| 1 | /* |
| 2 | * Copyright (C) 2004-2017 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 | AtomStringImpl* impl() const { return static_cast<AtomStringImpl *>(m_string.impl()); } |
| 94 | |
| 95 | bool is8Bit() const { return m_string.is8Bit(); } |
| 96 | const LChar* characters8() const { return m_string.characters8(); } |
| 97 | const UChar* characters16() const { return m_string.characters16(); } |
| 98 | unsigned length() const { return m_string.length(); } |
| 99 | |
| 100 | UChar operator[](unsigned int i) const { return m_string[i]; } |
| 101 | |
| 102 | WTF_EXPORT_PRIVATE static AtomString number(int); |
| 103 | WTF_EXPORT_PRIVATE static AtomString number(unsigned); |
| 104 | WTF_EXPORT_PRIVATE static AtomString number(unsigned long); |
| 105 | WTF_EXPORT_PRIVATE static AtomString number(unsigned long long); |
| 106 | WTF_EXPORT_PRIVATE static AtomString number(float); |
| 107 | WTF_EXPORT_PRIVATE static AtomString number(double); |
| 108 | // 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. |
| 109 | |
| 110 | bool contains(UChar character) const { return m_string.contains(character); } |
| 111 | bool contains(const LChar* string) const { return m_string.contains(string); } |
| 112 | bool contains(const String& string) const { return m_string.contains(string); } |
| 113 | bool containsIgnoringASCIICase(const String& string) const { return m_string.containsIgnoringASCIICase(string); } |
| 114 | |
| 115 | size_t find(UChar character, unsigned start = 0) const { return m_string.find(character, start); } |
| 116 | size_t find(const LChar* string, unsigned start = 0) const { return m_string.find(string, start); } |
| 117 | size_t find(const String& string, unsigned start = 0) const { return m_string.find(string, start); } |
| 118 | size_t findIgnoringASCIICase(const String& string) const { return m_string.findIgnoringASCIICase(string); } |
| 119 | size_t findIgnoringASCIICase(const String& string, unsigned startOffset) const { return m_string.findIgnoringASCIICase(string, startOffset); } |
| 120 | size_t find(CodeUnitMatchFunction matchFunction, unsigned start = 0) const { return m_string.find(matchFunction, start); } |
| 121 | |
| 122 | bool startsWith(const String& string) const { return m_string.startsWith(string); } |
| 123 | bool startsWithIgnoringASCIICase(const String& string) const { return m_string.startsWithIgnoringASCIICase(string); } |
| 124 | bool startsWith(UChar character) const { return m_string.startsWith(character); } |
| 125 | template<unsigned matchLength> bool startsWith(const char (&prefix)[matchLength]) const { return m_string.startsWith<matchLength>(prefix); } |
| 126 | |
| 127 | bool endsWith(const String& string) const { return m_string.endsWith(string); } |
| 128 | bool endsWithIgnoringASCIICase(const String& string) const { return m_string.endsWithIgnoringASCIICase(string); } |
| 129 | bool endsWith(UChar character) const { return m_string.endsWith(character); } |
| 130 | template<unsigned matchLength> bool endsWith(const char (&prefix)[matchLength]) const { return m_string.endsWith<matchLength>(prefix); } |
| 131 | |
| 132 | WTF_EXPORT_PRIVATE AtomString convertToASCIILowercase() const; |
| 133 | WTF_EXPORT_PRIVATE AtomString convertToASCIIUppercase() const; |
| 134 | |
| 135 | int toInt(bool* ok = 0) const { return m_string.toInt(ok); } |
| 136 | double toDouble(bool* ok = 0) const { return m_string.toDouble(ok); } |
| 137 | float toFloat(bool* ok = 0) const { return m_string.toFloat(ok); } |
| 138 | bool percentage(int& p) const { return m_string.percentage(p); } |
| 139 | |
| 140 | bool isNull() const { return m_string.isNull(); } |
| 141 | bool isEmpty() const { return m_string.isEmpty(); } |
| 142 | |
| 143 | #if USE(CF) |
| 144 | AtomString(CFStringRef); |
| 145 | #endif |
| 146 | |
| 147 | #ifdef __OBJC__ |
| 148 | AtomString(NSString*); |
| 149 | operator NSString*() const { return m_string; } |
| 150 | #endif |
| 151 | |
| 152 | #if OS(WINDOWS) && U_ICU_VERSION_MAJOR_NUM >= 59 |
| 153 | AtomString(const wchar_t* characters, unsigned length) |
| 154 | : AtomString(ucharFrom(characters), length) { } |
| 155 | |
| 156 | AtomString(const wchar_t* characters) |
| 157 | : AtomString(ucharFrom(characters)) { } |
| 158 | #endif |
| 159 | |
| 160 | // AtomString::fromUTF8 will return a null string if the input data contains invalid UTF-8 sequences. |
| 161 | static AtomString fromUTF8(const char*, size_t); |
| 162 | static AtomString fromUTF8(const char*); |
| 163 | |
| 164 | #ifndef NDEBUG |
| 165 | void show() const; |
| 166 | #endif |
| 167 | |
| 168 | private: |
| 169 | // The explicit constructors with AtomString::ConstructFromLiteral must be used for literals. |
| 170 | AtomString(ASCIILiteral); |
| 171 | |
| 172 | enum class CaseConvertType { Upper, Lower }; |
| 173 | template<CaseConvertType> AtomString convertASCIICase() const; |
| 174 | |
| 175 | WTF_EXPORT_PRIVATE static AtomString fromUTF8Internal(const char*, const char*); |
| 176 | |
| 177 | String m_string; |
| 178 | }; |
| 179 | |
| 180 | using AtomicString = AtomString; |
| 181 | |
| 182 | static_assert(sizeof(AtomString) == sizeof(String), "AtomString and String must be same size!" ); |
| 183 | |
| 184 | inline bool operator==(const AtomString& a, const AtomString& b) { return a.impl() == b.impl(); } |
| 185 | bool operator==(const AtomString&, const LChar*); |
| 186 | inline bool operator==(const AtomString& a, const char* b) { return WTF::equal(a.impl(), reinterpret_cast<const LChar*>(b)); } |
| 187 | inline bool operator==(const AtomString& a, const Vector<UChar>& b) { return a.impl() && equal(a.impl(), b.data(), b.size()); } |
| 188 | inline bool operator==(const AtomString& a, const String& b) { return equal(a.impl(), b.impl()); } |
| 189 | inline bool operator==(const LChar* a, const AtomString& b) { return b == a; } |
| 190 | inline bool operator==(const String& a, const AtomString& b) { return equal(a.impl(), b.impl()); } |
| 191 | inline bool operator==(const Vector<UChar>& a, const AtomString& b) { return b == a; } |
| 192 | |
| 193 | inline bool operator!=(const AtomString& a, const AtomString& b) { return a.impl() != b.impl(); } |
| 194 | inline bool operator!=(const AtomString& a, const LChar* b) { return !(a == b); } |
| 195 | inline bool operator!=(const AtomString& a, const char* b) { return !(a == b); } |
| 196 | inline bool operator!=(const AtomString& a, const String& b) { return !equal(a.impl(), b.impl()); } |
| 197 | inline bool operator!=(const AtomString& a, const Vector<UChar>& b) { return !(a == b); } |
| 198 | inline bool operator!=(const LChar* a, const AtomString& b) { return !(b == a); } |
| 199 | inline bool operator!=(const String& a, const AtomString& b) { return !equal(a.impl(), b.impl()); } |
| 200 | inline bool operator!=(const Vector<UChar>& a, const AtomString& b) { return !(a == b); } |
| 201 | |
| 202 | bool equalIgnoringASCIICase(const AtomString&, const AtomString&); |
| 203 | bool equalIgnoringASCIICase(const AtomString&, const String&); |
| 204 | bool equalIgnoringASCIICase(const String&, const AtomString&); |
| 205 | bool equalIgnoringASCIICase(const AtomString&, const char*); |
| 206 | |
| 207 | template<unsigned length> bool equalLettersIgnoringASCIICase(const AtomString&, const char (&lowercaseLetters)[length]); |
| 208 | |
| 209 | inline AtomString::AtomString() |
| 210 | { |
| 211 | } |
| 212 | |
| 213 | inline AtomString::AtomString(const LChar* string) |
| 214 | : m_string(AtomStringImpl::add(string)) |
| 215 | { |
| 216 | } |
| 217 | |
| 218 | inline AtomString::AtomString(const char* string) |
| 219 | : m_string(AtomStringImpl::add(string)) |
| 220 | { |
| 221 | } |
| 222 | |
| 223 | inline AtomString::AtomString(const LChar* string, unsigned length) |
| 224 | : m_string(AtomStringImpl::add(string, length)) |
| 225 | { |
| 226 | } |
| 227 | |
| 228 | inline AtomString::AtomString(const UChar* string, unsigned length) |
| 229 | : m_string(AtomStringImpl::add(string, length)) |
| 230 | { |
| 231 | } |
| 232 | |
| 233 | inline AtomString::AtomString(const UChar* string) |
| 234 | : m_string(AtomStringImpl::add(string)) |
| 235 | { |
| 236 | } |
| 237 | |
| 238 | inline AtomString::AtomString(AtomStringImpl* string) |
| 239 | : m_string(string) |
| 240 | { |
| 241 | } |
| 242 | |
| 243 | inline AtomString::AtomString(RefPtr<AtomStringImpl>&& string) |
| 244 | : m_string(WTFMove(string)) |
| 245 | { |
| 246 | } |
| 247 | |
| 248 | inline AtomString::AtomString(StringImpl* string) |
| 249 | : m_string(AtomStringImpl::add(string)) |
| 250 | { |
| 251 | } |
| 252 | |
| 253 | inline AtomString::AtomString(const StaticStringImpl* string) |
| 254 | : m_string(AtomStringImpl::add(string)) |
| 255 | { |
| 256 | } |
| 257 | |
| 258 | inline AtomString::AtomString(const String& string) |
| 259 | : m_string(AtomStringImpl::add(string.impl())) |
| 260 | { |
| 261 | } |
| 262 | |
| 263 | inline AtomString::AtomString(StringImpl* baseString, unsigned start, unsigned length) |
| 264 | : m_string(AtomStringImpl::add(baseString, start, length)) |
| 265 | { |
| 266 | } |
| 267 | |
| 268 | inline AtomString::AtomString(UniquedStringImpl* uid) |
| 269 | : m_string(uid) |
| 270 | { |
| 271 | } |
| 272 | |
| 273 | #if USE(CF) |
| 274 | |
| 275 | inline AtomString::AtomString(CFStringRef string) |
| 276 | : m_string(AtomStringImpl::add(string)) |
| 277 | { |
| 278 | } |
| 279 | |
| 280 | #endif |
| 281 | |
| 282 | #ifdef __OBJC__ |
| 283 | |
| 284 | inline AtomString::AtomString(NSString* string) |
| 285 | : m_string(AtomStringImpl::add((__bridge CFStringRef)string)) |
| 286 | { |
| 287 | } |
| 288 | |
| 289 | #endif |
| 290 | |
| 291 | // Define external global variables for the commonly used atomic strings. |
| 292 | // These are only usable from the main thread. |
| 293 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> nullAtomData; |
| 294 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> emptyAtomData; |
| 295 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> starAtomData; |
| 296 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> xmlAtomData; |
| 297 | extern WTF_EXPORT_PRIVATE LazyNeverDestroyed<AtomString> xmlnsAtomData; |
| 298 | |
| 299 | inline const AtomString& nullAtom() { return nullAtomData.get(); } |
| 300 | inline const AtomString& emptyAtom() { return emptyAtomData.get(); } |
| 301 | inline const AtomString& starAtom() { return starAtomData.get(); } |
| 302 | inline const AtomString& xmlAtom() { return xmlAtomData.get(); } |
| 303 | inline const AtomString& xmlnsAtom() { return xmlnsAtomData.get(); } |
| 304 | |
| 305 | inline AtomString AtomString::fromUTF8(const char* characters, size_t length) |
| 306 | { |
| 307 | if (!characters) |
| 308 | return nullAtom(); |
| 309 | if (!length) |
| 310 | return emptyAtom(); |
| 311 | return fromUTF8Internal(characters, characters + length); |
| 312 | } |
| 313 | |
| 314 | inline AtomString AtomString::fromUTF8(const char* characters) |
| 315 | { |
| 316 | if (!characters) |
| 317 | return nullAtom(); |
| 318 | if (!*characters) |
| 319 | return emptyAtom(); |
| 320 | return fromUTF8Internal(characters, nullptr); |
| 321 | } |
| 322 | |
| 323 | // AtomStringHash is the default hash for AtomString |
| 324 | template<typename T> struct DefaultHash; |
| 325 | template<> struct DefaultHash<AtomString> { |
| 326 | typedef AtomStringHash Hash; |
| 327 | }; |
| 328 | |
| 329 | template<unsigned length> inline bool equalLettersIgnoringASCIICase(const AtomString& string, const char (&lowercaseLetters)[length]) |
| 330 | { |
| 331 | return equalLettersIgnoringASCIICase(string.string(), lowercaseLetters); |
| 332 | } |
| 333 | |
| 334 | inline bool equalIgnoringASCIICase(const AtomString& a, const AtomString& b) |
| 335 | { |
| 336 | return equalIgnoringASCIICase(a.string(), b.string()); |
| 337 | } |
| 338 | |
| 339 | inline bool equalIgnoringASCIICase(const AtomString& a, const String& b) |
| 340 | { |
| 341 | return equalIgnoringASCIICase(a.string(), b); |
| 342 | } |
| 343 | |
| 344 | inline bool equalIgnoringASCIICase(const String& a, const AtomString& b) |
| 345 | { |
| 346 | return equalIgnoringASCIICase(a, b.string()); |
| 347 | } |
| 348 | |
| 349 | inline bool equalIgnoringASCIICase(const AtomString& a, const char* b) |
| 350 | { |
| 351 | return equalIgnoringASCIICase(a.string(), b); |
| 352 | } |
| 353 | |
| 354 | template<> struct IntegerToStringConversionTrait<AtomString> { |
| 355 | using ReturnType = AtomString; |
| 356 | using AdditionalArgumentType = void; |
| 357 | static AtomString flush(LChar* characters, unsigned length, void*) { return { characters, length }; } |
| 358 | }; |
| 359 | |
| 360 | } // namespace WTF |
| 361 | |
| 362 | using WTF::AtomString; |
| 363 | using WTF::AtomicString; |
| 364 | using WTF::nullAtom; |
| 365 | using WTF::emptyAtom; |
| 366 | using WTF::starAtom; |
| 367 | using WTF::xmlAtom; |
| 368 | using WTF::xmlnsAtom; |
| 369 | |
| 370 | #include <wtf/text/StringConcatenate.h> |
| 371 | |