runtime/UString.h

   1 /*
   2  * Copyright (C) 1999-2000 Harri Porten (porten@kde.org)
   3  * Copyright (C) 2004, 2005, 2006, 2007, 2008, 2009 Apple Inc. All rights reserved.
   4  * Copyright (C) 2009 Google Inc. All rights reserved.
   5  *
   6  * This library is free software; you can redistribute it and/or
   7  * modify it under the terms of the GNU Library General Public
   8  * License as published by the Free Software Foundation; either
   9  * version 2 of the License, or (at your option) any later version.
  10  *
  11  * This library is distributed in the hope that it will be useful,
  12  * but WITHOUT ANY WARRANTY; without even the implied warranty of
  13  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
  14  * Library General Public License for more details.
  15  *
  16  * You should have received a copy of the GNU Library General Public License
  17  * along with this library; see the file COPYING.LIB.  If not, write to
  18  * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
  19  * Boston, MA 02110-1301, USA.
  20  *
  21  */
  22
  23 #ifndef UString_h
  24 #define UString_h
  25
  26 #include <wtf/text/StringImpl.h>
  27
  28 namespace JSC {
  29
  30 class UString {
  31 public:
  32     // Construct a null string, distinguishable from an empty string.
  33     UString() { }
  34
  35     // Construct a string with UTF-16 data.
  36     UString(const UChar* characters, unsigned length);
  37
  38     // Construct a string with UTF-16 data, from a null-terminated source.
  39     UString(const UChar*);
  40
  41     // Construct a string with latin1 data.
  42     UString(const char* characters, unsigned length);
  43
  44     // Construct a string with latin1 data, from a null-terminated source.
  45     UString(const char* characters);
  46
  47     // Construct a string referencing an existing StringImpl.
  48     UString(StringImpl* impl) : m_impl(impl) { }
  49     UString(PassRefPtr<StringImpl> impl) : m_impl(impl) { }
  50     UString(RefPtr<StringImpl> impl) : m_impl(impl) { }
  51
  52     // Inline the destructor.
  53     ALWAYS_INLINE ~UString() { }
  54
  55     void swap(UString& o) { m_impl.swap(o.m_impl); }
  56
  57     template<size_t inlineCapacity>
  58     static UString adopt(Vector<UChar, inlineCapacity>& vector) { return StringImpl::adopt(vector); }
  59
  60     bool isNull() const { return !m_impl; }
  61     bool isEmpty() const { return !m_impl || !m_impl->length(); }
  62
  63     StringImpl* impl() const { return m_impl.get(); }
  64
  65     unsigned length() const
  66     {
  67         if (!m_impl)
  68             return 0;
  69         return m_impl->length();
  70     }
  71
  72     const UChar* characters() const
  73     {
  74         if (!m_impl)
  75             return 0;
  76         return m_impl->characters();
  77     }
  78
  79     CString ascii() const;
  80     CString latin1() const;
  81     CString utf8(bool strict = false) const;
  82
  83     UChar operator[](unsigned index) const
  84     {
  85         if (!m_impl || index >= m_impl->length())
  86             return 0;
  87         return m_impl->characters()[index];
  88     }
  89
  90     static UString number(int);
  91     static UString number(unsigned);
  92     static UString number(long);
  93     static UString number(long long);
  94     static UString number(double);
  95
  96     // Find a single character or string, also with match function & latin1 forms.
  97     size_t find(UChar c, unsigned start = 0) const
  98         { return m_impl ? m_impl->find(c, start) : notFound; }
  99     size_t find(const UString& str, unsigned start = 0) const
 100         { return m_impl ? m_impl->find(str.impl(), start) : notFound; }
 101     size_t find(const char* str, unsigned start = 0) const
 102         { return m_impl ? m_impl->find(str, start) : notFound; }
 103
 104     // Find the last instance of a single character or string.
 105     size_t reverseFind(UChar c, unsigned start = UINT_MAX) const
 106         { return m_impl ? m_impl->reverseFind(c, start) : notFound; }
 107     size_t reverseFind(const UString& str, unsigned start = UINT_MAX) const
 108         { return m_impl ? m_impl->reverseFind(str.impl(), start) : notFound; }
 109
 110     UString substringSharingImpl(unsigned pos, unsigned len = UINT_MAX) const;
 111
 112 private:
 113     RefPtr<StringImpl> m_impl;
 114 };
 115
 116 ALWAYS_INLINE bool operator==(const UString& s1, const UString& s2)
 117 {
 118     StringImpl* rep1 = s1.impl();
 119     StringImpl* rep2 = s2.impl();
 120     unsigned size1 = 0;
 121     unsigned size2 = 0;
 122
 123     if (rep1 == rep2) // If they're the same rep, they're equal.
 124         return true;
 125
 126     if (rep1)
 127         size1 = rep1->length();
 128
 129     if (rep2)
 130         size2 = rep2->length();
 131
 132     if (size1 != size2) // If the lengths are not the same, we're done.
 133         return false;
 134
 135     if (!size1)
 136         return true;
 137
 138     // At this point we know
 139     //   (a) that the strings are the same length and
 140     //   (b) that they are greater than zero length.
 141     const UChar* d1 = rep1->characters();
 142     const UChar* d2 = rep2->characters();
 143
 144     if (d1 == d2) // Check to see if the data pointers are the same.
 145         return true;
 146
 147     // Do quick checks for sizes 1 and 2.
 148     switch (size1) {
 149     case 1:
 150         return d1[0] == d2[0];
 151     case 2:
 152         return (d1[0] == d2[0]) & (d1[1] == d2[1]);
 153     default:
 154         return memcmp(d1, d2, size1 * sizeof(UChar)) == 0;
 155     }
 156 }
 157
 158
 159 inline bool operator!=(const UString& s1, const UString& s2)
 160 {
 161     return !JSC::operator==(s1, s2);
 162 }
 163
 164 bool operator<(const UString& s1, const UString& s2);
 165 bool operator>(const UString& s1, const UString& s2);
 166
 167 bool operator==(const UString& s1, const char* s2);
 168
 169 inline bool operator!=(const UString& s1, const char* s2)
 170 {
 171     return !JSC::operator==(s1, s2);
 172 }
 173
 174 inline bool operator==(const char *s1, const UString& s2)
 175 {
 176     return operator==(s2, s1);
 177 }
 178
 179 inline bool operator!=(const char *s1, const UString& s2)
 180 {
 181     return !JSC::operator==(s1, s2);
 182 }
 183
 184 inline int codePointCompare(const UString& s1, const UString& s2)
 185 {
 186     return codePointCompare(s1.impl(), s2.impl());
 187 }
 188
 189 struct UStringHash {
 190     static unsigned hash(StringImpl* key) { return key->hash(); }
 191     static bool equal(const StringImpl* a, const StringImpl* b)
 192     {
 193         if (a == b)
 194             return true;
 195         if (!a || !b)
 196             return false;
 197
 198         unsigned aLength = a->length();
 199         unsigned bLength = b->length();
 200         if (aLength != bLength)
 201             return false;
 202
 203         // FIXME: perhaps we should have a more abstract macro that indicates when
 204         // going 4 bytes at a time is unsafe
 205 #if CPU(ARM) || CPU(SH4) || CPU(MIPS)
 206         const UChar* aChars = a->characters();
 207         const UChar* bChars = b->characters();
 208         for (unsigned i = 0; i != aLength; ++i) {
 209             if (*aChars++ != *bChars++)
 210                 return false;
 211         }
 212         return true;
 213 #else
 214         /* Do it 4-bytes-at-a-time on architectures where it's safe */
 215         const uint32_t* aChars = reinterpret_cast<const uint32_t*>(a->characters());
 216         const uint32_t* bChars = reinterpret_cast<const uint32_t*>(b->characters());
 217
 218         unsigned halfLength = aLength >> 1;
 219         for (unsigned i = 0; i != halfLength; ++i)
 220             if (*aChars++ != *bChars++)
 221                 return false;
 222
 223         if (aLength & 1 && *reinterpret_cast<const uint16_t*>(aChars) != *reinterpret_cast<const uint16_t*>(bChars))
 224             return false;
 225
 226         return true;
 227 #endif
 228     }
 229
 230     static unsigned hash(const RefPtr<StringImpl>& key) { return key->hash(); }
 231     static bool equal(const RefPtr<StringImpl>& a, const RefPtr<StringImpl>& b)
 232     {
 233         return equal(a.get(), b.get());
 234     }
 235
 236     static unsigned hash(const UString& key) { return key.impl()->hash(); }
 237     static bool equal(const UString& a, const UString& b)
 238     {
 239         return equal(a.impl(), b.impl());
 240     }
 241
 242     static const bool safeToCompareToEmptyOrDeleted = false;
 243 };
 244
 245 } // namespace JSC
 246
 247 namespace WTF {
 248
 249 // UStringHash is the default hash for UString
 250 template<typename T> struct DefaultHash;
 251 template<> struct DefaultHash<JSC::UString> {
 252     typedef JSC::UStringHash Hash;
 253 };
 254
 255 template <> struct VectorTraits<JSC::UString> : SimpleClassVectorTraits { };
 256
 257 } // namespace WTF
 258
 259 #endif
 260