You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

750 lines
21 KiB

#ifndef RAPIDJSON_READER_H_
#define RAPIDJSON_READER_H_
// Copyright (c) 2011 Milo Yip (miloyip@gmail.com)
// Version 0.1
#include "rapidjson.h"
#include "internal/pow10.h"
#include "internal/stack.h"
#include <csetjmp>
// All part of denormalized parsing
#include <limits> // for numeric_limits
#include <cmath> // for fpclassify
#include <sstream>
#ifdef RAPIDJSON_SSE42
#include <nmmintrin.h>
#elif defined(RAPIDJSON_SSE2)
#include <emmintrin.h>
#endif
#ifdef _MSC_VER
#pragma warning(push)
#pragma warning(disable : 4127) // conditional expression is constant
#pragma warning(disable : 4702) // uncreachable code
#endif
#ifndef RAPIDJSON_PARSE_ERROR
#define RAPIDJSON_PARSE_ERROR(msg, offset) \
RAPIDJSON_MULTILINEMACRO_BEGIN \
parseError_ = msg; \
errorOffset_ = offset; \
longjmp(jmpbuf_, 1); \
RAPIDJSON_MULTILINEMACRO_END
#endif
namespace rapidjson {
///////////////////////////////////////////////////////////////////////////////
// ParseFlag
//! Combination of parseFlags
enum ParseFlag {
kParseDefaultFlags = 0, //!< Default parse flags. Non-destructive parsing. Text strings are decoded into allocated buffer.
kParseInsituFlag = 1 //!< In-situ(destructive) parsing.
};
///////////////////////////////////////////////////////////////////////////////
// Handler
/*! \class rapidjson::Handler
\brief Concept for receiving events from GenericReader upon parsing.
\code
concept Handler {
typename Ch;
void Null_();
void Bool_(bool b);
void Int(int i);
void Uint(unsigned i);
void Int64(int64_t i);
void Uint64(uint64_t i);
void Double(double d);
void String(const Ch* str, SizeType length, bool copy);
void StartObject();
void EndObject(SizeType memberCount);
void StartArray();
void EndArray(SizeType elementCount);
};
\endcode
*/
///////////////////////////////////////////////////////////////////////////////
// BaseReaderHandler
//! Default implementation of Handler.
/*! This can be used as base class of any reader handler.
\implements Handler
*/
template<typename Encoding = UTF8<> >
struct BaseReaderHandler {
typedef typename Encoding::Ch Ch;
void Default() {}
void Null_() { Default(); }
void Bool_(bool) { Default(); }
void Int(int) { Default(); }
void Uint(unsigned) { Default(); }
void Int64(int64_t) { Default(); }
void Uint64(uint64_t) { Default(); }
void Double(double) { Default(); }
void String(const Ch*, SizeType, bool) { Default(); }
void StartObject() { Default(); }
void EndObject(SizeType) { Default(); }
void StartArray() { Default(); }
void EndArray(SizeType) { Default(); }
};
///////////////////////////////////////////////////////////////////////////////
// SkipWhitespace
//! Skip the JSON white spaces in a stream.
/*! \param stream A input stream for skipping white spaces.
\note This function has SSE2/SSE4.2 specialization.
*/
template<typename Stream>
void SkipWhitespace(Stream& stream) {
Stream s = stream; // Use a local copy for optimization
while (s.Peek() == ' ' || s.Peek() == '\n' || s.Peek() == '\r' || s.Peek() == '\t')
s.Take();
stream = s;
}
#ifdef RAPIDJSON_SSE42
//! Skip whitespace with SSE 4.2 pcmpistrm instruction, testing 16 8-byte characters at once.
inline const char *SkipWhitespace_SIMD(const char* p) {
static const char whitespace[16] = " \n\r\t";
__m128i w = _mm_loadu_si128((const __m128i *)&whitespace[0]);
for (;;) {
__m128i s = _mm_loadu_si128((const __m128i *)p);
unsigned r = _mm_cvtsi128_si32(_mm_cmpistrm(w, s, _SIDD_UBYTE_OPS | _SIDD_CMP_EQUAL_ANY | _SIDD_BIT_MASK | _SIDD_NEGATIVE_POLARITY));
if (r == 0) // all 16 characters are whitespace
p += 16;
else { // some of characters may be non-whitespace
#ifdef _MSC_VER // Find the index of first non-whitespace
unsigned long offset;
if (_BitScanForward(&offset, r))
return p + offset;
#else
if (r != 0)
return p + __builtin_ffs(r) - 1;
#endif
}
}
}
#elif defined(RAPIDJSON_SSE2)
//! Skip whitespace with SSE2 instructions, testing 16 8-byte characters at once.
inline const char *SkipWhitespace_SIMD(const char* p) {
static const char whitespaces[4][17] = {
" ",
"\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n\n",
"\r\r\r\r\r\r\r\r\r\r\r\r\r\r\r\r",
"\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t\t"};
__m128i w0 = _mm_loadu_si128((const __m128i *)&whitespaces[0][0]);
__m128i w1 = _mm_loadu_si128((const __m128i *)&whitespaces[1][0]);
__m128i w2 = _mm_loadu_si128((const __m128i *)&whitespaces[2][0]);
__m128i w3 = _mm_loadu_si128((const __m128i *)&whitespaces[3][0]);
for (;;) {
__m128i s = _mm_loadu_si128((const __m128i *)p);
__m128i x = _mm_cmpeq_epi8(s, w0);
x = _mm_or_si128(x, _mm_cmpeq_epi8(s, w1));
x = _mm_or_si128(x, _mm_cmpeq_epi8(s, w2));
x = _mm_or_si128(x, _mm_cmpeq_epi8(s, w3));
unsigned short r = ~_mm_movemask_epi8(x);
if (r == 0) // all 16 characters are whitespace
p += 16;
else { // some of characters may be non-whitespace
#ifdef _MSC_VER // Find the index of first non-whitespace
unsigned long offset;
if (_BitScanForward(&offset, r))
return p + offset;
#else
if (r != 0)
return p + __builtin_ffs(r) - 1;
#endif
}
}
}
#endif // RAPIDJSON_SSE2
#ifdef RAPIDJSON_SIMD
//! Template function specialization for InsituStringStream
template<> inline void SkipWhitespace(InsituStringStream& stream) {
stream.src_ = const_cast<char*>(SkipWhitespace_SIMD(stream.src_));
}
//! Template function specialization for StringStream
template<> inline void SkipWhitespace(StringStream& stream) {
stream.src_ = SkipWhitespace_SIMD(stream.src_);
}
#endif // RAPIDJSON_SIMD
///////////////////////////////////////////////////////////////////////////////
// GenericReader
//! SAX-style JSON parser. Use Reader for UTF8 encoding and default allocator.
/*! GenericReader parses JSON text from a stream, and send events synchronously to an
object implementing Handler concept.
It needs to allocate a stack for storing a single decoded string during
non-destructive parsing.
For in-situ parsing, the decoded string is directly written to the source
text string, no temporary buffer is required.
A GenericReader object can be reused for parsing multiple JSON text.
\tparam Encoding Encoding of both the stream and the parse output.
\tparam Allocator Allocator type for stack.
*/
template <typename Encoding, typename Allocator = MemoryPoolAllocator<> >
class GenericReader {
public:
typedef typename Encoding::Ch Ch;
//! Constructor.
/*! \param allocator Optional allocator for allocating stack memory. (Only use for non-destructive parsing)
\param stackCapacity stack capacity in bytes for storing a single decoded string. (Only use for non-destructive parsing)
*/
GenericReader(Allocator* allocator = 0, size_t stackCapacity = kDefaultStackCapacity) : stack_(allocator, stackCapacity), parseError_(0), errorOffset_(0) {}
//! Parse JSON text.
/*! \tparam parseFlags Combination of ParseFlag.
\tparam Stream Type of input stream.
\tparam Handler Type of handler which must implement Handler concept.
\param stream Input stream to be parsed.
\param handler The handler to receive events.
\return Whether the parsing is successful.
*/
template <unsigned parseFlags, typename Stream, typename Handler>
bool Parse(Stream& stream, Handler& handler) {
parseError_ = 0;
errorOffset_ = 0;
#ifdef _MSC_VER
#pragma warning(push)
#pragma warning(disable : 4611) // interaction between '_setjmp' and C++ object destruction is non-portable
#endif
if (setjmp(jmpbuf_)) {
#ifdef _MSC_VER
#pragma warning(pop)
#endif
stack_.Clear();
return false;
}
SkipWhitespace(stream);
if (stream.Peek() == '\0')
RAPIDJSON_PARSE_ERROR("Text only contains white space(s)", stream.Tell());
else {
switch (stream.Peek()) {
case '{': ParseObject<parseFlags>(stream, handler); break;
case '[': ParseArray<parseFlags>(stream, handler); break;
default: RAPIDJSON_PARSE_ERROR("Expect either an object or array at root", stream.Tell());
}
SkipWhitespace(stream);
if (stream.Peek() != '\0' && stream.Peek() != static_cast<Ch>(std::char_traits<Ch>::eof()))
RAPIDJSON_PARSE_ERROR("Nothing should follow the root object or array.", stream.Tell());
}
return true;
}
bool HasParseError() const { return parseError_ != 0; }
const char* GetParseError() const { return parseError_; }
size_t GetErrorOffset() const { return errorOffset_; }
private:
// Parse object: { string : value, ... }
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseObject(Stream& stream, Handler& handler) {
RAPIDJSON_ASSERT(stream.Peek() == '{');
stream.Take(); // Skip '{'
handler.StartObject();
SkipWhitespace(stream);
if (stream.Peek() == '}') {
stream.Take();
handler.EndObject(0); // empty object
return;
}
for (SizeType memberCount = 0;;) {
if (stream.Peek() != '"') {
RAPIDJSON_PARSE_ERROR("Name of an object member must be a string", stream.Tell());
break;
}
ParseString<parseFlags>(stream, handler);
SkipWhitespace(stream);
if (stream.Take() != ':') {
RAPIDJSON_PARSE_ERROR("There must be a colon after the name of object member", stream.Tell());
break;
}
SkipWhitespace(stream);
ParseValue<parseFlags>(stream, handler);
SkipWhitespace(stream);
++memberCount;
switch(stream.Take()) {
case ',': SkipWhitespace(stream); break;
case '}': handler.EndObject(memberCount); return;
default: RAPIDJSON_PARSE_ERROR("Must be a comma or '}' after an object member", stream.Tell());
}
}
}
// Parse array: [ value, ... ]
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseArray(Stream& stream, Handler& handler) {
RAPIDJSON_ASSERT(stream.Peek() == '[');
stream.Take(); // Skip '['
handler.StartArray();
SkipWhitespace(stream);
if (stream.Peek() == ']') {
stream.Take();
handler.EndArray(0); // empty array
return;
}
for (SizeType elementCount = 0;;) {
ParseValue<parseFlags>(stream, handler);
++elementCount;
SkipWhitespace(stream);
switch (stream.Take()) {
case ',': SkipWhitespace(stream); break;
case ']': handler.EndArray(elementCount); return;
default: RAPIDJSON_PARSE_ERROR("Must be a comma or ']' after an array element.", stream.Tell());
}
}
}
// Parses for null or NaN
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseNaNNull_(Stream& stream, Handler& handler) {
RAPIDJSON_ASSERT(stream.Peek() == 'n');
stream.Take();
if( stream.Peek() == 'a' && stream.Take() == 'a' && stream.Take() == 'n' )
handler.Double( std::numeric_limits<double>::quiet_NaN() );
else if (stream.Take() == 'u' && stream.Take() == 'l' && stream.Take() == 'l')
handler.Null_();
else
RAPIDJSON_PARSE_ERROR("Invalid value", stream.Tell() - 1);
}
// Parses for infinity
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseInfinity(Stream& stream, Handler& handler) {
RAPIDJSON_ASSERT(stream.Peek() == 'i');
stream.Take();
if (stream.Take() == 'n' && stream.Take() == 'f')
handler.Double( std::numeric_limits<double>::infinity() );
else
RAPIDJSON_PARSE_ERROR("Invalid value", stream.Tell() - 1);
}
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseTrue(Stream& stream, Handler& handler) {
RAPIDJSON_ASSERT(stream.Peek() == 't');
stream.Take();
if (stream.Take() == 'r' && stream.Take() == 'u' && stream.Take() == 'e')
handler.Bool_(true);
else
RAPIDJSON_PARSE_ERROR("Invalid value", stream.Tell());
}
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseFalse(Stream& stream, Handler& handler) {
RAPIDJSON_ASSERT(stream.Peek() == 'f');
stream.Take();
if (stream.Take() == 'a' && stream.Take() == 'l' && stream.Take() == 's' && stream.Take() == 'e')
handler.Bool_(false);
else
RAPIDJSON_PARSE_ERROR("Invalid value", stream.Tell() - 1);
}
// Helper function to parse four hexidecimal digits in \uXXXX in ParseString().
template<typename Stream>
unsigned ParseHex4(Stream& stream) {
Stream s = stream; // Use a local copy for optimization
unsigned codepoint = 0;
for (int i = 0; i < 4; i++) {
Ch c = s.Take();
codepoint <<= 4;
codepoint += c;
if (c >= '0' && c <= '9')
codepoint -= '0';
else if (c >= 'A' && c <= 'F')
codepoint -= 'A' - 10;
else if (c >= 'a' && c <= 'f')
codepoint -= 'a' - 10;
else
RAPIDJSON_PARSE_ERROR("Incorrect hex digit after \\u escape", s.Tell() - 1);
}
stream = s; // Restore stream
return codepoint;
}
// cereal Temporary until constexpr support is added in RTM
#ifdef _MSC_VER
template <class Ch>
bool characterOk( Ch c )
{
return c < 256;
}
template <>
bool characterOk<char>( char )
{
return true;
}
#else
// As part of a fix for GCC 4.7
template <class T>
static constexpr int to_int( T t ){ return t; }
template<class Ch>
typename std::enable_if < to_int(std::numeric_limits<Ch>::max()) < to_int(256), bool>::type
characterOk( Ch )
{
return true;
}
template<class Ch>
typename std::enable_if< to_int(std::numeric_limits<Ch>::max()) >= to_int(256), bool>::type
characterOk(Ch c)
{ return c < 256; }
#endif
// Parse string, handling the prefix and suffix double quotes and escaping.
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseString(Stream& stream, Handler& handler) {
#define Z16 0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0
static const Ch escape[256] = {
Z16, Z16, 0, 0,'\"', 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,'/',
Z16, Z16, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,'\\', 0, 0, 0,
0, 0,'\b', 0, 0, 0,'\f', 0, 0, 0, 0, 0, 0, 0,'\n', 0,
0, 0,'\r', 0,'\t', 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
Z16, Z16, Z16, Z16, Z16, Z16, Z16, Z16
};
#undef Z16
Stream s = stream; // Use a local copy for optimization
RAPIDJSON_ASSERT(s.Peek() == '\"');
s.Take(); // Skip '\"'
Ch *head;
SizeType len;
if (parseFlags & kParseInsituFlag)
head = s.PutBegin();
else
len = 0;
#define RAPIDJSON_PUT(x) \
do { \
if (parseFlags & kParseInsituFlag) \
s.Put(x); \
else { \
*stack_.template Push<Ch>() = x; \
++len; \
} \
} while(false)
for (;;) {
Ch c = s.Take();
if (c == '\\') { // Escape
Ch e = s.Take();
if ((sizeof(Ch) == 1 || characterOk(e)) && escape[(unsigned char)e])
RAPIDJSON_PUT(escape[(unsigned char)e]);
else if (e == 'u') { // Unicode
unsigned codepoint = ParseHex4(s);
if (codepoint >= 0xD800 && codepoint <= 0xDBFF) { // Handle UTF-16 surrogate pair
if (s.Take() != '\\' || s.Take() != 'u') {
RAPIDJSON_PARSE_ERROR("Missing the second \\u in surrogate pair", s.Tell() - 2);
return;
}
unsigned codepoint2 = ParseHex4(s);
if (codepoint2 < 0xDC00 || codepoint2 > 0xDFFF) {
RAPIDJSON_PARSE_ERROR("The second \\u in surrogate pair is invalid", s.Tell() - 2);
return;
}
codepoint = (((codepoint - 0xD800) << 10) | (codepoint2 - 0xDC00)) + 0x10000;
}
Ch buffer[4];
SizeType count = SizeType(Encoding::Encode(buffer, codepoint) - &buffer[0]);
if (parseFlags & kParseInsituFlag)
for (SizeType i = 0; i < count; i++)
s.Put(buffer[i]);
else {
memcpy(stack_.template Push<Ch>(count), buffer, count * sizeof(Ch));
len += count;
}
}
else {
RAPIDJSON_PARSE_ERROR("Unknown escape character", stream.Tell() - 1);
return;
}
}
else if (c == '"') { // Closing double quote
if (parseFlags & kParseInsituFlag) {
size_t length = s.PutEnd(head);
RAPIDJSON_ASSERT(length <= 0xFFFFFFFF);
RAPIDJSON_PUT('\0'); // null-terminate the string
handler.String(head, SizeType(length), false);
}
else {
RAPIDJSON_PUT('\0');
handler.String(stack_.template Pop<Ch>(len), len - 1, true);
}
stream = s; // restore stream
return;
}
else if (c == '\0') {
RAPIDJSON_PARSE_ERROR("lacks ending quotation before the end of string", stream.Tell() - 1);
return;
}
else if ((unsigned)c < 0x20) { // RFC 4627: unescaped = %x20-21 / %x23-5B / %x5D-10FFFF
RAPIDJSON_PARSE_ERROR("Incorrect unescaped character in string", stream.Tell() - 1);
return;
}
else
RAPIDJSON_PUT(c); // Normal character, just copy
}
#undef RAPIDJSON_PUT
}
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseNumber(Stream& stream, Handler& handler) {
Stream s = stream; // Local copy for optimization
// Parse minus
bool minus = false;
if (s.Peek() == '-') {
minus = true;
s.Take();
}
// Parse int: zero / ( digit1-9 *DIGIT )
unsigned i;
bool try64bit = false;
if (s.Peek() == '0') {
i = 0;
s.Take();
}
else if (s.Peek() >= '1' && s.Peek() <= '9') {
i = s.Take() - '0';
if (minus)
while (s.Peek() >= '0' && s.Peek() <= '9') {
if (i >= 214748364) { // 2^31 = 2147483648
if (i != 214748364 || s.Peek() > '8') {
try64bit = true;
break;
}
}
i = i * 10 + (s.Take() - '0');
}
else
while (s.Peek() >= '0' && s.Peek() <= '9') {
if (i >= 429496729) { // 2^32 - 1 = 4294967295
if (i != 429496729 || s.Peek() > '5') {
try64bit = true;
break;
}
}
i = i * 10 + (s.Take() - '0');
}
}
else {
RAPIDJSON_PARSE_ERROR("Expect a value here.", stream.Tell());
return;
}
// Parse 64bit int
uint64_t i64 = 0;
bool useDouble = false;
if (try64bit) {
i64 = i;
if (minus)
while (s.Peek() >= '0' && s.Peek() <= '9') {
if (i64 >= 922337203685477580uLL) // 2^63 = 9223372036854775808
if (i64 != 922337203685477580uLL || s.Peek() > '8') {
useDouble = true;
break;
}
i64 = i64 * 10 + (s.Take() - '0');
}
else
while (s.Peek() >= '0' && s.Peek() <= '9') {
if (i64 >= 1844674407370955161uLL) // 2^64 - 1 = 18446744073709551615
if (i64 != 1844674407370955161uLL || s.Peek() > '5') {
useDouble = true;
break;
}
i64 = i64 * 10 + (s.Take() - '0');
}
}
// Force double for big integer
double d = 0.0;
if (useDouble) {
d = (double)i64;
while (s.Peek() >= '0' && s.Peek() <= '9') {
if (d >= 1E307) {
RAPIDJSON_PARSE_ERROR("Number too big to store in double", stream.Tell());
return;
}
d = d * 10 + (s.Take() - '0');
}
}
// Parse frac = decimal-point 1*DIGIT
int expFrac = 0;
if (s.Peek() == '.') {
if (!useDouble) {
d = try64bit ? (double)i64 : (double)i;
useDouble = true;
}
s.Take();
if (s.Peek() >= '0' && s.Peek() <= '9') {
d = d * 10 + (s.Take() - '0');
--expFrac;
}
else {
RAPIDJSON_PARSE_ERROR("At least one digit in fraction part", stream.Tell());
return;
}
while (s.Peek() >= '0' && s.Peek() <= '9') {
if (expFrac > -20) {
d = d * 10 + (s.Peek() - '0');
--expFrac;
}
s.Take();
}
}
// Parse exp = e [ minus / plus ] 1*DIGIT
int exp = 0;
if (s.Peek() == 'e' || s.Peek() == 'E') {
if (!useDouble) {
d = try64bit ? (double)i64 : (double)i;
useDouble = true;
}
s.Take();
bool expMinus = false;
if (s.Peek() == '+')
s.Take();
else if (s.Peek() == '-') {
s.Take();
expMinus = true;
}
if (s.Peek() >= '0' && s.Peek() <= '9') {
exp = s.Take() - '0';
while (s.Peek() >= '0' && s.Peek() <= '9') {
exp = exp * 10 + (s.Take() - '0');
if (exp > 308) {
// Attempt denormalized construction
std::stringstream ss;
ss.precision( std::numeric_limits<double>::max_digits10 );
ss << d * internal::Pow10(expFrac) << 'e' << (expMinus ? '-' : '+') << exp;
double dd;
ss >> dd;
if( std::fpclassify( dd ) == FP_SUBNORMAL )
handler.Double( dd );
else
RAPIDJSON_PARSE_ERROR("Number too big to store in double", stream.Tell());
return;
}
}
}
else {
RAPIDJSON_PARSE_ERROR("At least one digit in exponent", s.Tell());
return;
}
if (expMinus)
exp = -exp;
}
// Finish parsing, call event according to the type of number.
if (useDouble) {
d *= internal::Pow10(exp + expFrac);
handler.Double(minus ? -d : d);
}
else {
if (try64bit) {
if (minus)
handler.Int64(-(int64_t)i64);
else
handler.Uint64(i64);
}
else {
if (minus)
handler.Int(-(int)i);
else
handler.Uint(i);
}
}
stream = s; // restore stream
}
// Parse any JSON value
template<unsigned parseFlags, typename Stream, typename Handler>
void ParseValue(Stream& stream, Handler& handler) {
switch (stream.Peek()) {
case 'n': ParseNaNNull_ <parseFlags>(stream, handler); break;
case 'i': ParseInfinity <parseFlags>(stream, handler); break;
case 't': ParseTrue <parseFlags>(stream, handler); break;
case 'f': ParseFalse <parseFlags>(stream, handler); break;
case '"': ParseString <parseFlags>(stream, handler); break;
case '{': ParseObject <parseFlags>(stream, handler); break;
case '[': ParseArray <parseFlags>(stream, handler); break;
default : ParseNumber <parseFlags>(stream, handler);
}
}
static const size_t kDefaultStackCapacity = 256; //!< Default stack capacity in bytes for storing a single decoded string.
internal::Stack<Allocator> stack_; //!< A stack for storing decoded string temporarily during non-destructive parsing.
jmp_buf jmpbuf_; //!< setjmp buffer for fast exit from nested parsing function calls.
const char* parseError_;
size_t errorOffset_;
}; // class GenericReader
//! Reader with UTF8 encoding and default allocator.
typedef GenericReader<UTF8<> > Reader;
} // namespace rapidjson
#ifdef _MSC_VER
#pragma warning(pop)
#endif
#endif // RAPIDJSON_READER_H_