1// Copyright (C) 2022 The Qt Company Ltd.
2// SPDX-License-Identifier: LicenseRef-Qt-Commercial OR LGPL-3.0-only OR GPL-2.0-only OR GPL-3.0-only
3// Qt-Security score:critical reason:data-parser
4
5#ifndef QLATIN1STRINGMATCHER_H
6#define QLATIN1STRINGMATCHER_H
7
8#include <functional>
9#include <iterator>
10#include <limits>
11
12#include <QtCore/q20algorithm.h>
13#include <QtCore/qstring.h>
14
15QT_BEGIN_NAMESPACE
16
17namespace QtPrivate {
18template <typename T> constexpr inline bool isLatin1OrUtf16View = false;
19template <> constexpr inline bool isLatin1OrUtf16View<QLatin1StringView> = true;
20template <> constexpr inline bool isLatin1OrUtf16View<QStringView> = true;
21
22template<class RandomIt1,
23 class Hash = std::hash<typename std::iterator_traits<RandomIt1>::value_type>,
24 class BinaryPredicate = std::equal_to<>>
25class q_boyer_moore_searcher_hashed_needle
26{
27public:
28 constexpr q_boyer_moore_searcher_hashed_needle(RandomIt1 pat_first, RandomIt1 pat_last)
29 : m_skiptable{}
30 {
31 const size_t n = std::distance(pat_first, pat_last);
32 constexpr auto uchar_max = (std::numeric_limits<uchar>::max)();
33 uchar max = n > uchar_max ? uchar_max : uchar(n);
34 q20::fill(first: std::begin(arr&: m_skiptable), last: std::end(arr&: m_skiptable), value: max);
35
36 RandomIt1 pattern = pat_first;
37 pattern += n - max;
38 while (max--)
39 m_skiptable[uchar(*pattern++)] = max;
40 }
41
42 template<class RandomIt2>
43 constexpr auto operator()(RandomIt2 first, RandomIt2 last, RandomIt1 pat_first,
44 RandomIt1 pat_last) const
45 {
46 struct R
47 {
48 RandomIt2 begin, end;
49 };
50 Hash hf;
51 BinaryPredicate pred;
52 auto pat_length = std::distance(pat_first, pat_last);
53 if (pat_length == 0)
54 return R{ first, first };
55
56 const qsizetype pl_minus_one = qsizetype(pat_length - 1);
57 RandomIt2 current = first + pl_minus_one;
58
59 while (current < last) {
60 qsizetype skip = m_skiptable[hf(*current)];
61 if (!skip) {
62 // possible match
63 while (skip < pat_length) {
64 if (!pred(hf(*(current - skip)), uchar(pat_first[pl_minus_one - skip])))
65 break;
66 skip++;
67 }
68 if (skip > pl_minus_one) { // we have a match
69 auto match = current - skip + 1;
70 return R{ match, match + pat_length };
71 }
72
73 // If we don't have a match we are a bit inefficient as we only skip by one
74 // when we have the non matching char in the string.
75 if (m_skiptable[hf(*(current - skip))] == pat_length)
76 skip = pat_length - skip;
77 else
78 skip = 1;
79 }
80 current += skip;
81 }
82
83 return R{ last, last };
84 }
85
86private:
87 alignas(16) uchar m_skiptable[256];
88};
89
90struct QCaseSensitiveLatin1Hash
91{
92 constexpr QCaseSensitiveLatin1Hash() = default;
93
94 constexpr std::size_t operator()(char c) const noexcept { return std::size_t(uchar(c)); }
95};
96
97struct QCaseInsensitiveLatin1Hash
98{
99 constexpr QCaseInsensitiveLatin1Hash() = default;
100
101 constexpr std::size_t operator()(char c) const noexcept
102 {
103 return std::size_t(latin1Lower[uchar(c)]);
104 }
105
106 static int difference(char lhs, char rhs)
107 {
108 return int(latin1Lower[uchar(lhs)]) - int(latin1Lower[uchar(rhs)]);
109 }
110
111 static auto matcher(char ch)
112 {
113 return [sought = latin1Lower[uchar(ch)]](char other) {
114 return latin1Lower[uchar(other)] == sought;
115 };
116 }
117
118private:
119 static constexpr uchar latin1Lower[256] = {
120 0x00,0x01,0x02,0x03,0x04,0x05,0x06,0x07,0x08,0x09,0x0a,0x0b,0x0c,0x0d,0x0e,0x0f,
121 0x10,0x11,0x12,0x13,0x14,0x15,0x16,0x17,0x18,0x19,0x1a,0x1b,0x1c,0x1d,0x1e,0x1f,
122 0x20,0x21,0x22,0x23,0x24,0x25,0x26,0x27,0x28,0x29,0x2a,0x2b,0x2c,0x2d,0x2e,0x2f,
123 0x30,0x31,0x32,0x33,0x34,0x35,0x36,0x37,0x38,0x39,0x3a,0x3b,0x3c,0x3d,0x3e,0x3f,
124 0x40,0x61,0x62,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,
125 0x70,0x71,0x72,0x73,0x74,0x75,0x76,0x77,0x78,0x79,0x7a,0x5b,0x5c,0x5d,0x5e,0x5f,
126 0x60,0x61,0x62,0x63,0x64,0x65,0x66,0x67,0x68,0x69,0x6a,0x6b,0x6c,0x6d,0x6e,0x6f,
127 0x70,0x71,0x72,0x73,0x74,0x75,0x76,0x77,0x78,0x79,0x7a,0x7b,0x7c,0x7d,0x7e,0x7f,
128 0x80,0x81,0x82,0x83,0x84,0x85,0x86,0x87,0x88,0x89,0x8a,0x8b,0x8c,0x8d,0x8e,0x8f,
129 0x90,0x91,0x92,0x93,0x94,0x95,0x96,0x97,0x98,0x99,0x9a,0x9b,0x9c,0x9d,0x9e,0x9f,
130 0xa0,0xa1,0xa2,0xa3,0xa4,0xa5,0xa6,0xa7,0xa8,0xa9,0xaa,0xab,0xac,0xad,0xae,0xaf,
131 0xb0,0xb1,0xb2,0xb3,0xb4,0xb5,0xb6,0xb7,0xb8,0xb9,0xba,0xbb,0xbc,0xbd,0xbe,0xbf,
132 // 0xd7 (multiplication sign) and 0xdf (sz ligature) complicate life
133 0xe0,0xe1,0xe2,0xe3,0xe4,0xe5,0xe6,0xe7,0xe8,0xe9,0xea,0xeb,0xec,0xed,0xee,0xef,
134 0xf0,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xd7,0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xdf,
135 0xe0,0xe1,0xe2,0xe3,0xe4,0xe5,0xe6,0xe7,0xe8,0xe9,0xea,0xeb,0xec,0xed,0xee,0xef,
136 0xf0,0xf1,0xf2,0xf3,0xf4,0xf5,0xf6,0xf7,0xf8,0xf9,0xfa,0xfb,0xfc,0xfd,0xfe,0xff
137 };
138};
139}
140
141class QLatin1StringMatcher
142{
143public:
144 Q_CORE_EXPORT QLatin1StringMatcher() noexcept;
145 Q_CORE_EXPORT explicit QLatin1StringMatcher(
146 QLatin1StringView pattern, Qt::CaseSensitivity cs = Qt::CaseSensitive) noexcept;
147 Q_CORE_EXPORT ~QLatin1StringMatcher() noexcept;
148
149 Q_CORE_EXPORT void setPattern(QLatin1StringView pattern) noexcept;
150 Q_CORE_EXPORT QLatin1StringView pattern() const noexcept;
151 Q_CORE_EXPORT void setCaseSensitivity(Qt::CaseSensitivity cs) noexcept;
152 Q_CORE_EXPORT Qt::CaseSensitivity caseSensitivity() const noexcept;
153
154 Q_CORE_EXPORT qsizetype indexIn(QLatin1StringView haystack, qsizetype from = 0) const noexcept;
155 Q_CORE_EXPORT qsizetype indexIn(QStringView haystack, qsizetype from = 0) const noexcept;
156
157private:
158 void setSearcher() noexcept;
159 void freeSearcher() noexcept;
160
161 QLatin1StringView m_pattern;
162 Qt::CaseSensitivity m_cs;
163 typedef QtPrivate::q_boyer_moore_searcher_hashed_needle<const char *,
164 QtPrivate::QCaseSensitiveLatin1Hash>
165 CaseSensitiveSearcher;
166 typedef QtPrivate::q_boyer_moore_searcher_hashed_needle<const char *,
167 QtPrivate::QCaseInsensitiveLatin1Hash>
168 CaseInsensitiveSearcher;
169 union {
170 CaseSensitiveSearcher m_caseSensitiveSearcher;
171 CaseInsensitiveSearcher m_caseInsensitiveSearcher;
172 };
173
174 template <typename String>
175 qsizetype indexIn_helper(String haystack, qsizetype from) const noexcept;
176
177 char m_foldBuffer[256];
178};
179
180QT_END_NAMESPACE
181
182#endif // QLATIN1MATCHER_H
183

source code of qtbase/src/corelib/text/qlatin1stringmatcher.h