adjustLastDigit modifies d = x-currentDiff*ε, to get closest to d = x-targetDiff*ε, without becoming smaller than x-maxDiff*ε. It assumes that a decim

adjustLastDigit modifies d = x-currentDiff*ε, to get closest to d = x-targetDiff*ε, without becoming smaller than x-maxDiff*ε. It assumes that a decim

emptyInterface creates a new instance of the object sent It the returned interface is writable and contains the zero value.

extractFields returns a list of fields that JSON should recognize for the given type. The algorithm is breadth-first search over the set of structs to

%e: -d.ddddde±dd

%e: -d.ddddde±dd

%f: -ddddddd.ddddd

%f: -ddddddd.ddddd

formatBits computes the string representation of u in the given base. If neg is set, u is treated as negative int64 value.

Frexp10 is an analogue of math.Frexp for decimal powers. It scales f by an approximate power of ten 10^-exp, and returns exp10, so that f*10^exp10 has

Frexp10 is an analogue of math.Frexp for decimal powers. It scales f by an approximate power of ten 10^-exp, and returns exp10, so that f*10^exp10 has

getu4 decodes \uXXXX from the beginning of s, returning the hex value, or it returns -1.

We check if the last field is conditional.

Binary shift left (/ 2) by k bits. k <= maxShift to avoid overflow.

Binary shift left (/ 2) by k bits. k <= maxShift to avoid overflow.

readFloat reads a decimal mantissa and exponent from a float string representation. It sets ok to false if the number could not fit return types or is

readSlice is like ReadBytes but returns a reference to internal buffer data.

Binary shift right (* 2) by k bits. k <= maxShift to avoid overflow.

Binary shift right (* 2) by k bits. k <= maxShift to avoid overflow.

If we chop a at nd digits, should we round up?

If we chop a at nd digits, should we round up?

AssignComputeBounds sets f to the floating point value defined by mant, exp and precision given by flt. It returns lower, upper such that any number i

AssignComputeBounds sets f to the floating point value defined by mant, exp and precision given by flt. It returns lower, upper such that any number i

Captures an entire field value, including recursive objects, and converts them to a []byte suitable to pass to a sub-object's UnmarshalJSON

FixedDecimal stores in d the first n significant digits of the decimal representation of f. It returns false if it cannot be sure of the answer.

FixedDecimal stores in d the first n significant digits of the decimal representation of f. It returns false if it cannot be sure of the answer.

FormatBits2 computes the string representation of u in the given base. If neg is set, u is treated as negative int64 value. If append_ is set, the str

Marshal will act the same way as json.Marshal, except it will choose the ffjson marshal function before falling back to using json.Marshal. Using this

ParseUint is like ParseInt but for unsigned numbers, and oeprating on []byte

ParseUint is like ParseInt but for unsigned numbers, and oeprating on []byte

Send a buffer to the Pool to reuse for other instances. On servers where you have a lot of concurrent encoding going on, you can hand back the byte b

Read reads the next len(p) bytes from the buffer or until the buffer is drained. The return value n is the number of bytes read. If the buffer has n

Extract integer part, rounded appropriately. No guarantees about overflow.

ShortestDecimal stores in d the shortest decimal representation of f which belongs to the open interval (lower, upper), where f is supposed to lie. It

ShortestDecimal stores in d the shortest decimal representation of f which belongs to the open interval (lower, upper), where f is supposed to lie. It

TempFile creates a new temporary file in the directory dir with a name beginning with prefix, opens the file for reading and writing, and returns the

Unmarshal will act the same way as json.Unmarshal, except it will choose the ffjson unmarshal function before falling back to using json.Unmarshal. Th

adjustLastDigitFixed assumes d contains the representation of the integral part of some number, whose fractional part is num / (den << shift). The num

adjustLastDigitFixed assumes d contains the representation of the integral part of some number, whose fractional part is num / (den << shift). The num

If possible to compute mantissa*10^exp to 32-bit float f exactly, entirely in floating-point math, do so, avoiding the machinery above.

If possible to convert decimal representation to 64-bit float f exactly, entirely in floating-point math, do so, avoiding the expense of decimalToFloa

Return the first number n such that n*base >= 1<<64.

dominantField looks through the fields, all of which are known to have the same name, to find the single field that dominates the others using Go's em

%b: -ddddddddp±ddd

%b: -ddddddddp+ddd

frexp10Many applies a common shift by a power of ten to a, b, c.

frexp10Many applies a common shift by a power of ten to a, b, c.

Return a time no later than 5000 years from unix datum. JSON cannot handle dates after year 9999.