASSERT( IsValidUtf8LeadByte(*i) );
77 i += ms_utf8IterTable[(unsigned char)*i];
78 }
79
80 template<typename Iterator>
81 static void DecIter(Iterator& i)
82 {
83 wxASSERT( IsValidUtf8LeadByte(*i) );
84
85 // Non-lead bytes are all in the 0x80..0xBF range (i.e. 10xxxxxx in
86 // binary), so we just have to go back until we hit a byte that is
87 // either < 0x80 (i.e. 0xxxxxxx in binary) or 0xC0..0xFF (11xxxxxx in
88 // binary; this includes some invalid values, but we can ignore it
89 // here, because we assume valid UTF-8 input for the purpose of
90 // efficient implementation).
91 --i;
92 while ( ((*i) & 0xC0) == 0x80 /* 2 highest bits are '10' */ )
93 --i;
94 }
95
96 template<typename Iterator>
97 static Iterator AddToIter(const Iterator& i, ptrdiff_t n)
98 {
99 Iterator out(i);
100
101 if ( n > 0 )
102 {
103 for ( ptrdiff_t j = 0; j < n; ++j )
104 IncIter(out);
105 }
106 else if ( n < 0 )
107 {
108 for ( ptrdiff_t j = 0; j > n; --j )
109 DecIter(out);
110 }
111
112 return out;
113 }
114
115 template<typename Iterator>
116 static ptrdiff_t DiffIters(Iterator i1, Iterator i2)
117 {
118 ptrdiff_t dist = 0;
119
120 if ( i1 < i2 )
121 {
122 while ( i1 != i2 )
123 {
124 IncIter(i1);
125 dist--;
126 }
127 }
128 else if ( i2 < i1 )
129 {
130 while ( i2 != i1 )
131 {
132 IncIter(i2);
133 dist++;
134 }
135 }
136
137 return dist;
138 }
139
140 // encodes the character as UTF-8:
141 typedef wxUniChar::Utf8CharBuffer Utf8CharBuffer;
142 static Utf8CharBuffer EncodeChar(const wxUniChar& ch)
143 { return ch.AsUTF8(); }
144
145 // returns n copies of ch encoded in UTF-8 string
146 static wxCharBuffer EncodeNChars(size_t n, const wxUniChar& ch);
147
148 // returns the length of UTF-8 encoding of the character with lead byte 'c'
149 static size_t GetUtf8CharLength(char c)
150 {
151 wxASSERT( IsValidUtf8LeadByte(c) );
152 return ms_utf8IterTable[(unsigned char)c];
153 }
154
155 // decodes single UTF-8 character from UTF-8 string
156 static wxUniChar DecodeChar(wxStringImpl::const_iterator i)
157 {
158 if ( (unsigned char)*i < 0x80 )
159 return (int)*i;
160 return DecodeNonAsciiChar(i);
161 }
162
163 private:
164 static wxUniChar DecodeNonAsciiChar(wxStringImpl::const_iterator i);
165 };
166 #endif // wxUSE_UNICODE_UTF8
167
168
169 #if wxUSE_UNICODE_UTF8
170 typedef wxStringOperationsUtf8 wxStringOperations;
171 #else
172 typedef wxStringOperationsWchar wxStringOperations;
173 #endif
174
175 #endif // _WX_WXSTRINGOPS_H_