1 | /* |
2 | * Copyright (C) 2004-2019 Apple Inc. All rights reserved. |
3 | * |
4 | * This library is free software; you can redistribute it and/or |
5 | * modify it under the terms of the GNU Library General Public |
6 | * License as published by the Free Software Foundation; either |
7 | * version 2 of the License, or (at your option) any later version. |
8 | * |
9 | * This library is distributed in the hope that it will be useful, |
10 | * but WITHOUT ANY WARRANTY; without even the implied warranty of |
11 | * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU |
12 | * Library General Public License for more details. |
13 | * |
14 | * You should have received a copy of the GNU Library General Public License |
15 | * along with this library; see the file COPYING.LIB. If not, write to |
16 | * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor, |
17 | * Boston, MA 02110-1301, USA. |
18 | * |
19 | */ |
20 | |
21 | #pragma once |
22 | |
23 | #include <utility> |
24 | #include <wtf/NeverDestroyed.h> |
25 | #include <wtf/text/AtomStringImpl.h> |
26 | #include <wtf/text/IntegerToStringConversion.h> |
27 | #include <wtf/text/WTFString.h> |
28 | |
29 | #if OS(WINDOWS) |
30 | #include <wtf/text/win/WCharStringExtras.h> |
31 | #endif |
32 | |
33 | namespace WTF { |
34 | |
35 | struct AtomStringHash; |
36 | |
37 | class AtomString { |
38 | public: |
39 | WTF_EXPORT_PRIVATE static void init(); |
40 | |
41 | AtomString(); |
42 | AtomString(const LChar*); |
43 | AtomString(const char*); |
44 | AtomString(const LChar*, unsigned length); |
45 | AtomString(const UChar*, unsigned length); |
46 | AtomString(const UChar*); |
47 | |
48 | template<size_t inlineCapacity> |
49 | explicit AtomString(const Vector<UChar, inlineCapacity>& characters) |
50 | : m_string(AtomStringImpl::add(characters.data(), characters.size())) |
51 | { |
52 | } |
53 | |
54 | AtomString(AtomStringImpl*); |
55 | AtomString(RefPtr<AtomStringImpl>&&); |
56 | AtomString(const StaticStringImpl*); |
57 | AtomString(StringImpl*); |
58 | AtomString(const String&); |
59 | AtomString(StringImpl* baseString, unsigned start, unsigned length); |
60 | |
61 | // FIXME: AtomString doesn’t always have AtomStringImpl, so one of those two names needs to change. |
62 | AtomString(UniquedStringImpl* uid); |
63 | |
64 | enum ConstructFromLiteralTag { ConstructFromLiteral }; |
65 | AtomString(const char* characters, unsigned length, ConstructFromLiteralTag) |
66 | : m_string(AtomStringImpl::addLiteral(characters, length)) |
67 | { |
68 | } |
69 | |
70 | template<unsigned characterCount> ALWAYS_INLINE AtomString(const char (&characters)[characterCount], ConstructFromLiteralTag) |
71 | : m_string(AtomStringImpl::addLiteral(characters, characterCount - 1)) |
72 | { |
73 | COMPILE_ASSERT(characterCount > 1, AtomStringFromLiteralNotEmpty); |
74 | COMPILE_ASSERT((characterCount - 1 <= ((unsigned(~0) - sizeof(StringImpl)) / sizeof(LChar))), AtomStringFromLiteralCannotOverflow); |
75 | } |
76 | |
77 | // We have to declare the copy constructor and copy assignment operator as well, otherwise |
78 | // they'll be implicitly deleted by adding the move constructor and move assignment operator. |
79 | AtomString(const AtomString& other) : m_string(other.m_string) { } |
80 | AtomString(AtomString&& other) : m_string(WTFMove(other.m_string)) { } |
81 | AtomString& operator=(const AtomString& other) { m_string = other.m_string; return *this; } |
82 | AtomString& operator=(AtomString&& other) { m_string = WTFMove(other.m_string); return *this; } |
83 | |
84 | // Hash table deleted values, which are only constructed and never copied or destroyed. |
85 | AtomString(WTF::HashTableDeletedValueType) : m_string(WTF::HashTableDeletedValue) { } |
86 | bool isHashTableDeletedValue() const { return m_string.isHashTableDeletedValue(); } |
87 | |
88 | unsigned existingHash() const { return isNull() ? 0 : impl()->existingHash(); } |
89 | |
90 | operator const String&() const { return m_string; } |
91 | const String& string() const { return m_string; }; |
92 | |
93 | // FIXME: What guarantees this isn't a SymbolImpl rather than an AtomStringImpl? |
94 | AtomStringImpl* impl() const { return static_cast<AtomStringImpl*>(m_string.impl()); } |
95 | |
96 | bool is8Bit() const { return m_string.is8Bit(); } |
97 | const LChar* characters8() const { return m_string.characters8(); } |
98 | const UChar* characters16() const { return m_string.characters16(); } |
99 | unsigned length() const { return m_string.length(); } |
100 | |
101 | UChar operator[](unsigned int i) const { return m_string[i]; } |
102 | |
103 | WTF_EXPORT_PRIVATE static AtomString number(int); |
104 | WTF_EXPORT_PRIVATE static AtomString number(unsigned); |
105 | WTF_EXPORT_PRIVATE static AtomString number(unsigned long); |
106 | WTF_EXPORT_PRIVATE static AtomString number(unsigned long long); |
107 | WTF_EXPORT_PRIVATE static AtomString number(float); |
108 | WTF_EXPORT_PRIVATE static AtomString number(double); |
109 | // If we need more overloads of the number function, we can add all the others that String has, but these seem to do for now. |
110 | |
111 | bool contains(UChar character) const { return m_string.contains(character); } |
112 | bool contains(const LChar* string) const { return m_string.contains(string); } |
113 | bool contains(const String& string) const { return m_string.contains(string); } |
114 | bool containsIgnoringASCIICase(const String& string) const { return m_string.containsIgnoringASCIICase(string); } |
115 | |
116 | size_t find(UChar character, unsigned start = 0) const { return m_string.find(character, start); } |
117 | size_t find(const LChar* string, unsigned start = 0) const { return m_string.find(string, start); } |
118 | size_t find(const String& string, unsigned start = 0) const { return m_string.find(string, start); } |
119 | size_t findIgnoringASCIICase(const String& string) const { return m_string.findIgnoringASCIICase(string); } |
120 | size_t findIgnoringASCIICase(const String& string, unsigned startOffset) const { return m_string.findIgnoringASCIICase(string, startOffset); } |
121 | size_t find(CodeUnitMatchFunction matchFunction, unsigned start = 0) const { return m_string.find(matchFunction, start); } |
122 | |
123 | bool startsWith(const String& string) const { return m_string.startsWith(string); } |
124 | bool startsWithIgnoringASCIICase(const String& string) const { return m_string.startsWithIgnoringASCIICase(string); } |
125 | bool startsWith(UChar character) const { return m_string.startsWith(character); } |
126 | template<unsigned matchLength> bool startsWith(const char (&prefix)[matchLength]) const { return m_string.startsWith<matchLength>(prefix); } |
127 | |
128 | bool endsWith(const String& string) const { return m_string.endsWith(string); } |
129 | bool endsWithIgnoringASCIICase(const String& string) const { return m_string.endsWithIgnoringASCIICase(string); } |
130 | bool endsWith(UChar character) const { return m_string.endsWith(character); } |
131 | template<unsigned matchLength> bool endsWith(const char (&prefix)[matchLength]) const { return m_string.endsWith<matchLength>(prefix); } |
132 | |
133 | WTF_EXPORT_PRIVATE AtomString convertToASCIILowercase() const; |
134 | WTF_EXPORT_PRIVATE AtomString convertToASCIIUppercase() const; |
135 | |
136 | int toInt(bool* ok = nullptr) const { return m_string.toInt(ok); } |
137 | double toDouble(bool* ok = nullptr) const { return m_string.toDouble(ok); } |
138 | float toFloat(bool* ok = nullptr) const { return m_string.toFloat(ok); } |
139 | bool percentage(int& p) const { return m_string.percentage(p); } |
140 | |
141 | bool isNull() const { return m_string.isNull(); } |
142 | bool isEmpty() const { return m_string.isEmpty(); } |
143 | |
144 | #if USE(CF) |
145 | AtomString(CFStringRef); |
146 | #endif |
147 | |
148 | #ifdef __OBJC__ |
149 | AtomString(NSString *); |
150 | operator NSString *() const { return m_string; } |
151 | #endif |
152 | |
153 | #if OS(WINDOWS) && U_ICU_VERSION_MAJOR_NUM >= 59 |
154 | AtomString(const wchar_t* characters, unsigned length) |
155 | : AtomString(ucharFrom(characters), length) { } |
156 | |
157 | AtomString(const wchar_t* characters) |
158 | : AtomString(ucharFrom(characters)) { } |
159 | #endif |
160 | |
161 | // AtomString::fromUTF8 will return a null string if the input data contains invalid UTF-8 sequences. |
162 | static AtomString fromUTF8(const char*, size_t); |
163 | static AtomString fromUTF8(const char*); |
164 | |
165 | #ifndef NDEBUG |
166 | void show() const; |
167 | #endif |
168 | |
169 | private: |
170 | // The explicit constructors with AtomString::ConstructFromLiteral must be used for literals. |
171 | AtomString(ASCIILiteral); |
172 | |
173 | enum class CaseConvertType { Upper, Lower }; |
174 | template<CaseConvertType> AtomString convertASCIICase() const; |
175 | |
176 | WTF_EXPORT_PRIVATE static AtomString fromUTF8Internal(const char*, const char*); |
177 | |
178 | String m_string; |
179 | }; |
180 | |
181 | static_assert(sizeof(AtomString) == sizeof(String), "AtomString and String must be the same size!" ); |
182 | |
183 | inline bool operator==(const AtomString& a, const AtomString& b) { return a.impl() == b.impl(); } |
184 | bool operator==(const AtomString&, const LChar*); |
185 | inline bool operator==(const AtomString& a, const char* b) { return WTF::equal(a.impl(), reinterpret_cast<const LChar*>(b)); } |
186 | inline bool operator==(const AtomString& a, const Vector<UChar>& b) { return a.impl() && equal(a.impl(), b.data(), b.size()); } |
187 | inline bool operator==(const AtomString& a, const String& b) { return equal(a.impl(), b.impl()); } |
188 | inline bool operator==(const LChar* a, const AtomString& b) { return b == a; } |
189 | inline bool operator==(const String& a, const AtomString& b) { return equal(a.impl(), b.impl()); } |
190 | inline bool operator==(const Vector<UChar>& a, const AtomString& b) { return b == a; } |
191 | |
192 | inline bool operator!=(const AtomString& a, const AtomString& b) { return a.impl() != b.impl(); } |
193 | inline bool operator!=(const AtomString& a, const LChar* b) { return !(a == b); } |
194 | inline bool operator!=(const AtomString& a, const char* b) { return !(a == b); } |
195 | inline bool operator!=(const AtomString& a, const String& b) { return !equal(a.impl(), b.impl()); } |
196 | inline bool operator!=(const AtomString& a, const Vector<UChar>& b) { return !(a == b); } |
197 | inline bool operator!=(const LChar* a, const AtomString& b) { return !(b == a); } |
198 | inline bool operator!=(const String& a, const AtomString& b) { return !equal(a.impl(), b.impl()); } |
199 | inline bool operator!=(const Vector<UChar>& a, const AtomString& b) { return !(a == b); } |
200 | |
201 | bool equalIgnoringASCIICase(const AtomString&, const AtomString&); |
202 | bool equalIgnoringASCIICase(const AtomString&, const String&); |
203 | bool equalIgnoringASCIICase(const String&, const AtomString&); |
204 | bool equalIgnoringASCIICase(const AtomString&, const char*); |
205 | |
206 | template<unsigned length> bool equalLettersIgnoringASCIICase(const AtomString&, const char (&lowercaseLetters)[length]); |
207 | |
208 | inline AtomString::AtomString() |
209 | { |
210 | } |
211 | |
212 | inline AtomString::AtomString(const LChar* string) |
213 | : m_string(AtomStringImpl::add(string)) |
214 | { |
215 | } |
216 | |
217 | inline AtomString::AtomString(const char* string) |
218 | : m_string(AtomStringImpl::add(string)) |
219 | { |
220 | } |
221 | |
222 | inline AtomString::AtomString(const LChar* string, unsigned length) |
223 | : m_string(AtomStringImpl::add(string, length)) |
224 | { |
225 | } |
226 | |
227 | inline AtomString::AtomString(const UChar* string, unsigned length) |
228 | : m_string(AtomStringImpl::add(string, length)) |
229 | { |
230 | } |
231 | |
232 | inline AtomString::AtomString(const UChar* string) |
233 | : m_string(AtomStringImpl::add(string)) |
234 | { |
235 | } |
236 | |
237 | inline AtomString::AtomString(AtomStringImpl* string) |
238 | : m_string(string) |
239 | { |
240 | } |
241 | |
242 | inline AtomString::AtomString(RefPtr<AtomStringImpl>&& string) |
243 | : m_string(WTFMove(string)) |
244 | { |
245 | } |
246 | |
247 | inline AtomString::AtomString(StringImpl* string) |
248 | : m_string(AtomStringImpl::add(string)) |
249 | { |
250 | } |
251 | |
252 | inline AtomString::AtomString(const StaticStringImpl* string) |
253 | : m_string(AtomStringImpl::add(string)) |
254 | { |
255 | } |
256 | |
257 | inline AtomString::AtomString(const String& string) |
258 | : m_string(AtomStringImpl::add(string.impl())) |
259 | { |
260 | } |
261 | |
262 | inline AtomString::AtomString(StringImpl* baseString, unsigned start, unsigned length) |
263 | : m_string(AtomStringImpl::add(baseString, start, length)) |
264 | { |
265 | } |
266 | |
267 | inline AtomString::AtomString(UniquedStringImpl* uid) |
268 | : m_string(uid) |
269 | { |
270 | } |
271 | |
272 | #if USE(CF) |
273 | |
274 | inline AtomString::AtomString(CFStringRef string) |
275 | : m_string(AtomStringImpl::add(string)) |
276 | { |
277 | } |
278 | |
279 | #endif |
280 | |
281 | #ifdef __OBJC__ |
282 | |
283 | inline AtomString::AtomString(NSString *string) |
284 | : m_string(AtomStringImpl::add((__bridge CFStringRef)string)) |
285 | { |
286 | } |
287 | |
288 | #endif |
289 | |
290 | // Define external global variables for the commonly used atom strings. |
291 | // These are only usable from the main thread. |
292 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> nullAtomData; |
293 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> emptyAtomData; |
294 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> starAtomData; |
295 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> xmlAtomData; |
296 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> xmlnsAtomData; |
297 | |
298 | inline const AtomString& nullAtom() { return nullAtomData.get(); } |
299 | inline const AtomString& emptyAtom() { return emptyAtomData.get(); } |
300 | inline const AtomString& starAtom() { return starAtomData.get(); } |
301 | inline const AtomString& xmlAtom() { return xmlAtomData.get(); } |
302 | inline const AtomString& xmlnsAtom() { return xmlnsAtomData.get(); } |
303 | |
304 | inline AtomString AtomString::fromUTF8(const char* characters, size_t length) |
305 | { |
306 | if (!characters) |
307 | return nullAtom(); |
308 | if (!length) |
309 | return emptyAtom(); |
310 | return fromUTF8Internal(characters, characters + length); |
311 | } |
312 | |
313 | inline AtomString AtomString::fromUTF8(const char* characters) |
314 | { |
315 | if (!characters) |
316 | return nullAtom(); |
317 | if (!*characters) |
318 | return emptyAtom(); |
319 | return fromUTF8Internal(characters, nullptr); |
320 | } |
321 | |
322 | // AtomStringHash is the default hash for AtomString |
323 | template<typename T> struct DefaultHash; |
324 | template<> struct DefaultHash<AtomString> { |
325 | typedef AtomStringHash Hash; |
326 | }; |
327 | |
328 | template<unsigned length> inline bool equalLettersIgnoringASCIICase(const AtomString& string, const char (&lowercaseLetters)[length]) |
329 | { |
330 | return equalLettersIgnoringASCIICase(string.string(), lowercaseLetters); |
331 | } |
332 | |
333 | inline bool equalIgnoringASCIICase(const AtomString& a, const AtomString& b) |
334 | { |
335 | return equalIgnoringASCIICase(a.string(), b.string()); |
336 | } |
337 | |
338 | inline bool equalIgnoringASCIICase(const AtomString& a, const String& b) |
339 | { |
340 | return equalIgnoringASCIICase(a.string(), b); |
341 | } |
342 | |
343 | inline bool equalIgnoringASCIICase(const String& a, const AtomString& b) |
344 | { |
345 | return equalIgnoringASCIICase(a, b.string()); |
346 | } |
347 | |
348 | inline bool equalIgnoringASCIICase(const AtomString& a, const char* b) |
349 | { |
350 | return equalIgnoringASCIICase(a.string(), b); |
351 | } |
352 | |
353 | template<> struct IntegerToStringConversionTrait<AtomString> { |
354 | using ReturnType = AtomString; |
355 | using AdditionalArgumentType = void; |
356 | static AtomString flush(LChar* characters, unsigned length, void*) { return { characters, length }; } |
357 | }; |
358 | |
359 | } // namespace WTF |
360 | |
361 | using WTF::AtomString; |
362 | using WTF::nullAtom; |
363 | using WTF::emptyAtom; |
364 | using WTF::starAtom; |
365 | using WTF::xmlAtom; |
366 | using WTF::xmlnsAtom; |
367 | |
368 | #include <wtf/text/StringConcatenate.h> |
369 | |