AutoUTFInputStream< CharType, InputByteStream > Class Template Reference

Input stream wrapper with dynamically bound encoding and automatic encoding detection. More...

#include <encodedstream.h>

Public Types
typedef CharType	Ch

Public Member Functions
	AutoUTFInputStream (InputByteStream &is, UTFType type=kUTF8)
	Constructor. More...
UTFType	GetType () const
bool	HasBOM () const
Ch	Peek () const
Ch	Take ()
size_t	Tell () const
void	Put (Ch)
void	Flush ()
Ch *	PutBegin ()
size_t	PutEnd (Ch *)

Private Types
typedef Ch(*	TakeFunc) (InputByteStream &is)

Private Member Functions
	RAPIDJSON_STATIC_ASSERT (sizeof(typename InputByteStream::Ch)==1)
	AutoUTFInputStream (const AutoUTFInputStream &)
AutoUTFInputStream &	operator= (const AutoUTFInputStream &)
void	DetectType ()

Private Attributes
InputByteStream *	is_
UTFType	type_
Ch	current_
TakeFunc	takeFunc_
bool	hasBOM_

Detailed Description

template<typename CharType, typename InputByteStream>
class AutoUTFInputStream< CharType, InputByteStream >

Input stream wrapper with dynamically bound encoding and automatic encoding detection.

Template Parameters

CharType	Type of character for reading.
InputByteStream	type of input byte stream to be wrapped.

Definition at line 135 of file encodedstream.h.

Member Typedef Documentation

template<typename CharType , typename InputByteStream >

typedef CharType AutoUTFInputStream< CharType, InputByteStream >::Ch

Definition at line 138 of file encodedstream.h.

template<typename CharType , typename InputByteStream >

typedef Ch(* AutoUTFInputStream< CharType, InputByteStream >::TakeFunc) (InputByteStream &is)

private

Definition at line 219 of file encodedstream.h.

Constructor & Destructor Documentation

template<typename CharType , typename InputByteStream >

AutoUTFInputStream< CharType, InputByteStream >::AutoUTFInputStream ( InputByteStream &  is, UTFType  type = kUTF8  )

inline

Constructor.

Parameters

is	input stream to be wrapped.
type	UTF encoding type if it is not detected from the stream.

Definition at line 145 of file encodedstream.h.

                                                                  : is_(&is), type_(type), hasBOM_(false) {
        RAPIDJSON_ASSERT(type >= kUTF8 && type <= kUTF32BE);        
        DetectType();
        static const TakeFunc f[] = { RAPIDJSON_ENCODINGS_FUNC(Take) };
        takeFunc_ = f[type_];
        current_ = takeFunc_(*is_);
    }

template<typename CharType , typename InputByteStream >

AutoUTFInputStream< CharType, InputByteStream >::AutoUTFInputStream ( const AutoUTFInputStream< CharType, InputByteStream > &  )

private

Member Function Documentation

template<typename CharType , typename InputByteStream >

void AutoUTFInputStream< CharType, InputByteStream >::DetectType ( )

inlineprivate

Definition at line 171 of file encodedstream.h.

                      {
        // BOM (Byte Order Mark):
        // 00 00 FE FF  UTF-32BE
        // FF FE 00 00  UTF-32LE
        // FE FF        UTF-16BE
        // FF FE        UTF-16LE
        // EF BB BF     UTF-8

        const unsigned char* c = reinterpret_cast<const unsigned char *>(is_->Peek4());
        if (!c)
            return;

        unsigned bom = static_cast<unsigned>(c[0] | (c[1] << 8) | (c[2] << 16) | (c[3] << 24));
        hasBOM_ = false;
        if (bom == 0xFFFE0000)                  { type_ = kUTF32BE; hasBOM_ = true; is_->Take(); is_->Take(); is_->Take(); is_->Take(); }
        else if (bom == 0x0000FEFF)             { type_ = kUTF32LE; hasBOM_ = true; is_->Take(); is_->Take(); is_->Take(); is_->Take(); }
        else if ((bom & 0xFFFF) == 0xFFFE)      { type_ = kUTF16BE; hasBOM_ = true; is_->Take(); is_->Take();                           }
        else if ((bom & 0xFFFF) == 0xFEFF)      { type_ = kUTF16LE; hasBOM_ = true; is_->Take(); is_->Take();                           }
        else if ((bom & 0xFFFFFF) == 0xBFBBEF)  { type_ = kUTF8;    hasBOM_ = true; is_->Take(); is_->Take(); is_->Take();              }

        // RFC 4627: Section 3
        // "Since the first two characters of a JSON text will always be ASCII
        // characters [RFC0020], it is possible to determine whether an octet
        // stream is UTF-8, UTF-16 (BE or LE), or UTF-32 (BE or LE) by looking
        // at the pattern of nulls in the first four octets."
        // 00 00 00 xx  UTF-32BE
        // 00 xx 00 xx  UTF-16BE
        // xx 00 00 00  UTF-32LE
        // xx 00 xx 00  UTF-16LE
        // xx xx xx xx  UTF-8

        if (!hasBOM_) {
            int pattern = (c[0] ? 1 : 0) | (c[1] ? 2 : 0) | (c[2] ? 4 : 0) | (c[3] ? 8 : 0);
            switch (pattern) {
            case 0x08: type_ = kUTF32BE; break;
            case 0x0A: type_ = kUTF16BE; break;
            case 0x01: type_ = kUTF32LE; break;
            case 0x05: type_ = kUTF16LE; break;
            case 0x0F: type_ = kUTF8;    break;
            default: break; // Use type defined by user.
            }
        }

        // Runtime check whether the size of character type is sufficient. It only perform checks with assertion.
        if (type_ == kUTF16LE || type_ == kUTF16BE) RAPIDJSON_ASSERT(sizeof(Ch) >= 2);
        if (type_ == kUTF32LE || type_ == kUTF32BE) RAPIDJSON_ASSERT(sizeof(Ch) >= 4);
    }

template<typename CharType , typename InputByteStream >

void AutoUTFInputStream< CharType, InputByteStream >::Flush ( )

inline

Definition at line 162 of file encodedstream.h.

{ RAPIDJSON_ASSERT(false); } 

template<typename CharType , typename InputByteStream >

UTFType AutoUTFInputStream< CharType, InputByteStream >::GetType ( ) const

inline

Definition at line 153 of file encodedstream.h.

{ return type_; }

template<typename CharType , typename InputByteStream >

bool AutoUTFInputStream< CharType, InputByteStream >::HasBOM ( ) const

inline

Definition at line 154 of file encodedstream.h.

{ return hasBOM_; }

template<typename CharType , typename InputByteStream >

AutoUTFInputStream& AutoUTFInputStream< CharType, InputByteStream >::operator= ( const AutoUTFInputStream< CharType, InputByteStream > &  )

private

template<typename CharType , typename InputByteStream >

Ch AutoUTFInputStream< CharType, InputByteStream >::Peek ( ) const

inline

Definition at line 156 of file encodedstream.h.

{ return current_; }

template<typename CharType , typename InputByteStream >

void AutoUTFInputStream< CharType, InputByteStream >::Put ( Ch  )

inline

Definition at line 161 of file encodedstream.h.

{ RAPIDJSON_ASSERT(false); }

template<typename CharType , typename InputByteStream >

Ch* AutoUTFInputStream< CharType, InputByteStream >::PutBegin ( )

inline

Definition at line 163 of file encodedstream.h.

{ RAPIDJSON_ASSERT(false); return 0; }

template<typename CharType , typename InputByteStream >

size_t AutoUTFInputStream< CharType, InputByteStream >::PutEnd ( Ch *  )

inline

Definition at line 164 of file encodedstream.h.

{ RAPIDJSON_ASSERT(false); return 0; }

template<typename CharType , typename InputByteStream >

AutoUTFInputStream< CharType, InputByteStream >::RAPIDJSON_STATIC_ASSERT ( sizeof(typename InputByteStream::Ch)  = =1 )

private

template<typename CharType , typename InputByteStream >

Ch AutoUTFInputStream< CharType, InputByteStream >::Take ( )

inline

Definition at line 157 of file encodedstream.h.

{ Ch c = current_; current_ = takeFunc_(*is_); return c; }

template<typename CharType , typename InputByteStream >

size_t AutoUTFInputStream< CharType, InputByteStream >::Tell ( ) const

inline

Definition at line 158 of file encodedstream.h.

{ return is_->Tell(); }

Member Data Documentation

template<typename CharType , typename InputByteStream >

Ch AutoUTFInputStream< CharType, InputByteStream >::current_

private

Definition at line 222 of file encodedstream.h.

template<typename CharType , typename InputByteStream >

bool AutoUTFInputStream< CharType, InputByteStream >::hasBOM_

private

Definition at line 224 of file encodedstream.h.

template<typename CharType , typename InputByteStream >

InputByteStream* AutoUTFInputStream< CharType, InputByteStream >::is_

private

Definition at line 220 of file encodedstream.h.

template<typename CharType , typename InputByteStream >

TakeFunc AutoUTFInputStream< CharType, InputByteStream >::takeFunc_

private

Definition at line 223 of file encodedstream.h.

template<typename CharType , typename InputByteStream >

UTFType AutoUTFInputStream< CharType, InputByteStream >::type_

private

Definition at line 221 of file encodedstream.h.

---

The documentation for this class was generated from the following file:

• art/rapidjson/encodedstream.h