1/*
2 * (C) 1999 Lars Knoll (knoll@kde.org)
3 * Copyright (C) 2004-2018 Apple Inc. All rights reserved.
4 *
5 * This library is free software; you can redistribute it and/or
6 * modify it under the terms of the GNU Library General Public
7 * License as published by the Free Software Foundation; either
8 * version 2 of the License, or (at your option) any later version.
9 *
10 * This library is distributed in the hope that it will be useful,
11 * but WITHOUT ANY WARRANTY; without even the implied warranty of
12 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
13 * Library General Public License for more details.
14 *
15 * You should have received a copy of the GNU Library General Public License
16 * along with this library; see the file COPYING.LIB. If not, write to
17 * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
18 * Boston, MA 02110-1301, USA.
19 *
20 */
21
22#pragma once
23
24// This file would be called String.h, but that conflicts with <string.h>
25// on systems without case-sensitive file systems.
26
27#include <stdarg.h>
28#include <wtf/Function.h>
29#include <wtf/text/ASCIILiteral.h>
30#include <wtf/text/IntegerToStringConversion.h>
31#include <wtf/text/StringImpl.h>
32
33#ifdef __OBJC__
34#include <objc/objc.h>
35#endif
36
37#if OS(WINDOWS)
38#include <wtf/text/win/WCharStringExtras.h>
39#endif
40
41namespace WTF {
42
43// Declarations of string operations
44
45WTF_EXPORT_PRIVATE int charactersToIntStrict(const LChar*, size_t, bool* ok = nullptr, int base = 10);
46WTF_EXPORT_PRIVATE int charactersToIntStrict(const UChar*, size_t, bool* ok = nullptr, int base = 10);
47WTF_EXPORT_PRIVATE unsigned charactersToUIntStrict(const LChar*, size_t, bool* ok = nullptr, int base = 10);
48WTF_EXPORT_PRIVATE unsigned charactersToUIntStrict(const UChar*, size_t, bool* ok = nullptr, int base = 10);
49int64_t charactersToInt64Strict(const LChar*, size_t, bool* ok = nullptr, int base = 10);
50int64_t charactersToInt64Strict(const UChar*, size_t, bool* ok = nullptr, int base = 10);
51WTF_EXPORT_PRIVATE uint64_t charactersToUInt64Strict(const LChar*, size_t, bool* ok = nullptr, int base = 10);
52WTF_EXPORT_PRIVATE uint64_t charactersToUInt64Strict(const UChar*, size_t, bool* ok = nullptr, int base = 10);
53intptr_t charactersToIntPtrStrict(const LChar*, size_t, bool* ok = nullptr, int base = 10);
54intptr_t charactersToIntPtrStrict(const UChar*, size_t, bool* ok = nullptr, int base = 10);
55
56WTF_EXPORT_PRIVATE int charactersToInt(const LChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
57WTF_EXPORT_PRIVATE int charactersToInt(const UChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
58unsigned charactersToUInt(const LChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
59unsigned charactersToUInt(const UChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
60int64_t charactersToInt64(const LChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
61int64_t charactersToInt64(const UChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
62uint64_t charactersToUInt64(const LChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
63WTF_EXPORT_PRIVATE uint64_t charactersToUInt64(const UChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
64intptr_t charactersToIntPtr(const LChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
65intptr_t charactersToIntPtr(const UChar*, size_t, bool* ok = nullptr); // ignores trailing garbage
66
67// FIXME: Like the strict functions above, these give false for "ok" when there is trailing garbage.
68// Like the non-strict functions above, these return the value when there is trailing garbage.
69// It would be better if these were more consistent with the above functions instead.
70WTF_EXPORT_PRIVATE double charactersToDouble(const LChar*, size_t, bool* ok = nullptr);
71WTF_EXPORT_PRIVATE double charactersToDouble(const UChar*, size_t, bool* ok = nullptr);
72WTF_EXPORT_PRIVATE float charactersToFloat(const LChar*, size_t, bool* ok = nullptr);
73WTF_EXPORT_PRIVATE float charactersToFloat(const UChar*, size_t, bool* ok = nullptr);
74WTF_EXPORT_PRIVATE float charactersToFloat(const LChar*, size_t, size_t& parsedLength);
75WTF_EXPORT_PRIVATE float charactersToFloat(const UChar*, size_t, size_t& parsedLength);
76
77template<bool isSpecialCharacter(UChar), typename CharacterType> bool isAllSpecialCharacters(const CharacterType*, size_t);
78
79enum TrailingZerosTruncatingPolicy { KeepTrailingZeros, TruncateTrailingZeros };
80
81class String {
82public:
83 // Construct a null string, distinguishable from an empty string.
84 String() = default;
85
86 // Construct a string with UTF-16 data.
87 WTF_EXPORT_PRIVATE String(const UChar* characters, unsigned length);
88
89 // Construct a string by copying the contents of a vector. To avoid
90 // copying, consider using String::adopt instead.
91 // This method will never create a null string. Vectors with size() == 0
92 // will return the empty string.
93 // NOTE: This is different from String(vector.data(), vector.size())
94 // which will sometimes return a null string when vector.data() is null
95 // which can only occur for vectors without inline capacity.
96 // See: https://bugs.webkit.org/show_bug.cgi?id=109792
97 template<size_t inlineCapacity, typename OverflowHandler>
98 explicit String(const Vector<UChar, inlineCapacity, OverflowHandler>&);
99
100 // Construct a string with UTF-16 data, from a null-terminated source.
101 WTF_EXPORT_PRIVATE String(const UChar*);
102
103 // Construct a string with latin1 data.
104 WTF_EXPORT_PRIVATE String(const LChar* characters, unsigned length);
105 WTF_EXPORT_PRIVATE String(const char* characters, unsigned length);
106
107 // Construct a string with latin1 data, from a null-terminated source.
108 WTF_EXPORT_PRIVATE String(const LChar* characters);
109 WTF_EXPORT_PRIVATE String(const char* characters);
110
111 // Construct a string referencing an existing StringImpl.
112 String(StringImpl&);
113 String(StringImpl*);
114 String(Ref<StringImpl>&&);
115 String(RefPtr<StringImpl>&&);
116
117 String(Ref<AtomicStringImpl>&&);
118 String(RefPtr<AtomicStringImpl>&&);
119
120 String(StaticStringImpl&);
121 String(StaticStringImpl*);
122
123 // Construct a string from a constant string literal.
124 WTF_EXPORT_PRIVATE String(ASCIILiteral);
125
126 // Construct a string from a constant string literal.
127 // This constructor is the "big" version, as it put the length in the function call and generate bigger code.
128 enum ConstructFromLiteralTag { ConstructFromLiteral };
129 template<unsigned characterCount> String(const char (&characters)[characterCount], ConstructFromLiteralTag) : m_impl(StringImpl::createFromLiteral<characterCount>(characters)) { }
130
131 // FIXME: Why do we have to define these explicitly given that we just want the default versions?
132 // We have verified empirically that we do.
133 String(const String&) = default;
134 String(String&&) = default;
135 String& operator=(const String&) = default;
136 String& operator=(String&&) = default;
137
138 ALWAYS_INLINE ~String() = default;
139
140 void swap(String& o) { m_impl.swap(o.m_impl); }
141
142 static String adopt(StringBuffer<LChar>&& buffer) { return StringImpl::adopt(WTFMove(buffer)); }
143 static String adopt(StringBuffer<UChar>&& buffer) { return StringImpl::adopt(WTFMove(buffer)); }
144 template<typename CharacterType, size_t inlineCapacity, typename OverflowHandler, size_t minCapacity>
145 static String adopt(Vector<CharacterType, inlineCapacity, OverflowHandler, minCapacity>&& vector) { return StringImpl::adopt(WTFMove(vector)); }
146
147 bool isNull() const { return !m_impl; }
148 bool isEmpty() const { return !m_impl || m_impl->isEmpty(); }
149
150 StringImpl* impl() const { return m_impl.get(); }
151 RefPtr<StringImpl> releaseImpl() { return WTFMove(m_impl); }
152
153 unsigned length() const { return m_impl ? m_impl->length() : 0; }
154 const LChar* characters8() const { return m_impl ? m_impl->characters8() : nullptr; }
155 const UChar* characters16() const { return m_impl ? m_impl->characters16() : nullptr; }
156
157 // Return characters8() or characters16() depending on CharacterType.
158 template<typename CharacterType> const CharacterType* characters() const;
159
160 bool is8Bit() const { return !m_impl || m_impl->is8Bit(); }
161
162 unsigned sizeInBytes() const { return m_impl ? m_impl->length() * (is8Bit() ? sizeof(LChar) : sizeof(UChar)) : 0; }
163
164 WTF_EXPORT_PRIVATE CString ascii() const;
165 WTF_EXPORT_PRIVATE CString latin1() const;
166
167 WTF_EXPORT_PRIVATE CString utf8(ConversionMode) const;
168 WTF_EXPORT_PRIVATE CString utf8() const;
169
170 WTF_EXPORT_PRIVATE Expected<CString, UTF8ConversionError> tryGetUtf8(ConversionMode) const;
171 WTF_EXPORT_PRIVATE Expected<CString, UTF8ConversionError> tryGetUtf8() const;
172
173 UChar characterAt(unsigned index) const;
174 UChar operator[](unsigned index) const { return characterAt(index); }
175
176 WTF_EXPORT_PRIVATE static String number(int);
177 WTF_EXPORT_PRIVATE static String number(unsigned);
178 WTF_EXPORT_PRIVATE static String number(long);
179 WTF_EXPORT_PRIVATE static String number(unsigned long);
180 WTF_EXPORT_PRIVATE static String number(long long);
181 WTF_EXPORT_PRIVATE static String number(unsigned long long);
182 // FIXME: Change to call numberToStringShortest.
183 static String number(float) = delete;
184 static String number(double) = delete;
185
186 WTF_EXPORT_PRIVATE static String numberToStringShortest(float);
187 WTF_EXPORT_PRIVATE static String numberToStringShortest(double);
188 WTF_EXPORT_PRIVATE static String numberToStringFixedPrecision(float, unsigned precision = 6, TrailingZerosTruncatingPolicy = TruncateTrailingZeros);
189 WTF_EXPORT_PRIVATE static String numberToStringFixedPrecision(double, unsigned precision = 6, TrailingZerosTruncatingPolicy = TruncateTrailingZeros);
190 WTF_EXPORT_PRIVATE static String numberToStringFixedWidth(float, unsigned decimalPlaces);
191 WTF_EXPORT_PRIVATE static String numberToStringFixedWidth(double, unsigned decimalPlaces);
192
193 // FIXME: Delete in favor of the name numberToStringShortest or just number.
194 static String numberToStringECMAScript(float) = delete;
195 static String numberToStringECMAScript(double);
196
197 // Find a single character or string, also with match function & latin1 forms.
198 size_t find(UChar character, unsigned start = 0) const { return m_impl ? m_impl->find(character, start) : notFound; }
199
200 size_t find(const String& string) const { return m_impl ? m_impl->find(string.impl()) : notFound; }
201 size_t find(const String& string, unsigned start) const { return m_impl ? m_impl->find(string.impl(), start) : notFound; }
202 size_t findIgnoringASCIICase(const String& string) const { return m_impl ? m_impl->findIgnoringASCIICase(string.impl()) : notFound; }
203 size_t findIgnoringASCIICase(const String& string, unsigned startOffset) const { return m_impl ? m_impl->findIgnoringASCIICase(string.impl(), startOffset) : notFound; }
204
205 size_t find(CodeUnitMatchFunction matchFunction, unsigned start = 0) const { return m_impl ? m_impl->find(matchFunction, start) : notFound; }
206 size_t find(const LChar* string, unsigned start = 0) const { return m_impl ? m_impl->find(string, start) : notFound; }
207
208 // Find the last instance of a single character or string.
209 size_t reverseFind(UChar character, unsigned start = MaxLength) const { return m_impl ? m_impl->reverseFind(character, start) : notFound; }
210 size_t reverseFind(const String& string, unsigned start = MaxLength) const { return m_impl ? m_impl->reverseFind(string.impl(), start) : notFound; }
211
212 WTF_EXPORT_PRIVATE Vector<UChar> charactersWithNullTermination() const;
213
214 WTF_EXPORT_PRIVATE UChar32 characterStartingAt(unsigned) const;
215
216 bool contains(UChar character) const { return find(character) != notFound; }
217 bool contains(const LChar* string) const { return find(string) != notFound; }
218 bool contains(const String& string) const { return find(string) != notFound; }
219 bool containsIgnoringASCIICase(const String& string) const { return findIgnoringASCIICase(string) != notFound; }
220 bool containsIgnoringASCIICase(const String& string, unsigned startOffset) const { return findIgnoringASCIICase(string, startOffset) != notFound; }
221
222 bool startsWith(const String& string) const { return m_impl ? m_impl->startsWith(string.impl()) : string.isEmpty(); }
223 bool startsWithIgnoringASCIICase(const String& string) const { return m_impl ? m_impl->startsWithIgnoringASCIICase(string.impl()) : string.isEmpty(); }
224 bool startsWith(UChar character) const { return m_impl && m_impl->startsWith(character); }
225 template<unsigned matchLength> bool startsWith(const char (&prefix)[matchLength]) const { return m_impl ? m_impl->startsWith<matchLength>(prefix) : !matchLength; }
226 bool hasInfixStartingAt(const String& prefix, unsigned startOffset) const { return m_impl && prefix.impl() && m_impl->hasInfixStartingAt(*prefix.impl(), startOffset); }
227
228 bool endsWith(const String& string) const { return m_impl ? m_impl->endsWith(string.impl()) : string.isEmpty(); }
229 bool endsWithIgnoringASCIICase(const String& string) const { return m_impl ? m_impl->endsWithIgnoringASCIICase(string.impl()) : string.isEmpty(); }
230 bool endsWith(UChar character) const { return m_impl && m_impl->endsWith(character); }
231 bool endsWith(char character) const { return endsWith(static_cast<UChar>(character)); }
232 template<unsigned matchLength> bool endsWith(const char (&prefix)[matchLength]) const { return m_impl ? m_impl->endsWith<matchLength>(prefix) : !matchLength; }
233 bool hasInfixEndingAt(const String& suffix, unsigned endOffset) const { return m_impl && suffix.impl() && m_impl->hasInfixEndingAt(*suffix.impl(), endOffset); }
234
235 WTF_EXPORT_PRIVATE void append(const String&);
236 WTF_EXPORT_PRIVATE void append(LChar);
237 void append(char character) { append(static_cast<LChar>(character)); };
238 WTF_EXPORT_PRIVATE void append(UChar);
239 WTF_EXPORT_PRIVATE void append(const LChar*, unsigned length);
240 WTF_EXPORT_PRIVATE void append(const UChar*, unsigned length);
241 WTF_EXPORT_PRIVATE void insert(const String&, unsigned position);
242
243 String& replace(UChar target, UChar replacement);
244 String& replace(UChar target, const String& replacement);
245 String& replace(const String& target, const String& replacement);
246 String& replace(unsigned start, unsigned length, const String& replacement);
247 template<unsigned characterCount> String& replaceWithLiteral(UChar target, const char (&replacement)[characterCount]);
248
249 WTF_EXPORT_PRIVATE void truncate(unsigned length);
250 WTF_EXPORT_PRIVATE void remove(unsigned position, unsigned length = 1);
251
252 WTF_EXPORT_PRIVATE String substring(unsigned position, unsigned length = MaxLength) const;
253 WTF_EXPORT_PRIVATE String substringSharingImpl(unsigned position, unsigned length = MaxLength) const;
254 String left(unsigned length) const { return substring(0, length); }
255 String right(unsigned length) const { return substring(this->length() - length, length); }
256
257 WTF_EXPORT_PRIVATE String convertToASCIILowercase() const;
258 WTF_EXPORT_PRIVATE String convertToASCIIUppercase() const;
259 WTF_EXPORT_PRIVATE String convertToLowercaseWithoutLocale() const;
260 WTF_EXPORT_PRIVATE String convertToLowercaseWithoutLocaleStartingAtFailingIndex8Bit(unsigned) const;
261 WTF_EXPORT_PRIVATE String convertToUppercaseWithoutLocale() const;
262 WTF_EXPORT_PRIVATE String convertToLowercaseWithLocale(const AtomicString& localeIdentifier) const;
263 WTF_EXPORT_PRIVATE String convertToUppercaseWithLocale(const AtomicString& localeIdentifier) const;
264
265 WTF_EXPORT_PRIVATE String stripWhiteSpace() const;
266 WTF_EXPORT_PRIVATE String simplifyWhiteSpace() const;
267 WTF_EXPORT_PRIVATE String simplifyWhiteSpace(CodeUnitMatchFunction) const;
268
269 WTF_EXPORT_PRIVATE String stripLeadingAndTrailingCharacters(CodeUnitMatchFunction) const;
270 WTF_EXPORT_PRIVATE String removeCharacters(CodeUnitMatchFunction) const;
271
272 // Returns the string with case folded for case insensitive comparison.
273 // Use convertToASCIILowercase instead if ASCII case insensitive comparison is desired.
274 WTF_EXPORT_PRIVATE String foldCase() const;
275
276 // Returns an uninitialized string. The characters needs to be written
277 // into the buffer returned in data before the returned string is used.
278 static String createUninitialized(unsigned length, UChar*& data) { return StringImpl::createUninitialized(length, data); }
279 static String createUninitialized(unsigned length, LChar*& data) { return StringImpl::createUninitialized(length, data); }
280
281 using SplitFunctor = WTF::Function<void(const StringView&)>;
282
283 WTF_EXPORT_PRIVATE void split(UChar separator, const SplitFunctor&) const;
284 WTF_EXPORT_PRIVATE Vector<String> split(UChar separator) const;
285 WTF_EXPORT_PRIVATE Vector<String> split(const String& separator) const;
286
287 WTF_EXPORT_PRIVATE void splitAllowingEmptyEntries(UChar separator, const SplitFunctor&) const;
288 WTF_EXPORT_PRIVATE Vector<String> splitAllowingEmptyEntries(UChar separator) const;
289 WTF_EXPORT_PRIVATE Vector<String> splitAllowingEmptyEntries(const String& separator) const;
290
291 WTF_EXPORT_PRIVATE int toIntStrict(bool* ok = nullptr, int base = 10) const;
292 WTF_EXPORT_PRIVATE unsigned toUIntStrict(bool* ok = nullptr, int base = 10) const;
293 WTF_EXPORT_PRIVATE int64_t toInt64Strict(bool* ok = nullptr, int base = 10) const;
294 WTF_EXPORT_PRIVATE uint64_t toUInt64Strict(bool* ok = nullptr, int base = 10) const;
295 WTF_EXPORT_PRIVATE intptr_t toIntPtrStrict(bool* ok = nullptr, int base = 10) const;
296
297 WTF_EXPORT_PRIVATE int toInt(bool* ok = nullptr) const;
298 WTF_EXPORT_PRIVATE unsigned toUInt(bool* ok = nullptr) const;
299 WTF_EXPORT_PRIVATE int64_t toInt64(bool* ok = nullptr) const;
300 WTF_EXPORT_PRIVATE uint64_t toUInt64(bool* ok = nullptr) const;
301 WTF_EXPORT_PRIVATE intptr_t toIntPtr(bool* ok = nullptr) const;
302
303 // FIXME: Like the strict functions above, these give false for "ok" when there is trailing garbage.
304 // Like the non-strict functions above, these return the value when there is trailing garbage.
305 // It would be better if these were more consistent with the above functions instead.
306 WTF_EXPORT_PRIVATE double toDouble(bool* ok = nullptr) const;
307 WTF_EXPORT_PRIVATE float toFloat(bool* ok = nullptr) const;
308
309 bool percentage(int& percentage) const;
310
311 WTF_EXPORT_PRIVATE String isolatedCopy() const &;
312 WTF_EXPORT_PRIVATE String isolatedCopy() &&;
313
314 WTF_EXPORT_PRIVATE bool isSafeToSendToAnotherThread() const;
315
316 // Prevent Strings from being implicitly convertable to bool as it will be ambiguous on any platform that
317 // allows implicit conversion to another pointer type (e.g., Mac allows implicit conversion to NSString *).
318 typedef struct ImplicitConversionFromWTFStringToBoolDisallowedA* (String::*UnspecifiedBoolTypeA);
319 typedef struct ImplicitConversionFromWTFStringToBoolDisallowedB* (String::*UnspecifiedBoolTypeB);
320 operator UnspecifiedBoolTypeA() const;
321 operator UnspecifiedBoolTypeB() const;
322
323#if USE(CF)
324 WTF_EXPORT_PRIVATE String(CFStringRef);
325 WTF_EXPORT_PRIVATE RetainPtr<CFStringRef> createCFString() const;
326#endif
327
328#ifdef __OBJC__
329 WTF_EXPORT_PRIVATE String(NSString *);
330
331 // This conversion converts the null string to an empty NSString rather than to nil.
332 // Given Cocoa idioms, this is a more useful default. Clients that need to preserve the
333 // null string can check isNull explicitly.
334 operator NSString *() const;
335#endif
336
337#if OS(WINDOWS)
338#if U_ICU_VERSION_MAJOR_NUM >= 59
339 String(const wchar_t* characters, unsigned length)
340 : String(ucharFrom(characters), length) { }
341
342 String(const wchar_t* characters)
343 : String(ucharFrom(characters)) { }
344#endif
345
346 WTF_EXPORT_PRIVATE Vector<wchar_t> wideCharacters() const;
347#endif
348
349 WTF_EXPORT_PRIVATE static String make8BitFrom16BitSource(const UChar*, size_t);
350 template<size_t inlineCapacity> static String make8BitFrom16BitSource(const Vector<UChar, inlineCapacity>&);
351
352 WTF_EXPORT_PRIVATE static String make16BitFrom8BitSource(const LChar*, size_t);
353
354 // String::fromUTF8 will return a null string if
355 // the input data contains invalid UTF-8 sequences.
356 WTF_EXPORT_PRIVATE static String fromUTF8(const LChar*, size_t);
357 WTF_EXPORT_PRIVATE static String fromUTF8(const LChar*);
358 static String fromUTF8(const char* characters, size_t length) { return fromUTF8(reinterpret_cast<const LChar*>(characters), length); };
359 static String fromUTF8(const char* string) { return fromUTF8(reinterpret_cast<const LChar*>(string)); };
360 WTF_EXPORT_PRIVATE static String fromUTF8(const CString&);
361 static String fromUTF8(const Vector<LChar>& characters);
362
363 // Tries to convert the passed in string to UTF-8, but will fall back to Latin-1 if the string is not valid UTF-8.
364 WTF_EXPORT_PRIVATE static String fromUTF8WithLatin1Fallback(const LChar*, size_t);
365 static String fromUTF8WithLatin1Fallback(const char* characters, size_t length) { return fromUTF8WithLatin1Fallback(reinterpret_cast<const LChar*>(characters), length); };
366
367 // Determines the writing direction using the Unicode Bidi Algorithm rules P2 and P3.
368 UCharDirection defaultWritingDirection(bool* hasStrongDirectionality = nullptr) const;
369
370 bool isAllASCII() const { return !m_impl || m_impl->isAllASCII(); }
371 bool isAllLatin1() const { return !m_impl || m_impl->isAllLatin1(); }
372 template<bool isSpecialCharacter(UChar)> bool isAllSpecialCharacters() const { return !m_impl || m_impl->isAllSpecialCharacters<isSpecialCharacter>(); }
373
374 // Hash table deleted values, which are only constructed and never copied or destroyed.
375 String(WTF::HashTableDeletedValueType) : m_impl(WTF::HashTableDeletedValue) { }
376 bool isHashTableDeletedValue() const { return m_impl.isHashTableDeletedValue(); }
377
378 unsigned hash() const { return isNull() ? 0 : impl()->hash(); }
379 unsigned existingHash() const { return isNull() ? 0 : impl()->existingHash(); }
380
381#ifndef NDEBUG
382 WTF_EXPORT_PRIVATE void show() const;
383#endif
384
385 // Turns this String empty if the StringImpl is not referenced by anyone else.
386 // This is useful for clearing String-based caches.
387 void clearImplIfNotShared();
388
389 static constexpr unsigned MaxLength = StringImpl::MaxLength;
390
391private:
392 template<typename CharacterType> void removeInternal(const CharacterType*, unsigned, unsigned);
393
394 template<bool allowEmptyEntries> void splitInternal(UChar separator, const SplitFunctor&) const;
395 template<bool allowEmptyEntries> Vector<String> splitInternal(UChar separator) const;
396 template<bool allowEmptyEntries> Vector<String> splitInternal(const String& separator) const;
397
398 RefPtr<StringImpl> m_impl;
399};
400
401static_assert(sizeof(String) == sizeof(void*), "String should effectively be a pointer to a StringImpl, and efficient to pass by value");
402
403inline bool operator==(const String& a, const String& b) { return equal(a.impl(), b.impl()); }
404inline bool operator==(const String& a, const LChar* b) { return equal(a.impl(), b); }
405inline bool operator==(const String& a, const char* b) { return equal(a.impl(), reinterpret_cast<const LChar*>(b)); }
406inline bool operator==(const String& a, ASCIILiteral b) { return equal(a.impl(), reinterpret_cast<const LChar*>(b.characters())); }
407inline bool operator==(const LChar* a, const String& b) { return equal(a, b.impl()); }
408inline bool operator==(const char* a, const String& b) { return equal(reinterpret_cast<const LChar*>(a), b.impl()); }
409inline bool operator==(ASCIILiteral a, const String& b) { return equal(reinterpret_cast<const LChar*>(a.characters()), b.impl()); }
410template<size_t inlineCapacity> inline bool operator==(const Vector<char, inlineCapacity>& a, const String& b) { return equal(b.impl(), a.data(), a.size()); }
411template<size_t inlineCapacity> inline bool operator==(const String& a, const Vector<char, inlineCapacity>& b) { return b == a; }
412
413inline bool operator!=(const String& a, const String& b) { return !equal(a.impl(), b.impl()); }
414inline bool operator!=(const String& a, const LChar* b) { return !equal(a.impl(), b); }
415inline bool operator!=(const String& a, const char* b) { return !equal(a.impl(), reinterpret_cast<const LChar*>(b)); }
416inline bool operator!=(const String& a, ASCIILiteral b) { return !equal(a.impl(), reinterpret_cast<const LChar*>(b.characters())); }
417inline bool operator!=(const LChar* a, const String& b) { return !equal(a, b.impl()); }
418inline bool operator!=(const char* a, const String& b) { return !equal(reinterpret_cast<const LChar*>(a), b.impl()); }
419inline bool operator!=(ASCIILiteral a, const String& b) { return !equal(reinterpret_cast<const LChar*>(a.characters()), b.impl()); }
420template<size_t inlineCapacity> inline bool operator!=(const Vector<char, inlineCapacity>& a, const String& b) { return !(a == b); }
421template<size_t inlineCapacity> inline bool operator!=(const String& a, const Vector<char, inlineCapacity>& b) { return b != a; }
422
423bool equalIgnoringASCIICase(const String&, const String&);
424bool equalIgnoringASCIICase(const String&, const char*);
425
426template<unsigned length> bool equalLettersIgnoringASCIICase(const String&, const char (&lowercaseLetters)[length]);
427template<unsigned length> bool startsWithLettersIgnoringASCIICase(const String&, const char (&lowercaseLetters)[length]);
428
429inline bool equalIgnoringNullity(const String& a, const String& b) { return equalIgnoringNullity(a.impl(), b.impl()); }
430template<size_t inlineCapacity> inline bool equalIgnoringNullity(const Vector<UChar, inlineCapacity>& a, const String& b) { return equalIgnoringNullity(a, b.impl()); }
431
432inline bool operator!(const String& string) { return string.isNull(); }
433
434inline void swap(String& a, String& b) { a.swap(b); }
435
436#ifdef __OBJC__
437
438// Used in a small number of places where the long standing behavior has been "nil if empty".
439NSString * nsStringNilIfEmpty(const String&);
440
441#endif
442
443WTF_EXPORT_PRIVATE int codePointCompare(const String&, const String&);
444bool codePointCompareLessThan(const String&, const String&);
445
446template<typename CharacterType> void appendNumber(Vector<CharacterType>&, unsigned char number);
447
448// Shared global empty and null string.
449WTF_EXPORT_PRIVATE const String& emptyString();
450WTF_EXPORT_PRIVATE const String& nullString();
451
452template<typename> struct DefaultHash;
453template<> struct DefaultHash<String> { using Hash = StringHash; };
454template<> struct VectorTraits<String> : VectorTraitsBase<false, void> {
455 static const bool canInitializeWithMemset = true;
456 static const bool canMoveWithMemcpy = true;
457};
458
459template<> struct IntegerToStringConversionTrait<String> {
460 using ReturnType = String;
461 using AdditionalArgumentType = void;
462 static String flush(LChar* characters, unsigned length, void*) { return { characters, length }; }
463};
464
465// Definitions of string operations
466
467inline String::String(StringImpl& string)
468 : m_impl(&string)
469{
470}
471
472inline String::String(StringImpl* string)
473 : m_impl(string)
474{
475}
476
477inline String::String(Ref<StringImpl>&& string)
478 : m_impl(WTFMove(string))
479{
480}
481
482inline String::String(RefPtr<StringImpl>&& string)
483 : m_impl(WTFMove(string))
484{
485}
486
487inline String::String(Ref<AtomicStringImpl>&& string)
488 : m_impl(WTFMove(string))
489{
490}
491
492inline String::String(RefPtr<AtomicStringImpl>&& string)
493 : m_impl(WTFMove(string))
494{
495}
496
497inline String::String(StaticStringImpl& string)
498 : m_impl(reinterpret_cast<StringImpl*>(&string))
499{
500}
501
502inline String::String(StaticStringImpl* string)
503 : m_impl(reinterpret_cast<StringImpl*>(string))
504{
505}
506
507template<size_t inlineCapacity, typename OverflowHandler> String::String(const Vector<UChar, inlineCapacity, OverflowHandler>& vector)
508 : m_impl(vector.size() ? StringImpl::create(vector.data(), vector.size()) : Ref<StringImpl> { *StringImpl::empty() })
509{
510}
511
512template<> inline const LChar* String::characters<LChar>() const
513{
514 return characters8();
515}
516
517template<> inline const UChar* String::characters<UChar>() const
518{
519 return characters16();
520}
521
522inline UChar String::characterAt(unsigned index) const
523{
524 if (!m_impl || index >= m_impl->length())
525 return 0;
526 return (*m_impl)[index];
527}
528
529inline String& String::replace(UChar target, UChar replacement)
530{
531 if (m_impl)
532 m_impl = m_impl->replace(target, replacement);
533 return *this;
534}
535
536inline String& String::replace(UChar target, const String& replacement)
537{
538 if (m_impl)
539 m_impl = m_impl->replace(target, replacement.impl());
540 return *this;
541}
542
543inline String& String::replace(const String& target, const String& replacement)
544{
545 if (m_impl)
546 m_impl = m_impl->replace(target.impl(), replacement.impl());
547 return *this;
548}
549
550inline String& String::replace(unsigned start, unsigned length, const String& replacement)
551{
552 if (m_impl)
553 m_impl = m_impl->replace(start, length, replacement.impl());
554 return *this;
555}
556
557template<unsigned characterCount> ALWAYS_INLINE String& String::replaceWithLiteral(UChar target, const char (&characters)[characterCount])
558{
559 if (m_impl)
560 m_impl = m_impl->replace(target, characters, characterCount - 1);
561 return *this;
562}
563
564template<size_t inlineCapacity> inline String String::make8BitFrom16BitSource(const Vector<UChar, inlineCapacity>& buffer)
565{
566 return make8BitFrom16BitSource(buffer.data(), buffer.size());
567}
568
569inline UCharDirection String::defaultWritingDirection(bool* hasStrongDirectionality) const
570{
571 if (m_impl)
572 return m_impl->defaultWritingDirection(hasStrongDirectionality);
573 if (hasStrongDirectionality)
574 *hasStrongDirectionality = false;
575 return U_LEFT_TO_RIGHT;
576}
577
578inline void String::clearImplIfNotShared()
579{
580 if (m_impl && m_impl->hasOneRef())
581 m_impl = nullptr;
582}
583
584#ifdef __OBJC__
585
586inline String::operator NSString *() const
587{
588 if (!m_impl)
589 return @"";
590 return *m_impl;
591}
592
593inline NSString * nsStringNilIfEmpty(const String& string)
594{
595 if (string.isEmpty())
596 return nil;
597 return *string.impl();
598}
599
600#endif
601
602inline bool codePointCompareLessThan(const String& a, const String& b)
603{
604 return codePointCompare(a.impl(), b.impl()) < 0;
605}
606
607template<typename CharacterType>
608inline void appendNumber(Vector<CharacterType>& vector, unsigned char number)
609{
610 int numberLength = number > 99 ? 3 : (number > 9 ? 2 : 1);
611 size_t vectorSize = vector.size();
612 vector.grow(vectorSize + numberLength);
613
614 switch (numberLength) {
615 case 3:
616 vector[vectorSize + 2] = number % 10 + '0';
617 number /= 10;
618 FALLTHROUGH;
619
620 case 2:
621 vector[vectorSize + 1] = number % 10 + '0';
622 number /= 10;
623 FALLTHROUGH;
624
625 case 1:
626 vector[vectorSize] = number % 10 + '0';
627 }
628}
629
630inline String String::fromUTF8(const Vector<LChar>& characters)
631{
632 if (characters.isEmpty())
633 return emptyString();
634 return fromUTF8(characters.data(), characters.size());
635}
636
637template<unsigned length> inline bool equalLettersIgnoringASCIICase(const String& string, const char (&lowercaseLetters)[length])
638{
639 return equalLettersIgnoringASCIICase(string.impl(), lowercaseLetters);
640}
641
642inline bool equalIgnoringASCIICase(const String& a, const String& b)
643{
644 return equalIgnoringASCIICase(a.impl(), b.impl());
645}
646
647inline bool equalIgnoringASCIICase(const String& a, const char* b)
648{
649 return equalIgnoringASCIICase(a.impl(), b);
650}
651
652template<unsigned length> inline bool startsWithLettersIgnoringASCIICase(const String& string, const char (&lowercaseLetters)[length])
653{
654 return startsWithLettersIgnoringASCIICase(string.impl(), lowercaseLetters);
655}
656
657inline String String::numberToStringECMAScript(double number)
658{
659 return numberToStringShortest(number);
660}
661
662inline namespace StringLiterals {
663
664inline String operator"" _str(const char* characters, size_t)
665{
666 return ASCIILiteral::fromLiteralUnsafe(characters);
667}
668
669} // inline StringLiterals
670
671} // namespace WTF
672
673using WTF::KeepTrailingZeros;
674using WTF::String;
675using WTF::appendNumber;
676using WTF::charactersToDouble;
677using WTF::charactersToFloat;
678using WTF::charactersToInt64;
679using WTF::charactersToInt64Strict;
680using WTF::charactersToInt;
681using WTF::charactersToIntPtr;
682using WTF::charactersToIntPtrStrict;
683using WTF::charactersToIntStrict;
684using WTF::charactersToUInt64;
685using WTF::charactersToUInt64Strict;
686using WTF::charactersToUInt;
687using WTF::charactersToUIntStrict;
688using WTF::emptyString;
689using WTF::nullString;
690using WTF::equal;
691using WTF::find;
692using WTF::isAllSpecialCharacters;
693using WTF::isSpaceOrNewline;
694using WTF::reverseFind;
695
696#include <wtf/text/AtomicString.h>
697