Carla/source/modules/water/text/String.cpp

/*
  ==============================================================================

   This file is part of the Water library.
   Copyright (c) 2016 ROLI Ltd.
   Copyright (C) 2017-2018 Filipe Coelho <falktx@falktx.com>

   Permission is granted to use this software under the terms of the ISC license
   http://www.isc.org/downloads/software-support-policy/isc-license/

   Permission to use, copy, modify, and/or distribute this software for any
   purpose with or without fee is hereby granted, provided that the above
   copyright notice and this permission notice appear in all copies.

   THE SOFTWARE IS PROVIDED "AS IS" AND ISC DISCLAIMS ALL WARRANTIES WITH REGARD
   TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF MERCHANTABILITY AND
   FITNESS. IN NO EVENT SHALL ISC BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT,
   OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF
   USE, DATA OR PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
   TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR PERFORMANCE
   OF THIS SOFTWARE.

  ==============================================================================
*/

#include "String.h"
#include "NewLine.h"
#include "../maths/MathsFunctions.h"
#include "../memory/HeapBlock.h"
#include "../streams/OutputStream.h"

#include "CarlaJuceUtils.hpp"

#include <locale>
#include <iostream>

#ifdef CARLA_OS_MAC
// FIXME
// # import <CoreData/CoreData.h>
#endif

// #import <Carbon/UnicodeConverter.h>

namespace water {

NewLine newLine;

//==============================================================================
// (Mirrors the structure of StringHolder, but without the atomic member, so can be statically constructed)
struct EmptyString
{
    int refCount;
    size_t allocatedBytes;
    String::CharPointerType::CharType text;
};

static const EmptyString emptyString = { 0x3fffffff, sizeof (String::CharPointerType::CharType), 0 };

//==============================================================================
class StringHolder
{
public:
    StringHolder() WATER_DELETED_FUNCTION;

    typedef String::CharPointerType CharPointerType;
    typedef String::CharPointerType::CharType CharType;

    //==============================================================================
    static CharPointerType createUninitialisedBytes (size_t numBytes)
    {
        numBytes = (numBytes + 3) & ~(size_t) 3;
        StringHolder* const s = reinterpret_cast<StringHolder*> (new char [sizeof (StringHolder) - sizeof (CharType) + numBytes]);
        s->refCount.value = 0;
        s->allocatedNumBytes = numBytes;
        return CharPointerType (s->text);
    }

    template <class CharPointer>
    static CharPointerType createFromCharPointer (const CharPointer text)
    {
        if (text.getAddress() == nullptr || text.isEmpty())
            return CharPointerType (&(emptyString.text));

        const size_t bytesNeeded = sizeof (CharType) + CharPointerType::getBytesRequiredFor (text);
        const CharPointerType dest (createUninitialisedBytes (bytesNeeded));
        CharPointerType (dest).writeAll (text);
        return dest;
    }

    template <class CharPointer>
    static CharPointerType createFromCharPointer (const CharPointer text, size_t maxChars)
    {
        if (text.getAddress() == nullptr || text.isEmpty() || maxChars == 0)
            return CharPointerType (&(emptyString.text));

        CharPointer end (text);
        size_t numChars = 0;
        size_t bytesNeeded = sizeof (CharType);

        while (numChars < maxChars && ! end.isEmpty())
        {
            bytesNeeded += CharPointerType::getBytesRequiredFor (end.getAndAdvance());
            ++numChars;
        }

        const CharPointerType dest (createUninitialisedBytes (bytesNeeded));
        CharPointerType (dest).writeWithCharLimit (text, (int) numChars + 1);
        return dest;
    }

    template <class CharPointer>
    static CharPointerType createFromCharPointer (const CharPointer start, const CharPointer end)
    {
        if (start.getAddress() == nullptr || start.isEmpty())
            return CharPointerType (&(emptyString.text));

        CharPointer e (start);
        int numChars = 0;
        size_t bytesNeeded = sizeof (CharType);

        while (e < end && ! e.isEmpty())
        {
            bytesNeeded += CharPointerType::getBytesRequiredFor (e.getAndAdvance());
            ++numChars;
        }

        const CharPointerType dest (createUninitialisedBytes (bytesNeeded));
        CharPointerType (dest).writeWithCharLimit (start, numChars + 1);
        return dest;
    }

    static CharPointerType createFromCharPointer (const CharPointerType start, const CharPointerType end)
    {
        if (start.getAddress() == nullptr || start.isEmpty())
            return CharPointerType (&(emptyString.text));

        const size_t numBytes = (size_t) (reinterpret_cast<const char*> (end.getAddress())
                                           - reinterpret_cast<const char*> (start.getAddress()));
        const CharPointerType dest (createUninitialisedBytes (numBytes + sizeof (CharType)));
        memcpy (dest.getAddress(), start, numBytes);
        dest.getAddress()[numBytes / sizeof (CharType)] = 0;
        return dest;
    }

    static CharPointerType createFromFixedLength (const char* const src, const size_t numChars)
    {
        const CharPointerType dest (createUninitialisedBytes (numChars * sizeof (CharType) + sizeof (CharType)));
        CharPointerType (dest).writeWithCharLimit (CharPointer_UTF8 (src), (int) (numChars + 1));
        return dest;
    }

    //==============================================================================
    static void retain (const CharPointerType text) noexcept
    {
        StringHolder* const b = bufferFromText (text);

        if (b != (StringHolder*) &emptyString)
            ++(b->refCount);
    }

    static inline void release (StringHolder* const b) noexcept
    {
        if (b != (StringHolder*) &emptyString)
            if (--(b->refCount) == -1)
                delete[] reinterpret_cast<char*> (b);
    }

    static void release (const CharPointerType text) noexcept
    {
        release (bufferFromText (text));
    }

    static inline int getReferenceCount (const CharPointerType text) noexcept
    {
        return bufferFromText (text)->refCount.get() + 1;
    }

    //==============================================================================
    static CharPointerType makeUniqueWithByteSize (const CharPointerType text, size_t numBytes)
    {
        StringHolder* const b = bufferFromText (text);

        if (b == (StringHolder*) &emptyString)
        {
            CharPointerType newText (createUninitialisedBytes (numBytes));
            newText.writeNull();
            return newText;
        }

        if (b->allocatedNumBytes >= numBytes && b->refCount.get() <= 0)
            return text;

        CharPointerType newText (createUninitialisedBytes (jmax (b->allocatedNumBytes, numBytes)));
        memcpy (newText.getAddress(), text.getAddress(), b->allocatedNumBytes);
        release (b);

        return newText;
    }

    static size_t getAllocatedNumBytes (const CharPointerType text) noexcept
    {
        return bufferFromText (text)->allocatedNumBytes;
    }

    //==============================================================================
    Atomic<int> refCount;
    size_t allocatedNumBytes;
    CharType text[1];

private:
    static inline StringHolder* bufferFromText (const CharPointerType text) noexcept
    {
        // (Can't use offsetof() here because of warnings about this not being a POD)
        return reinterpret_cast<StringHolder*> (reinterpret_cast<char*> (text.getAddress())
                    - (reinterpret_cast<size_t> (reinterpret_cast<StringHolder*> (1)->text) - 1));
    }
};

//==============================================================================
String::String() noexcept  : text (&(emptyString.text))
{
}

String::~String() noexcept
{
    StringHolder::release (text);
}

String::String (const String& other) noexcept   : text (other.text)
{
    StringHolder::retain (text);
}

void String::swapWith (String& other) noexcept
{
    std::swap (text, other.text);
}

void String::clear() noexcept
{
    StringHolder::release (text);
    text = &(emptyString.text);
}

String& String::operator= (const String& other) noexcept
{
    StringHolder::retain (other.text);
    StringHolder::release (text.atomicSwap (other.text));
    return *this;
}

#if WATER_COMPILER_SUPPORTS_MOVE_SEMANTICS
String::String (String&& other) noexcept   : text (other.text)
{
    other.text = &(emptyString.text);
}

String& String::operator= (String&& other) noexcept
{
    std::swap (text, other.text);
    return *this;
}
#endif

inline String::PreallocationBytes::PreallocationBytes (const size_t num) noexcept : numBytes (num) {}

String::String (const PreallocationBytes& preallocationSize)
    : text (StringHolder::createUninitialisedBytes (preallocationSize.numBytes + sizeof (CharPointerType::CharType)))
{
}

void String::preallocateBytes (const size_t numBytesNeeded)
{
    text = StringHolder::makeUniqueWithByteSize (text, numBytesNeeded + sizeof (CharPointerType::CharType));
}

int String::getReferenceCount() const noexcept
{
    return StringHolder::getReferenceCount (text);
}

//==============================================================================
String::String (const char* const t)
    : text (StringHolder::createFromCharPointer (CharPointer_UTF8 (t)))
{
    /*  If you get an assertion here, then you're trying to create a string from 8-bit data
        that contains values greater than 127. These can NOT be correctly converted to unicode
        because there's no way for the String class to know what encoding was used to
        create them. The source data could be UTF-8, ASCII or one of many local code-pages.

        To get around this problem, you must be more explicit when you pass an ambiguous 8-bit
        string to the String class - so for example if your source data is actually UTF-8,
        you'd call String (CharPointer_UTF8 ("my utf8 string..")), and it would be able to
        correctly convert the multi-byte characters to unicode. It's *highly* recommended that
        you use UTF-8 with escape characters in your source code to represent extended characters,
        because there's no other way to represent these strings in a way that isn't dependent on
        the compiler, source code editor and platform.
    */
    CARLA_SAFE_ASSERT (t == nullptr || CharPointer_UTF8::isValidString (t, std::numeric_limits<int>::max()));
}

String::String (const char* const t, const size_t maxChars)
    : text (StringHolder::createFromCharPointer (CharPointer_UTF8 (t), maxChars))
{
    /*  If you get an assertion here, then you're trying to create a string from 8-bit data
        that contains values greater than 127. These can NOT be correctly converted to unicode
        because there's no way for the String class to know what encoding was used to
        create them. The source data could be UTF-8, ASCII or one of many local code-pages.

        To get around this problem, you must be more explicit when you pass an ambiguous 8-bit
        string to the String class - so for example if your source data is actually UTF-8,
        you'd call String (CharPointer_UTF8 ("my utf8 string..")), and it would be able to
        correctly convert the multi-byte characters to unicode. It's *highly* recommended that
        you use UTF-8 with escape characters in your source code to represent extended characters,
        because there's no other way to represent these strings in a way that isn't dependent on
        the compiler, source code editor and platform.
    */
    CARLA_SAFE_ASSERT (t == nullptr || CharPointer_UTF8::isValidString (t, (int) maxChars));
}

String::String (const CharPointer_UTF8  t)   : text (StringHolder::createFromCharPointer (t)) {}
String::String (const CharPointer_UTF8  t, const size_t maxChars)   : text (StringHolder::createFromCharPointer (t, maxChars)) {}
String::String (const CharPointer_UTF8  start, const CharPointer_UTF8  end)  : text (StringHolder::createFromCharPointer (start, end)) {}

String::String (const std::string& s) : text (StringHolder::createFromFixedLength (s.data(), s.size())) {}
String::String (StringRef s)          : text (StringHolder::createFromCharPointer (s.text)) {}

String String::charToString (const water_uchar character)
{
    String result (PreallocationBytes (CharPointerType::getBytesRequiredFor (character)));
    CharPointerType t (result.text);
    t.write (character);
    t.writeNull();
    return result;
}

//==============================================================================
namespace NumberToStringConverters
{
    static char* numberToString (char* t, uint64 v) noexcept
    {
        return printDigits (t, v);
    }

    static char* numberToString (char* t, const int n) noexcept
    {
        if (n >= 0)
            return printDigits (t, static_cast<unsigned int> (n));

        // NB: this needs to be careful not to call -std::numeric_limits<int>::min(),
        // which has undefined behaviour
        t = printDigits (t, static_cast<unsigned int> (-(n + 1)) + 1);
        *--t = '-';
        return t;
    }

    static char* numberToString (char* t, const unsigned int v) noexcept
    {
        return printDigits (t, v);
    }

    static char* numberToString (char* t, const long n) noexcept
    {
        if (n >= 0)
            return printDigits (t, static_cast<unsigned long> (n));

        t = printDigits (t, static_cast<unsigned long> (-(n + 1)) + 1);
        *--t = '-';
        return t;
    }

    static char* numberToString (char* t, const unsigned long v) noexcept
    {
        return printDigits (t, v);
    }

    struct StackArrayStream  : public std::basic_streambuf<char, std::char_traits<char> >
    {
        explicit StackArrayStream (char* d)
        {
            static const std::locale classicLocale (std::locale::classic());
            imbue (classicLocale);
            setp (d, d + charsNeededForDouble);
        }

        size_t writeDouble (double n, int numDecPlaces)
        {
            {
                std::ostream o (this);

                if (numDecPlaces > 0)
                    o.precision ((std::streamsize) numDecPlaces);

                o << n;
            }

            return (size_t) (pptr() - pbase());
        }
    };

    static char* doubleToString (char* buffer, const int numChars, double n, int numDecPlaces, size_t& len) noexcept
    {
        if (numDecPlaces > 0 && numDecPlaces < 7 && n > -1.0e20 && n < 1.0e20)
        {
            char* const end = buffer + numChars;
            char* t = end;
            int64 v = (int64) (pow (10.0, numDecPlaces) * std::abs (n) + 0.5);
            *--t = (char) 0;

            while (numDecPlaces >= 0 || v > 0)
            {
                if (numDecPlaces == 0)
                    *--t = '.';

                *--t = (char) ('0' + (v % 10));

                v /= 10;
                --numDecPlaces;
            }

            if (n < 0)
                *--t = '-';

            len = (size_t) (end - t - 1);
            return t;
        }

        StackArrayStream strm (buffer);
        len = strm.writeDouble (n, numDecPlaces);
        jassert (len <= charsNeededForDouble);
        return buffer;
    }

    template <typename IntegerType>
    static String::CharPointerType createFromInteger (const IntegerType number)
    {
        char buffer [charsNeededForInt];
        char* const end = buffer + numElementsInArray (buffer);
        char* const start = numberToString (end, number);
        return StringHolder::createFromFixedLength (start, (size_t) (end - start - 1));
    }

    static String::CharPointerType createFromDouble (const double number, const int numberOfDecimalPlaces)
    {
        char buffer [charsNeededForDouble];
        size_t len;
        char* const start = doubleToString (buffer, numElementsInArray (buffer), (double) number, numberOfDecimalPlaces, len);
        return StringHolder::createFromFixedLength (start, len);
    }
}

//==============================================================================
String::String (const int number)            : text (NumberToStringConverters::createFromInteger (number)) {}
String::String (const unsigned int number)   : text (NumberToStringConverters::createFromInteger (number)) {}
String::String (const short number)          : text (NumberToStringConverters::createFromInteger ((int) number)) {}
String::String (const unsigned short number) : text (NumberToStringConverters::createFromInteger ((unsigned int) number)) {}
String::String (const int64  number)         : text (NumberToStringConverters::createFromInteger (number)) {}
String::String (const uint64 number)         : text (NumberToStringConverters::createFromInteger (number)) {}
String::String (const long number)           : text (NumberToStringConverters::createFromInteger (number)) {}
String::String (const unsigned long number)  : text (NumberToStringConverters::createFromInteger (number)) {}

String::String (const float  number)         : text (NumberToStringConverters::createFromDouble ((double) number, 0)) {}
String::String (const double number)         : text (NumberToStringConverters::createFromDouble (number, 0)) {}
String::String (const float  number, const int numberOfDecimalPlaces)  : text (NumberToStringConverters::createFromDouble ((double) number, numberOfDecimalPlaces)) {}
String::String (const double number, const int numberOfDecimalPlaces)  : text (NumberToStringConverters::createFromDouble (number, numberOfDecimalPlaces)) {}

//==============================================================================
int String::length() const noexcept
{
    return (int) text.length();
}

static size_t findByteOffsetOfEnd (String::CharPointerType text) noexcept
{
    return (size_t) (((char*) text.findTerminatingNull().getAddress()) - (char*) text.getAddress());
}

size_t String::getByteOffsetOfEnd() const noexcept
{
    return findByteOffsetOfEnd (text);
}

water_uchar String::operator[] (int index) const noexcept
{
    jassert (index == 0 || (index > 0 && index <= (int) text.lengthUpTo ((size_t) index + 1)));
    return text [index];
}

//==============================================================================
bool operator== (const String& s1, const String& s2) noexcept            { return s1.compare (s2) == 0; }
bool operator!= (const String& s1, const String& s2) noexcept            { return s1.compare (s2) != 0; }
bool operator== (const String& s1, const char* s2) noexcept              { return s1.compare (s2) == 0; }
bool operator!= (const String& s1, const char* s2) noexcept              { return s1.compare (s2) != 0; }
bool operator== (const String& s1, StringRef s2) noexcept                { return s1.getCharPointer().compare (s2.text) == 0; }
bool operator!= (const String& s1, StringRef s2) noexcept                { return s1.getCharPointer().compare (s2.text) != 0; }
bool operator== (const String& s1, const CharPointer_UTF8 s2) noexcept   { return s1.getCharPointer().compare (s2) == 0; }
bool operator!= (const String& s1, const CharPointer_UTF8 s2) noexcept   { return s1.getCharPointer().compare (s2) != 0; }
bool operator>  (const String& s1, const String& s2) noexcept            { return s1.compare (s2) > 0; }
bool operator<  (const String& s1, const String& s2) noexcept            { return s1.compare (s2) < 0; }
bool operator>= (const String& s1, const String& s2) noexcept            { return s1.compare (s2) >= 0; }
bool operator<= (const String& s1, const String& s2) noexcept            { return s1.compare (s2) <= 0; }

bool String::equalsIgnoreCase (const char* const t) const noexcept
{
    return t != nullptr ? text.compareIgnoreCase (CharPointer_UTF8 (t)) == 0
                        : isEmpty();
}

bool String::equalsIgnoreCase (StringRef t) const noexcept
{
    return text.compareIgnoreCase (t.text) == 0;
}

bool String::equalsIgnoreCase (const String& other) const noexcept
{
    return text == other.text
            || text.compareIgnoreCase (other.text) == 0;
}

int String::compare (const String& other) const noexcept           { return (text == other.text) ? 0 : text.compare (other.text); }
int String::compare (const char* const other) const noexcept       { return text.compare (CharPointer_UTF8 (other)); }
int String::compareIgnoreCase (const String& other) const noexcept { return (text == other.text) ? 0 : text.compareIgnoreCase (other.text); }

static int stringCompareRight (String::CharPointerType s1, String::CharPointerType s2) noexcept
{
    for (int bias = 0;;)
    {
        const water_uchar c1 = s1.getAndAdvance();
        const bool isDigit1 = CharacterFunctions::isDigit (c1);

        const water_uchar c2 = s2.getAndAdvance();
        const bool isDigit2 = CharacterFunctions::isDigit (c2);

        if (! (isDigit1 || isDigit2))   return bias;
        if (! isDigit1)                 return -1;
        if (! isDigit2)                 return 1;

        if (c1 != c2 && bias == 0)
            bias = c1 < c2 ? -1 : 1;

        jassert (c1 != 0 && c2 != 0);
    }
}

static int stringCompareLeft (String::CharPointerType s1, String::CharPointerType s2) noexcept
{
    for (;;)
    {
        const water_uchar c1 = s1.getAndAdvance();
        const bool isDigit1 = CharacterFunctions::isDigit (c1);

        const water_uchar c2 = s2.getAndAdvance();
        const bool isDigit2 = CharacterFunctions::isDigit (c2);

        if (! (isDigit1 || isDigit2))   return 0;
        if (! isDigit1)                 return -1;
        if (! isDigit2)                 return 1;
        if (c1 < c2)                    return -1;
        if (c1 > c2)                    return 1;
    }
}

static int naturalStringCompare (String::CharPointerType s1, String::CharPointerType s2, bool isCaseSensitive) noexcept
{
    bool firstLoop = true;

    for (;;)
    {
        const bool hasSpace1 = s1.isWhitespace();
        const bool hasSpace2 = s2.isWhitespace();

        if ((! firstLoop) && (hasSpace1 ^ hasSpace2))
            return hasSpace2 ? 1 : -1;

        firstLoop = false;

        if (hasSpace1)  s1 = s1.findEndOfWhitespace();
        if (hasSpace2)  s2 = s2.findEndOfWhitespace();

        if (s1.isDigit() && s2.isDigit())
        {
            const int result = (*s1 == '0' || *s2 == '0') ? stringCompareLeft  (s1, s2)
                                                          : stringCompareRight (s1, s2);

            if (result != 0)
                return result;
        }

        water_uchar c1 = s1.getAndAdvance();
        water_uchar c2 = s2.getAndAdvance();

        if (c1 != c2 && ! isCaseSensitive)
        {
            c1 = CharacterFunctions::toUpperCase (c1);
            c2 = CharacterFunctions::toUpperCase (c2);
        }

        if (c1 == c2)
        {
            if (c1 == 0)
                return 0;
        }
        else
        {
            const bool isAlphaNum1 = CharacterFunctions::isLetterOrDigit (c1);
            const bool isAlphaNum2 = CharacterFunctions::isLetterOrDigit (c2);

            if (isAlphaNum2 && ! isAlphaNum1) return -1;
            if (isAlphaNum1 && ! isAlphaNum2) return 1;

            return c1 < c2 ? -1 : 1;
        }

        jassert (c1 != 0 && c2 != 0);
    }
}

int String::compareNatural (StringRef other, bool isCaseSensitive) const noexcept
{
    return naturalStringCompare (getCharPointer(), other.text, isCaseSensitive);
}

//==============================================================================
void String::append (const String& textToAppend, size_t maxCharsToTake)
{
    appendCharPointer (this == &textToAppend ? String (textToAppend).text
                                             : textToAppend.text, maxCharsToTake);
}

void String::appendCharPointer (const CharPointerType textToAppend)
{
    appendCharPointer (textToAppend, textToAppend.findTerminatingNull());
}

void String::appendCharPointer (const CharPointerType startOfTextToAppend,
                                const CharPointerType endOfTextToAppend)
{
    jassert (startOfTextToAppend.getAddress() != nullptr && endOfTextToAppend.getAddress() != nullptr);

    const int extraBytesNeeded = getAddressDifference (endOfTextToAppend.getAddress(),
                                                       startOfTextToAppend.getAddress());
    jassert (extraBytesNeeded >= 0);

    if (extraBytesNeeded > 0)
    {
        const size_t byteOffsetOfNull = getByteOffsetOfEnd();
        preallocateBytes (byteOffsetOfNull + (size_t) extraBytesNeeded);

        CharPointerType::CharType* const newStringStart = addBytesToPointer (text.getAddress(), (int) byteOffsetOfNull);
        memcpy (newStringStart, startOfTextToAppend.getAddress(), (size_t) extraBytesNeeded);
        CharPointerType (addBytesToPointer (newStringStart, extraBytesNeeded)).writeNull();
    }
}

String& String::operator+= (const char* const t)
{
    appendCharPointer (CharPointer_UTF8 (t)); // (using UTF8 here triggers a faster code-path than ascii)
    return *this;
}

String& String::operator+= (const String& other)
{
    if (isEmpty())
        return operator= (other);

    if (this == &other)
        return operator+= (String (*this));

    appendCharPointer (other.text);
    return *this;
}

String& String::operator+= (StringRef other)
{
    return operator+= (String (other));
}

String& String::operator+= (const char ch)
{
    const char asString[] = { ch, 0 };
    return operator+= (asString);
}

String& String::operator+= (const water_uchar ch)
{
    return operator+= (charToString(ch));
}

namespace StringHelpers
{
    template <typename T>
    inline String& operationAddAssign (String& str, const T number)
    {
        char buffer [(sizeof(T) * 8) / 2];
        char* end = buffer + numElementsInArray (buffer);
        char* start = NumberToStringConverters::numberToString (end, number);

        str.appendCharPointer (String::CharPointerType (start), String::CharPointerType (end));
        return str;
    }
}

String& String::operator+= (const int number)          { return StringHelpers::operationAddAssign<int>          (*this, number); }
String& String::operator+= (const int64 number)        { return StringHelpers::operationAddAssign<int64>        (*this, number); }
String& String::operator+= (const uint64 number)       { return StringHelpers::operationAddAssign<uint64>       (*this, number); }

//==============================================================================
String operator+ (const char* const s1, const String& s2)    { String s (s1); return s += s2; }
String operator+ (const char s1, const String& s2)           { return String::charToString ((water_uchar) (uint8) s1) + s2; }
String operator+ (String s1, const String& s2)               { return s1 += s2; }
String operator+ (String s1, const char* const s2)           { return s1 += s2; }
String operator+ (String s1, const char s2)                  { return s1 += s2; }

String operator+ (const water_uchar s1, const String& s2)     { return String::charToString (s1) + s2; }
String operator+ (String s1, const water_uchar s2)            { return s1 += s2; }
String& operator<< (String& s1, const water_uchar s2)         { return s1 += s2; }

String& operator<< (String& s1, const char s2)               { return s1 += s2; }
String& operator<< (String& s1, const char* const s2)        { return s1 += s2; }
String& operator<< (String& s1, const String& s2)            { return s1 += s2; }
String& operator<< (String& s1, StringRef s2)                { return s1 += s2; }

String& operator<< (String& s1, const int number)            { return s1 += number; }
String& operator<< (String& s1, const short number)          { return s1 += (int) number; }
String& operator<< (String& s1, const unsigned short number) { return s1 += (uint64) number; }
String& operator<< (String& s1, const long number)           { return s1 += String (number); }
String& operator<< (String& s1, const unsigned long number)  { return s1 += String (number); }
String& operator<< (String& s1, const int64 number)          { return s1 += String (number); }
String& operator<< (String& s1, const uint64 number)         { return s1 += String (number); }
String& operator<< (String& s1, const float number)          { return s1 += String (number); }
String& operator<< (String& s1, const double number)         { return s1 += String (number); }

OutputStream& operator<< (OutputStream& stream, const String& text)
{
    return operator<< (stream, StringRef (text));
}

OutputStream& operator<< (OutputStream& stream, StringRef text)
{
    const size_t numBytes = CharPointer_UTF8::getBytesRequiredFor (text.text);

    stream.write (text.text.getAddress(), numBytes);
    return stream;
}

//==============================================================================
int String::indexOfChar (const water_uchar character) const noexcept
{
    return text.indexOf (character);
}

int String::indexOfChar (const int startIndex, const water_uchar character) const noexcept
{
    CharPointerType t (text);

    for (int i = 0; ! t.isEmpty(); ++i)
    {
        if (i >= startIndex)
        {
            if (t.getAndAdvance() == character)
                return i;
        }
        else
        {
            ++t;
        }
    }

    return -1;
}

int String::lastIndexOfChar (const water_uchar character) const noexcept
{
    CharPointerType t (text);
    int last = -1;

    for (int i = 0; ! t.isEmpty(); ++i)
        if (t.getAndAdvance() == character)
            last = i;

    return last;
}

int String::indexOfAnyOf (StringRef charactersToLookFor, const int startIndex, const bool ignoreCase) const noexcept
{
    CharPointerType t (text);

    for (int i = 0; ! t.isEmpty(); ++i)
    {
        if (i >= startIndex)
        {
            if (charactersToLookFor.text.indexOf (t.getAndAdvance(), ignoreCase) >= 0)
                return i;
        }
        else
        {
            ++t;
        }
    }

    return -1;
}

int String::indexOf (StringRef other) const noexcept
{
    return other.isEmpty() ? 0 : text.indexOf (other.text);
}

int String::indexOfIgnoreCase (StringRef other) const noexcept
{
    return other.isEmpty() ? 0 : CharacterFunctions::indexOfIgnoreCase (text, other.text);
}

int String::indexOf (const int startIndex, StringRef other) const noexcept
{
    if (other.isEmpty())
        return -1;

    CharPointerType t (text);

    for (int i = startIndex; --i >= 0;)
    {
        if (t.isEmpty())
            return -1;

        ++t;
    }

    int found = t.indexOf (other.text);
    if (found >= 0)
        found += startIndex;
    return found;
}

int String::indexOfIgnoreCase (const int startIndex, StringRef other) const noexcept
{
    if (other.isEmpty())
        return -1;

    CharPointerType t (text);

    for (int i = startIndex; --i >= 0;)
    {
        if (t.isEmpty())
            return -1;

        ++t;
    }

    int found = CharacterFunctions::indexOfIgnoreCase (t, other.text);
    if (found >= 0)
        found += startIndex;
    return found;
}

int String::lastIndexOf (StringRef other) const noexcept
{
    if (other.isNotEmpty())
    {
        const int len = other.length();
        int i = length() - len;

        if (i >= 0)
        {
            for (CharPointerType n (text + i); i >= 0; --i)
            {
                if (n.compareUpTo (other.text, len) == 0)
                    return i;

                --n;
            }
        }
    }

    return -1;
}

int String::lastIndexOfIgnoreCase (StringRef other) const noexcept
{
    if (other.isNotEmpty())
    {
        const int len = other.length();
        int i = length() - len;

        if (i >= 0)
        {
            for (CharPointerType n (text + i); i >= 0; --i)
            {
                if (n.compareIgnoreCaseUpTo (other.text, len) == 0)
                    return i;

                --n;
            }
        }
    }

    return -1;
}

int String::lastIndexOfAnyOf (StringRef charactersToLookFor, const bool ignoreCase) const noexcept
{
    CharPointerType t (text);
    int last = -1;

    for (int i = 0; ! t.isEmpty(); ++i)
        if (charactersToLookFor.text.indexOf (t.getAndAdvance(), ignoreCase) >= 0)
            last = i;

    return last;
}

bool String::contains (StringRef other) const noexcept
{
    return indexOf (other) >= 0;
}

bool String::containsChar (const water_uchar character) const noexcept
{
    return text.indexOf (character) >= 0;
}

bool String::containsIgnoreCase (StringRef t) const noexcept
{
    return indexOfIgnoreCase (t) >= 0;
}

int String::indexOfWholeWord (StringRef word) const noexcept
{
    if (word.isNotEmpty())
    {
        CharPointerType t (text);
        const int wordLen = word.length();
        const int end = (int) t.length() - wordLen;

        for (int i = 0; i <= end; ++i)
        {
            if (t.compareUpTo (word.text, wordLen) == 0
                  && (i == 0 || ! (t - 1).isLetterOrDigit())
                  && ! (t + wordLen).isLetterOrDigit())
                return i;

            ++t;
        }
    }

    return -1;
}

int String::indexOfWholeWordIgnoreCase (StringRef word) const noexcept
{
    if (word.isNotEmpty())
    {
        CharPointerType t (text);
        const int wordLen = word.length();
        const int end = (int) t.length() - wordLen;

        for (int i = 0; i <= end; ++i)
        {
            if (t.compareIgnoreCaseUpTo (word.text, wordLen) == 0
                  && (i == 0 || ! (t - 1).isLetterOrDigit())
                  && ! (t + wordLen).isLetterOrDigit())
                return i;

            ++t;
        }
    }

    return -1;
}

bool String::containsWholeWord (StringRef wordToLookFor) const noexcept
{
    return indexOfWholeWord (wordToLookFor) >= 0;
}

bool String::containsWholeWordIgnoreCase (StringRef wordToLookFor) const noexcept
{
    return indexOfWholeWordIgnoreCase (wordToLookFor) >= 0;
}

//==============================================================================
template <typename CharPointer>
struct WildCardMatcher
{
    static bool matches (CharPointer wildcard, CharPointer test, const bool ignoreCase) noexcept
    {
        for (;;)
        {
            const water_uchar wc = wildcard.getAndAdvance();

            if (wc == '*')
                return wildcard.isEmpty() || matchesAnywhere (wildcard, test, ignoreCase);

            if (! characterMatches (wc, test.getAndAdvance(), ignoreCase))
                return false;

            if (wc == 0)
                return true;
        }
    }

    static bool characterMatches (const water_uchar wc, const water_uchar tc, const bool ignoreCase) noexcept
    {
        return (wc == tc) || (wc == '?' && tc != 0)
                || (ignoreCase && CharacterFunctions::toLowerCase (wc) == CharacterFunctions::toLowerCase (tc));
    }

    static bool matchesAnywhere (const CharPointer wildcard, CharPointer test, const bool ignoreCase) noexcept
    {
        for (; ! test.isEmpty(); ++test)
            if (matches (wildcard, test, ignoreCase))
                return true;

        return false;
    }
};

bool String::matchesWildcard (StringRef wildcard, const bool ignoreCase) const noexcept
{
    return WildCardMatcher<CharPointerType>::matches (wildcard.text, text, ignoreCase);
}

//==============================================================================
String String::repeatedString (StringRef stringToRepeat, int numberOfTimesToRepeat)
{
    if (numberOfTimesToRepeat <= 0)
        return String();

    String result (PreallocationBytes (findByteOffsetOfEnd (stringToRepeat) * (size_t) numberOfTimesToRepeat));
    CharPointerType n (result.text);

    while (--numberOfTimesToRepeat >= 0)
        n.writeAll (stringToRepeat.text);

    return result;
}

String String::paddedLeft (const water_uchar padCharacter, int minimumLength) const
{
    jassert (padCharacter != 0);

    int extraChars = minimumLength;
    CharPointerType end (text);

    while (! end.isEmpty())
    {
        --extraChars;
        ++end;
    }

    if (extraChars <= 0 || padCharacter == 0)
        return *this;

    const size_t currentByteSize = (size_t) (((char*) end.getAddress()) - (char*) text.getAddress());
    String result (PreallocationBytes (currentByteSize + (size_t) extraChars * CharPointerType::getBytesRequiredFor (padCharacter)));
    CharPointerType n (result.text);

    while (--extraChars >= 0)
        n.write (padCharacter);

    n.writeAll (text);
    return result;
}

String String::paddedRight (const water_uchar padCharacter, int minimumLength) const
{
    CARLA_SAFE_ASSERT_RETURN (padCharacter != 0, *this);

    int extraChars = minimumLength;
    CharPointerType end (text);

    while (! end.isEmpty())
    {
        --extraChars;
        ++end;
    }

    if (extraChars <= 0 || padCharacter == 0)
        return *this;

    const size_t currentByteSize = (size_t) (((char*) end.getAddress()) - (char*) text.getAddress());
    String result (PreallocationBytes (currentByteSize + (size_t) extraChars * CharPointerType::getBytesRequiredFor (padCharacter)));
    CharPointerType n (result.text);

    n.writeAll (text);

    while (--extraChars >= 0)
        n.write (padCharacter);

    n.writeNull();
    return result;
}

//==============================================================================
String String::replaceSection (int index, int numCharsToReplace, StringRef stringToInsert) const
{
    if (index < 0)
    {
        // a negative index to replace from?
        jassertfalse;
        index = 0;
    }

    if (numCharsToReplace < 0)
    {
        // replacing a negative number of characters?
        numCharsToReplace = 0;
        jassertfalse;
    }

    CharPointerType insertPoint (text);

    for (int i = 0; i < index; ++i)
    {
        if (insertPoint.isEmpty())
        {
            // replacing beyond the end of the string?
            jassertfalse;
            return *this + stringToInsert;
        }

        ++insertPoint;
    }

    CharPointerType startOfRemainder (insertPoint);

    for (int i = 0; i < numCharsToReplace && ! startOfRemainder.isEmpty(); ++i)
        ++startOfRemainder;

    if (insertPoint == text && startOfRemainder.isEmpty())
        return stringToInsert.text;

    const size_t initialBytes = (size_t) (((char*) insertPoint.getAddress()) - (char*) text.getAddress());
    const size_t newStringBytes = findByteOffsetOfEnd (stringToInsert);
    const size_t remainderBytes = (size_t) (((char*) startOfRemainder.findTerminatingNull().getAddress()) - (char*) startOfRemainder.getAddress());

    const size_t newTotalBytes = initialBytes + newStringBytes + remainderBytes;
    if (newTotalBytes <= 0)
        return String();

    String result (PreallocationBytes ((size_t) newTotalBytes));

    char* dest = (char*) result.text.getAddress();
    memcpy (dest, text.getAddress(), initialBytes);
    dest += initialBytes;
    memcpy (dest, stringToInsert.text.getAddress(), newStringBytes);
    dest += newStringBytes;
    memcpy (dest, startOfRemainder.getAddress(), remainderBytes);
    dest += remainderBytes;
    CharPointerType ((CharPointerType::CharType*) dest).writeNull();

    return result;
}

String String::replace (StringRef stringToReplace, StringRef stringToInsert, const bool ignoreCase) const
{
    const int stringToReplaceLen = stringToReplace.length();
    const int stringToInsertLen = stringToInsert.length();

    int i = 0;
    String result (*this);

    while ((i = (ignoreCase ? result.indexOfIgnoreCase (i, stringToReplace)
                            : result.indexOf (i, stringToReplace))) >= 0)
    {
        result = result.replaceSection (i, stringToReplaceLen, stringToInsert);
        i += stringToInsertLen;
    }

    return result;
}

class StringCreationHelper
{
public:
    StringCreationHelper (const size_t initialBytes)
        : source (nullptr), dest (nullptr), allocatedBytes (initialBytes), bytesWritten (0)
    {
        result.preallocateBytes (allocatedBytes);
        dest = result.getCharPointer();
    }

    StringCreationHelper (const String::CharPointerType s)
        : source (s), dest (nullptr), allocatedBytes (StringHolder::getAllocatedNumBytes (s)), bytesWritten (0)
    {
        result.preallocateBytes (allocatedBytes);
        dest = result.getCharPointer();
    }

    void write (water_uchar c)
    {
        bytesWritten += String::CharPointerType::getBytesRequiredFor (c);

        if (bytesWritten > allocatedBytes)
        {
            allocatedBytes += jmax ((size_t) 8, allocatedBytes / 16);
            const size_t destOffset = (size_t) (((char*) dest.getAddress()) - (char*) result.getCharPointer().getAddress());
            result.preallocateBytes (allocatedBytes);
            dest = addBytesToPointer (result.getCharPointer().getAddress(), (int) destOffset);
        }

        dest.write (c);
    }

    String result;
    String::CharPointerType source;

private:
    String::CharPointerType dest;
    size_t allocatedBytes, bytesWritten;
};

String String::replaceCharacter (const water_uchar charToReplace, const water_uchar charToInsert) const
{
    if (! containsChar (charToReplace))
        return *this;

    StringCreationHelper builder (text);

    for (;;)
    {
        water_uchar c = builder.source.getAndAdvance();

        if (c == charToReplace)
            c = charToInsert;

        builder.write (c);

        if (c == 0)
            break;
    }

    return builder.result;
}

String String::replaceCharacters (StringRef charactersToReplace, StringRef charactersToInsertInstead) const
{
    // Each character in the first string must have a matching one in the
    // second, so the two strings must be the same length.
    jassert (charactersToReplace.length() == charactersToInsertInstead.length());

    StringCreationHelper builder (text);

    for (;;)
    {
        water_uchar c = builder.source.getAndAdvance();

        const int index = charactersToReplace.text.indexOf (c);
        if (index >= 0)
            c = charactersToInsertInstead [index];

        builder.write (c);

        if (c == 0)
            break;
    }

    return builder.result;
}

//==============================================================================
bool String::startsWith (StringRef other) const noexcept
{
    return text.compareUpTo (other.text, other.length()) == 0;
}

bool String::startsWithIgnoreCase (StringRef other) const noexcept
{
    return text.compareIgnoreCaseUpTo (other.text, other.length()) == 0;
}

bool String::startsWithChar (const water_uchar character) const noexcept
{
    // strings can't contain a null character!
    CARLA_SAFE_ASSERT_RETURN (character != 0, false);

    return *text == character;
}

bool String::endsWithChar (const water_uchar character) const noexcept
{
    // strings can't contain a null character!
    CARLA_SAFE_ASSERT_RETURN (character != 0, false);

    if (text.isEmpty())
        return false;

    CharPointerType t (text.findTerminatingNull());
    return *--t == character;
}

bool String::endsWith (StringRef other) const noexcept
{
    CharPointerType end (text.findTerminatingNull());
    CharPointerType otherEnd (other.text.findTerminatingNull());

    while (end > text && otherEnd > other.text)
    {
        --end;
        --otherEnd;

        if (*end != *otherEnd)
            return false;
    }

    return otherEnd == other.text;
}

bool String::endsWithIgnoreCase (StringRef other) const noexcept
{
    CharPointerType end (text.findTerminatingNull());
    CharPointerType otherEnd (other.text.findTerminatingNull());

    while (end > text && otherEnd > other.text)
    {
        --end;
        --otherEnd;

        if (end.toLowerCase() != otherEnd.toLowerCase())
            return false;
    }

    return otherEnd == other.text;
}

//==============================================================================
String String::toUpperCase() const
{
    StringCreationHelper builder (text);

    for (;;)
    {
        const water_uchar c = builder.source.toUpperCase();
        builder.write (c);

        if (c == 0)
            break;

        ++(builder.source);
    }

    return builder.result;
}

String String::toLowerCase() const
{
    StringCreationHelper builder (text);

    for (;;)
    {
        const water_uchar c = builder.source.toLowerCase();
        builder.write (c);

        if (c == 0)
            break;

        ++(builder.source);
    }

    return builder.result;
}

//==============================================================================
water_uchar String::getLastCharacter() const noexcept
{
    return isEmpty() ? water_uchar() : text [length() - 1];
}

String String::substring (int start, const int end) const
{
    if (start < 0)
        start = 0;

    if (end <= start)
        return String();

    int i = 0;
    CharPointerType t1 (text);

    while (i < start)
    {
        if (t1.isEmpty())
            return String();

        ++i;
        ++t1;
    }

    CharPointerType t2 (t1);
    while (i < end)
    {
        if (t2.isEmpty())
        {
            if (start == 0)
                return *this;

            break;
        }

        ++i;
        ++t2;
    }

    return String (t1, t2);
}

String String::substring (int start) const
{
    if (start <= 0)
        return *this;

    CharPointerType t (text);

    while (--start >= 0)
    {
        if (t.isEmpty())
            return String();

        ++t;
    }

    return String (t);
}

String String::dropLastCharacters (const int numberToDrop) const
{
    return String (text, (size_t) jmax (0, length() - numberToDrop));
}

String String::getLastCharacters (const int numCharacters) const
{
    return String (text + jmax (0, length() - jmax (0, numCharacters)));
}

String String::fromFirstOccurrenceOf (StringRef sub,
                                      const bool includeSubString,
                                      const bool ignoreCase) const
{
    const int i = ignoreCase ? indexOfIgnoreCase (sub)
                             : indexOf (sub);
    if (i < 0)
        return String();

    return substring (includeSubString ? i : i + sub.length());
}

String String::fromLastOccurrenceOf (StringRef sub,
                                     const bool includeSubString,
                                     const bool ignoreCase) const
{
    const int i = ignoreCase ? lastIndexOfIgnoreCase (sub)
                             : lastIndexOf (sub);
    if (i < 0)
        return *this;

    return substring (includeSubString ? i : i + sub.length());
}

String String::upToFirstOccurrenceOf (StringRef sub,
                                      const bool includeSubString,
                                      const bool ignoreCase) const
{
    const int i = ignoreCase ? indexOfIgnoreCase (sub)
                             : indexOf (sub);
    if (i < 0)
        return *this;

    return substring (0, includeSubString ? i + sub.length() : i);
}

String String::upToLastOccurrenceOf (StringRef sub,
                                     const bool includeSubString,
                                     const bool ignoreCase) const
{
    const int i = ignoreCase ? lastIndexOfIgnoreCase (sub)
                             : lastIndexOf (sub);
    if (i < 0)
        return *this;

    return substring (0, includeSubString ? i + sub.length() : i);
}

bool String::isQuotedString() const
{
    const water_uchar trimmedStart = trimStart()[0];

    return trimmedStart == '"'
        || trimmedStart == '\'';
}

String String::unquoted() const
{
    const int len = length();

    if (len == 0)
        return String();

    const water_uchar lastChar = text [len - 1];
    const int dropAtStart = (*text == '"' || *text == '\'') ? 1 : 0;
    const int dropAtEnd = (lastChar == '"' || lastChar == '\'') ? 1 : 0;

    return substring (dropAtStart, len - dropAtEnd);
}

String String::quoted (water_uchar quoteCharacter) const
{
    if (isEmpty())
        return charToString (quoteCharacter) + quoteCharacter;

    String t (*this);

    if (! t.startsWithChar (quoteCharacter))
        t = charToString (quoteCharacter) + t;

    if (! t.endsWithChar (quoteCharacter))
        t += quoteCharacter;

    return t;
}

//==============================================================================
static String::CharPointerType findTrimmedEnd (const String::CharPointerType start,
                                               String::CharPointerType end)
{
    while (end > start)
    {
        if (! (--end).isWhitespace())
        {
            ++end;
            break;
        }
    }

    return end;
}

String String::trim() const
{
    if (isNotEmpty())
    {
        CharPointerType start (text.findEndOfWhitespace());

        const CharPointerType end (start.findTerminatingNull());
        CharPointerType trimmedEnd (findTrimmedEnd (start, end));

        if (trimmedEnd <= start)
            return String();

        if (text < start || trimmedEnd < end)
            return String (start, trimmedEnd);
    }

    return *this;
}

String String::trimStart() const
{
    if (isNotEmpty())
    {
        const CharPointerType t (text.findEndOfWhitespace());

        if (t != text)
            return String (t);
    }

    return *this;
}

String String::trimEnd() const
{
    if (isNotEmpty())
    {
        const CharPointerType end (text.findTerminatingNull());
        CharPointerType trimmedEnd (findTrimmedEnd (text, end));

        if (trimmedEnd < end)
            return String (text, trimmedEnd);
    }

    return *this;
}

String String::trimCharactersAtStart (StringRef charactersToTrim) const
{
    CharPointerType t (text);

    while (charactersToTrim.text.indexOf (*t) >= 0)
        ++t;

    return t == text ? *this : String (t);
}

String String::trimCharactersAtEnd (StringRef charactersToTrim) const
{
    if (isNotEmpty())
    {
        const CharPointerType end (text.findTerminatingNull());
        CharPointerType trimmedEnd (end);

        while (trimmedEnd > text)
        {
            if (charactersToTrim.text.indexOf (*--trimmedEnd) < 0)
            {
                ++trimmedEnd;
                break;
            }
        }

        if (trimmedEnd < end)
            return String (text, trimmedEnd);
    }

    return *this;
}

//==============================================================================
String String::retainCharacters (StringRef charactersToRetain) const
{
    if (isEmpty())
        return String();

    StringCreationHelper builder (text);

    for (;;)
    {
        water_uchar c = builder.source.getAndAdvance();

        if (charactersToRetain.text.indexOf (c) >= 0)
            builder.write (c);

        if (c == 0)
            break;
    }

    builder.write (0);
    return builder.result;
}

String String::removeCharacters (StringRef charactersToRemove) const
{
    if (isEmpty())
        return String();

    StringCreationHelper builder (text);

    for (;;)
    {
        water_uchar c = builder.source.getAndAdvance();

        if (charactersToRemove.text.indexOf (c) < 0)
            builder.write (c);

        if (c == 0)
            break;
    }

    return builder.result;
}

String String::initialSectionContainingOnly (StringRef permittedCharacters) const
{
    for (CharPointerType t (text); ! t.isEmpty(); ++t)
        if (permittedCharacters.text.indexOf (*t) < 0)
            return String (text, t);

    return *this;
}

String String::initialSectionNotContaining (StringRef charactersToStopAt) const
{
    for (CharPointerType t (text); ! t.isEmpty(); ++t)
        if (charactersToStopAt.text.indexOf (*t) >= 0)
            return String (text, t);

    return *this;
}

bool String::containsOnly (StringRef chars) const noexcept
{
    for (CharPointerType t (text); ! t.isEmpty();)
        if (chars.text.indexOf (t.getAndAdvance()) < 0)
            return false;

    return true;
}

bool String::containsAnyOf (StringRef chars) const noexcept
{
    for (CharPointerType t (text); ! t.isEmpty();)
        if (chars.text.indexOf (t.getAndAdvance()) >= 0)
            return true;

    return false;
}

bool String::containsNonWhitespaceChars() const noexcept
{
    for (CharPointerType t (text); ! t.isEmpty(); ++t)
        if (! t.isWhitespace())
            return true;

    return false;
}

//=====================================================================================================================
static String getStringFromWindows1252Codepage (const char* data, size_t num)
{
    HeapBlock<char> unicode;
    CARLA_SAFE_ASSERT_RETURN(unicode.malloc(num + 1), String());

    for (size_t i = 0; i < num; ++i)
        unicode[i] = CharacterFunctions::getUnicodeCharFromWindows1252Codepage ((uint8) data[i]);

    unicode[num] = 0;
    return CharPointer_UTF8 (unicode);
}

String String::createStringFromData (const void* const unknownData, int size)
{
    const uint8* const data = static_cast<const uint8*> (unknownData);

    if (size <= 0 || data == nullptr)
        return String();

    if (size == 1)
        return charToString ((water_uchar) data[0]);

    const char* start = (const char*) data;

    if (size >= 3 && CharPointer_UTF8::isByteOrderMark (data))
    {
        start += 3;
        size -= 3;
    }

    if (CharPointer_UTF8::isValidString (start, size))
        return String (CharPointer_UTF8 (start),
                       CharPointer_UTF8 (start + size));

    return getStringFromWindows1252Codepage (start, (size_t) size);
}

// Note! The format parameter here MUST NOT be a reference, otherwise MS's va_start macro fails to work (but still compiles).
String String::formatted (const String pf, ... )
{
    size_t bufferSize = 256;

    HeapBlock<char> temp;
    CARLA_SAFE_ASSERT_RETURN(temp.malloc(bufferSize), String());

    for (;;)
    {
        va_list args;
        va_start (args, pf);

        // FIXME - needed?
        temp.clear (bufferSize);

        const int num = vsnprintf (temp.getData(), bufferSize - 1, pf.toRawUTF8(), args);

        va_end (args);

        if (num > 0)
            return String (temp);

        bufferSize += 256;

        if (num == 0 || bufferSize > 65536) // the upper limit is a sanity check to avoid situations where vsnprintf repeatedly
            break;                          // returns -1 because of an error rather than because it needs more space.
    }

    return String();
}

//==============================================================================
int String::getIntValue() const noexcept            { return text.getIntValue32(); }
int64 String::getLargeIntValue() const noexcept     { return text.getIntValue64(); }
float String::getFloatValue() const noexcept        { return (float) getDoubleValue(); }
double String::getDoubleValue() const noexcept      { return text.getDoubleValue(); }

int String::getTrailingIntValue() const noexcept
{
    int n = 0;
    int mult = 1;
    CharPointerType t (text.findTerminatingNull());

    while (--t >= text)
    {
        if (! t.isDigit())
        {
            if (*t == '-')
                n = -n;

            break;
        }

        n += mult * (*t - '0');
        mult *= 10;
    }

    return n;
}

static const char hexDigits[] = "0123456789abcdef";

template <typename Type>
static String hexToString (Type v)
{
    String::CharPointerType::CharType buffer[32];
    String::CharPointerType::CharType* const end = buffer + numElementsInArray (buffer) - 1;
    String::CharPointerType::CharType* t = end;
    *t = 0;

    do
    {
        *--t = hexDigits [(int) (v & 15)];
        v >>= 4;

    } while (v != 0);

    return String (String::CharPointerType (t),
                   String::CharPointerType (end));
}

String String::toHexString (int number)       { return hexToString ((unsigned int) number); }
String String::toHexString (int64 number)     { return hexToString ((uint64) number); }
String String::toHexString (short number)     { return toHexString ((int) (unsigned short) number); }

String String::toHexString (const void* const d, const int size, const int groupSize)
{
    if (size <= 0)
        return String();

    int numChars = (size * 2) + 2;
    if (groupSize > 0)
        numChars += size / groupSize;

    String s (PreallocationBytes (sizeof (CharPointerType::CharType) * (size_t) numChars));

    const unsigned char* data = static_cast<const unsigned char*> (d);
    CharPointerType dest (s.text);

    for (int i = 0; i < size; ++i)
    {
        const unsigned char nextByte = *data++;
        dest.write ((water_uchar) hexDigits [nextByte >> 4]);
        dest.write ((water_uchar) hexDigits [nextByte & 0xf]);

        if (groupSize > 0 && (i % groupSize) == (groupSize - 1) && i < (size - 1))
            dest.write ((water_uchar) ' ');
    }

    dest.writeNull();
    return s;
}

int   String::getHexValue32() const noexcept    { return CharacterFunctions::HexParser<int>  ::parse (text); }
int64 String::getHexValue64() const noexcept    { return CharacterFunctions::HexParser<int64>::parse (text); }

//==============================================================================
static const water_uchar emptyChar = 0;

template <class CharPointerType_Src, class CharPointerType_Dest>
struct StringEncodingConverter
{
    static CharPointerType_Dest convert (const String& s)
    {
        String& source = const_cast<String&> (s);

        typedef typename CharPointerType_Dest::CharType DestChar;

        if (source.isEmpty())
            return CharPointerType_Dest (reinterpret_cast<const DestChar*> (&emptyChar));

        CharPointerType_Src text (source.getCharPointer());
        const size_t extraBytesNeeded = CharPointerType_Dest::getBytesRequiredFor (text) + sizeof (typename CharPointerType_Dest::CharType);
        const size_t endOffset = (text.sizeInBytes() + 3) & ~3u; // the new string must be word-aligned or many Windows
                                                                // functions will fail to read it correctly!
        source.preallocateBytes (endOffset + extraBytesNeeded);
        text = source.getCharPointer();

        void* const newSpace = addBytesToPointer (text.getAddress(), (int) endOffset);
        const CharPointerType_Dest extraSpace (static_cast<DestChar*> (newSpace));

       #ifdef DEBUG // (This just avoids spurious warnings from valgrind about the uninitialised bytes at the end of the buffer..)
        const size_t bytesToClear = (size_t) jmin ((int) extraBytesNeeded, 4);
        zeromem (addBytesToPointer (newSpace, extraBytesNeeded - bytesToClear), bytesToClear);
       #endif

        CharPointerType_Dest (extraSpace).writeAll (text);
        return extraSpace;
    }
};

template <>
struct StringEncodingConverter<CharPointer_UTF8, CharPointer_UTF8>
{
    static CharPointer_UTF8 convert (const String& source) noexcept   { return CharPointer_UTF8 ((CharPointer_UTF8::CharType*) source.getCharPointer().getAddress()); }
};

CharPointer_UTF8  String::toUTF8()  const { return StringEncodingConverter<CharPointerType, CharPointer_UTF8 >::convert (*this); }

const char* String::toRawUTF8() const
{
    return toUTF8().getAddress();
}

std::string String::toStdString() const
{
    return std::string (toRawUTF8());
}

//==============================================================================
template <class CharPointerType_Src, class CharPointerType_Dest>
struct StringCopier
{
    static size_t copyToBuffer (const CharPointerType_Src source, typename CharPointerType_Dest::CharType* const buffer, const size_t maxBufferSizeBytes)
    {
        jassert (((ssize_t) maxBufferSizeBytes) >= 0); // keep this value positive!

        if (buffer == nullptr)
            return CharPointerType_Dest::getBytesRequiredFor (source) + sizeof (typename CharPointerType_Dest::CharType);

        return CharPointerType_Dest (buffer).writeWithDestByteLimit (source, maxBufferSizeBytes);
    }
};

size_t String::copyToUTF8 (CharPointer_UTF8::CharType* const buffer, size_t maxBufferSizeBytes) const noexcept
{
    return StringCopier<CharPointerType, CharPointer_UTF8>::copyToBuffer (text, buffer, maxBufferSizeBytes);
}

//==============================================================================
size_t String::getNumBytesAsUTF8() const noexcept
{
    return CharPointer_UTF8::getBytesRequiredFor (text);
}

String String::fromUTF8 (const char* const buffer, int bufferSizeBytes)
{
    if (buffer != nullptr)
    {
        if (bufferSizeBytes < 0)
            return String (CharPointer_UTF8 (buffer));

        if (bufferSizeBytes > 0)
        {
            jassert (CharPointer_UTF8::isValidString (buffer, bufferSizeBytes));
            return String (CharPointer_UTF8 (buffer), CharPointer_UTF8 (buffer + bufferSizeBytes));
        }
    }

    return String();
}

#ifdef CARLA_OS_MAC

String String::convertToPrecomposedUnicode() const
{
#if 0
    UnicodeMapping map;

    map.unicodeEncoding = CreateTextEncoding (kTextEncodingUnicodeDefault,
                                              kUnicodeNoSubset,
                                              kTextEncodingDefaultFormat);

    map.otherEncoding = CreateTextEncoding (kTextEncodingUnicodeDefault,
                                            kUnicodeCanonicalCompVariant,
                                            kTextEncodingDefaultFormat);

    map.mappingVersion = kUnicodeUseLatestMapping;

    UnicodeToTextInfo conversionInfo = 0;
    String result;

    if (CreateUnicodeToTextInfo (&map, &conversionInfo) == noErr)
    {
        const size_t bytesNeeded = CharPointer_UTF16::getBytesRequiredFor (getCharPointer());

        HeapBlock<char> tempOut;
        tempOut.calloc (bytesNeeded + 4);

        ByteCount bytesRead = 0;
        ByteCount outputBufferSize = 0;

        if (ConvertFromUnicodeToText (conversionInfo,
                                      bytesNeeded, (ConstUniCharArrayPtr) toUTF16().getAddress(),
                                      kUnicodeDefaultDirectionMask,
                                      0, 0, 0, 0,
                                      bytesNeeded, &bytesRead,
                                      &outputBufferSize, tempOut) == noErr)
        {
            result = String (CharPointer_UTF16 ((CharPointer_UTF16::CharType*) tempOut.getData()));
        }

        DisposeUnicodeToTextInfo (&conversionInfo);
    }

    return result;
#else
    return *this;
#endif
}
#endif

//==============================================================================
StringRef::StringRef() noexcept  : text ((const String::CharPointerType::CharType*) "\0\0\0")
{
}

StringRef::StringRef (const char* stringLiteral) noexcept
    : text (stringLiteral)
{
    jassert (stringLiteral != nullptr); // This must be a valid string literal, not a null pointer!!

    /*  If you get an assertion here, then you're trying to create a string from 8-bit data
        that contains values greater than 127. These can NOT be correctly converted to unicode
        because there's no way for the String class to know what encoding was used to
        create them. The source data could be UTF-8, ASCII or one of many local code-pages.

        To get around this problem, you must be more explicit when you pass an ambiguous 8-bit
        string to the StringRef class - so for example if your source data is actually UTF-8,
        you'd call StringRef (CharPointer_UTF8 ("my utf8 string..")), and it would be able to
        correctly convert the multi-byte characters to unicode. It's *highly* recommended that
        you use UTF-8 with escape characters in your source code to represent extended characters,
        because there's no other way to represent these strings in a way that isn't dependent on
        the compiler, source code editor and platform.
    */
    jassert (CharPointer_UTF8::isValidString (stringLiteral, std::numeric_limits<int>::max()));
}

StringRef::StringRef (String::CharPointerType stringLiteral) noexcept  : text (stringLiteral)
{
    jassert (stringLiteral.getAddress() != nullptr); // This must be a valid string literal, not a null pointer!!
}

StringRef::StringRef (const String& string) noexcept  : text (string.getCharPointer()) {}

}

//==============================================================================