Source file src/encoding/json/decode.go

     1  // Copyright 2010 The Go Authors. All rights reserved.
     2  // Use of this source code is governed by a BSD-style
     3  // license that can be found in the LICENSE file.
     4  
     5  // Represents JSON data structure using native Go types: booleans, floats,
     6  // strings, arrays, and maps.
     7  
     8  //go:build !goexperiment.jsonv2
     9  
    10  package json
    11  
    12  import (
    13  	"encoding"
    14  	"encoding/base64"
    15  	"fmt"
    16  	"reflect"
    17  	"strconv"
    18  	"strings"
    19  	"unicode"
    20  	"unicode/utf16"
    21  	"unicode/utf8"
    22  )
    23  
    24  // Unmarshal parses the JSON-encoded data and stores the result
    25  // in the value pointed to by v. If v is nil or not a pointer,
    26  // Unmarshal returns an [InvalidUnmarshalError].
    27  //
    28  // Unmarshal uses the inverse of the encodings that
    29  // [Marshal] uses, allocating maps, slices, and pointers as necessary,
    30  // with the following additional rules:
    31  //
    32  // To unmarshal JSON into a pointer, Unmarshal first handles the case of
    33  // the JSON being the JSON literal null. In that case, Unmarshal sets
    34  // the pointer to nil. Otherwise, Unmarshal unmarshals the JSON into
    35  // the value pointed at by the pointer. If the pointer is nil, Unmarshal
    36  // allocates a new value for it to point to.
    37  //
    38  // To unmarshal JSON into a value implementing [Unmarshaler],
    39  // Unmarshal calls that value's [Unmarshaler.UnmarshalJSON] method, including
    40  // when the input is a JSON null.
    41  // Otherwise, if the value implements [encoding.TextUnmarshaler]
    42  // and the input is a JSON quoted string, Unmarshal calls
    43  // [encoding.TextUnmarshaler.UnmarshalText] with the unquoted form of the string.
    44  //
    45  // To unmarshal JSON into a struct, Unmarshal matches incoming object keys to
    46  // the keys used by [Marshal] (either the struct field name or its tag),
    47  // ignoring case. If multiple struct fields match an object key, an exact case
    48  // match is preferred over a case-insensitive one.
    49  //
    50  // Incoming object members are processed in the order observed. If an object
    51  // includes duplicate keys, later duplicates will replace or be merged into
    52  // prior values.
    53  //
    54  // To unmarshal JSON into an interface value,
    55  // Unmarshal stores one of these in the interface value:
    56  //
    57  //   - bool, for JSON booleans
    58  //   - float64, for JSON numbers
    59  //   - string, for JSON strings
    60  //   - []any, for JSON arrays
    61  //   - map[string]any, for JSON objects
    62  //   - nil for JSON null
    63  //
    64  // To unmarshal a JSON array into a slice, Unmarshal resets the slice length
    65  // to zero and then appends each element to the slice.
    66  // As a special case, to unmarshal an empty JSON array into a slice,
    67  // Unmarshal replaces the slice with a new empty slice.
    68  //
    69  // To unmarshal a JSON array into a Go array, Unmarshal decodes
    70  // JSON array elements into corresponding Go array elements.
    71  // If the Go array is smaller than the JSON array,
    72  // the additional JSON array elements are discarded.
    73  // If the JSON array is smaller than the Go array,
    74  // the additional Go array elements are set to zero values.
    75  //
    76  // To unmarshal a JSON object into a map, Unmarshal first establishes a map to
    77  // use. If the map is nil, Unmarshal allocates a new map. Otherwise Unmarshal
    78  // reuses the existing map, keeping existing entries. Unmarshal then stores
    79  // key-value pairs from the JSON object into the map. The map's key type must
    80  // either be any string type, an integer, or implement [encoding.TextUnmarshaler].
    81  //
    82  // If the JSON-encoded data contain a syntax error, Unmarshal returns a [SyntaxError].
    83  //
    84  // If a JSON value is not appropriate for a given target type,
    85  // or if a JSON number overflows the target type, Unmarshal
    86  // skips that field and completes the unmarshaling as best it can.
    87  // If no more serious errors are encountered, Unmarshal returns
    88  // an [UnmarshalTypeError] describing the earliest such error. In any
    89  // case, it's not guaranteed that all the remaining fields following
    90  // the problematic one will be unmarshaled into the target object.
    91  //
    92  // The JSON null value unmarshals into an interface, map, pointer, or slice
    93  // by setting that Go value to nil. Because null is often used in JSON to mean
    94  // “not present,” unmarshaling a JSON null into any other Go type has no effect
    95  // on the value and produces no error.
    96  //
    97  // When unmarshaling quoted strings, invalid UTF-8 or
    98  // invalid UTF-16 surrogate pairs are not treated as an error.
    99  // Instead, they are replaced by the Unicode replacement
   100  // character U+FFFD.
   101  func Unmarshal(data []byte, v any) error {
   102  	// Check for well-formedness.
   103  	// Avoids filling out half a data structure
   104  	// before discovering a JSON syntax error.
   105  	var d decodeState
   106  	err := checkValid(data, &d.scan)
   107  	if err != nil {
   108  		return err
   109  	}
   110  
   111  	d.init(data)
   112  	return d.unmarshal(v)
   113  }
   114  
   115  // Unmarshaler is the interface implemented by types
   116  // that can unmarshal a JSON description of themselves.
   117  // The input can be assumed to be a valid encoding of
   118  // a JSON value. UnmarshalJSON must copy the JSON data
   119  // if it wishes to retain the data after returning.
   120  type Unmarshaler interface {
   121  	UnmarshalJSON([]byte) error
   122  }
   123  
   124  // An UnmarshalTypeError describes a JSON value that was
   125  // not appropriate for a value of a specific Go type.
   126  type UnmarshalTypeError struct {
   127  	Value  string       // description of JSON value - "bool", "array", "number -5"
   128  	Type   reflect.Type // type of Go value it could not be assigned to
   129  	Offset int64        // error occurred after reading Offset bytes
   130  	Struct string       // name of the struct type containing the field
   131  	Field  string       // the full path from root node to the field, include embedded struct
   132  }
   133  
   134  func (e *UnmarshalTypeError) Error() string {
   135  	if e.Struct != "" || e.Field != "" {
   136  		return "json: cannot unmarshal " + e.Value + " into Go struct field " + e.Struct + "." + e.Field + " of type " + e.Type.String()
   137  	}
   138  	return "json: cannot unmarshal " + e.Value + " into Go value of type " + e.Type.String()
   139  }
   140  
   141  // An UnmarshalFieldError describes a JSON object key that
   142  // led to an unexported (and therefore unwritable) struct field.
   143  //
   144  // Deprecated: No longer used; kept for compatibility.
   145  type UnmarshalFieldError struct {
   146  	Key   string
   147  	Type  reflect.Type
   148  	Field reflect.StructField
   149  }
   150  
   151  func (e *UnmarshalFieldError) Error() string {
   152  	return "json: cannot unmarshal object key " + strconv.Quote(e.Key) + " into unexported field " + e.Field.Name + " of type " + e.Type.String()
   153  }
   154  
   155  // An InvalidUnmarshalError describes an invalid argument passed to [Unmarshal].
   156  // (The argument to [Unmarshal] must be a non-nil pointer.)
   157  type InvalidUnmarshalError struct {
   158  	Type reflect.Type
   159  }
   160  
   161  func (e *InvalidUnmarshalError) Error() string {
   162  	if e.Type == nil {
   163  		return "json: Unmarshal(nil)"
   164  	}
   165  
   166  	if e.Type.Kind() != reflect.Pointer {
   167  		return "json: Unmarshal(non-pointer " + e.Type.String() + ")"
   168  	}
   169  	return "json: Unmarshal(nil " + e.Type.String() + ")"
   170  }
   171  
   172  func (d *decodeState) unmarshal(v any) error {
   173  	rv := reflect.ValueOf(v)
   174  	if rv.Kind() != reflect.Pointer || rv.IsNil() {
   175  		return &InvalidUnmarshalError{reflect.TypeOf(v)}
   176  	}
   177  
   178  	d.scan.reset()
   179  	d.scanWhile(scanSkipSpace)
   180  	// We decode rv not rv.Elem because the Unmarshaler interface
   181  	// test must be applied at the top level of the value.
   182  	err := d.value(rv)
   183  	if err != nil {
   184  		return d.addErrorContext(err)
   185  	}
   186  	return d.savedError
   187  }
   188  
   189  // A Number represents a JSON number literal.
   190  type Number string
   191  
   192  // String returns the literal text of the number.
   193  func (n Number) String() string { return string(n) }
   194  
   195  // Float64 returns the number as a float64.
   196  func (n Number) Float64() (float64, error) {
   197  	return strconv.ParseFloat(string(n), 64)
   198  }
   199  
   200  // Int64 returns the number as an int64.
   201  func (n Number) Int64() (int64, error) {
   202  	return strconv.ParseInt(string(n), 10, 64)
   203  }
   204  
   205  // An errorContext provides context for type errors during decoding.
   206  type errorContext struct {
   207  	Struct     reflect.Type
   208  	FieldStack []string
   209  }
   210  
   211  // decodeState represents the state while decoding a JSON value.
   212  type decodeState struct {
   213  	data                  []byte
   214  	off                   int // next read offset in data
   215  	opcode                int // last read result
   216  	scan                  scanner
   217  	errorContext          *errorContext
   218  	savedError            error
   219  	useNumber             bool
   220  	disallowUnknownFields bool
   221  }
   222  
   223  // readIndex returns the position of the last byte read.
   224  func (d *decodeState) readIndex() int {
   225  	return d.off - 1
   226  }
   227  
   228  // phasePanicMsg is used as a panic message when we end up with something that
   229  // shouldn't happen. It can indicate a bug in the JSON decoder, or that
   230  // something is editing the data slice while the decoder executes.
   231  const phasePanicMsg = "JSON decoder out of sync - data changing underfoot?"
   232  
   233  func (d *decodeState) init(data []byte) *decodeState {
   234  	d.data = data
   235  	d.off = 0
   236  	d.savedError = nil
   237  	if d.errorContext != nil {
   238  		d.errorContext.Struct = nil
   239  		// Reuse the allocated space for the FieldStack slice.
   240  		d.errorContext.FieldStack = d.errorContext.FieldStack[:0]
   241  	}
   242  	return d
   243  }
   244  
   245  // saveError saves the first err it is called with,
   246  // for reporting at the end of the unmarshal.
   247  func (d *decodeState) saveError(err error) {
   248  	if d.savedError == nil {
   249  		d.savedError = d.addErrorContext(err)
   250  	}
   251  }
   252  
   253  // addErrorContext returns a new error enhanced with information from d.errorContext
   254  func (d *decodeState) addErrorContext(err error) error {
   255  	if d.errorContext != nil && (d.errorContext.Struct != nil || len(d.errorContext.FieldStack) > 0) {
   256  		switch err := err.(type) {
   257  		case *UnmarshalTypeError:
   258  			err.Struct = d.errorContext.Struct.Name()
   259  			fieldStack := d.errorContext.FieldStack
   260  			if err.Field != "" {
   261  				fieldStack = append(fieldStack, err.Field)
   262  			}
   263  			err.Field = strings.Join(fieldStack, ".")
   264  		}
   265  	}
   266  	return err
   267  }
   268  
   269  // skip scans to the end of what was started.
   270  func (d *decodeState) skip() {
   271  	s, data, i := &d.scan, d.data, d.off
   272  	depth := len(s.parseState)
   273  	for {
   274  		op := s.step(s, data[i])
   275  		i++
   276  		if len(s.parseState) < depth {
   277  			d.off = i
   278  			d.opcode = op
   279  			return
   280  		}
   281  	}
   282  }
   283  
   284  // scanNext processes the byte at d.data[d.off].
   285  func (d *decodeState) scanNext() {
   286  	if d.off < len(d.data) {
   287  		d.opcode = d.scan.step(&d.scan, d.data[d.off])
   288  		d.off++
   289  	} else {
   290  		d.opcode = d.scan.eof()
   291  		d.off = len(d.data) + 1 // mark processed EOF with len+1
   292  	}
   293  }
   294  
   295  // scanWhile processes bytes in d.data[d.off:] until it
   296  // receives a scan code not equal to op.
   297  func (d *decodeState) scanWhile(op int) {
   298  	s, data, i := &d.scan, d.data, d.off
   299  	for i < len(data) {
   300  		newOp := s.step(s, data[i])
   301  		i++
   302  		if newOp != op {
   303  			d.opcode = newOp
   304  			d.off = i
   305  			return
   306  		}
   307  	}
   308  
   309  	d.off = len(data) + 1 // mark processed EOF with len+1
   310  	d.opcode = d.scan.eof()
   311  }
   312  
   313  // rescanLiteral is similar to scanWhile(scanContinue), but it specialises the
   314  // common case where we're decoding a literal. The decoder scans the input
   315  // twice, once for syntax errors and to check the length of the value, and the
   316  // second to perform the decoding.
   317  //
   318  // Only in the second step do we use decodeState to tokenize literals, so we
   319  // know there aren't any syntax errors. We can take advantage of that knowledge,
   320  // and scan a literal's bytes much more quickly.
   321  func (d *decodeState) rescanLiteral() {
   322  	data, i := d.data, d.off
   323  Switch:
   324  	switch data[i-1] {
   325  	case '"': // string
   326  		for ; i < len(data); i++ {
   327  			switch data[i] {
   328  			case '\\':
   329  				i++ // escaped char
   330  			case '"':
   331  				i++ // tokenize the closing quote too
   332  				break Switch
   333  			}
   334  		}
   335  	case '0', '1', '2', '3', '4', '5', '6', '7', '8', '9', '-': // number
   336  		for ; i < len(data); i++ {
   337  			switch data[i] {
   338  			case '0', '1', '2', '3', '4', '5', '6', '7', '8', '9',
   339  				'.', 'e', 'E', '+', '-':
   340  			default:
   341  				break Switch
   342  			}
   343  		}
   344  	case 't': // true
   345  		i += len("rue")
   346  	case 'f': // false
   347  		i += len("alse")
   348  	case 'n': // null
   349  		i += len("ull")
   350  	}
   351  	if i < len(data) {
   352  		d.opcode = stateEndValue(&d.scan, data[i])
   353  	} else {
   354  		d.opcode = scanEnd
   355  	}
   356  	d.off = i + 1
   357  }
   358  
   359  // value consumes a JSON value from d.data[d.off-1:], decoding into v, and
   360  // reads the following byte ahead. If v is invalid, the value is discarded.
   361  // The first byte of the value has been read already.
   362  func (d *decodeState) value(v reflect.Value) error {
   363  	switch d.opcode {
   364  	default:
   365  		panic(phasePanicMsg)
   366  
   367  	case scanBeginArray:
   368  		if v.IsValid() {
   369  			if err := d.array(v); err != nil {
   370  				return err
   371  			}
   372  		} else {
   373  			d.skip()
   374  		}
   375  		d.scanNext()
   376  
   377  	case scanBeginObject:
   378  		if v.IsValid() {
   379  			if err := d.object(v); err != nil {
   380  				return err
   381  			}
   382  		} else {
   383  			d.skip()
   384  		}
   385  		d.scanNext()
   386  
   387  	case scanBeginLiteral:
   388  		// All bytes inside literal return scanContinue op code.
   389  		start := d.readIndex()
   390  		d.rescanLiteral()
   391  
   392  		if v.IsValid() {
   393  			if err := d.literalStore(d.data[start:d.readIndex()], v, false); err != nil {
   394  				return err
   395  			}
   396  		}
   397  	}
   398  	return nil
   399  }
   400  
   401  type unquotedValue struct{}
   402  
   403  // valueQuoted is like value but decodes a
   404  // quoted string literal or literal null into an interface value.
   405  // If it finds anything other than a quoted string literal or null,
   406  // valueQuoted returns unquotedValue{}.
   407  func (d *decodeState) valueQuoted() any {
   408  	switch d.opcode {
   409  	default:
   410  		panic(phasePanicMsg)
   411  
   412  	case scanBeginArray, scanBeginObject:
   413  		d.skip()
   414  		d.scanNext()
   415  
   416  	case scanBeginLiteral:
   417  		v := d.literalInterface()
   418  		switch v.(type) {
   419  		case nil, string:
   420  			return v
   421  		}
   422  	}
   423  	return unquotedValue{}
   424  }
   425  
   426  // indirect walks down v allocating pointers as needed,
   427  // until it gets to a non-pointer.
   428  // If it encounters an Unmarshaler, indirect stops and returns that.
   429  // If decodingNull is true, indirect stops at the first settable pointer so it
   430  // can be set to nil.
   431  func indirect(v reflect.Value, decodingNull bool) (Unmarshaler, encoding.TextUnmarshaler, reflect.Value) {
   432  	// Issue #24153 indicates that it is generally not a guaranteed property
   433  	// that you may round-trip a reflect.Value by calling Value.Addr().Elem()
   434  	// and expect the value to still be settable for values derived from
   435  	// unexported embedded struct fields.
   436  	//
   437  	// The logic below effectively does this when it first addresses the value
   438  	// (to satisfy possible pointer methods) and continues to dereference
   439  	// subsequent pointers as necessary.
   440  	//
   441  	// After the first round-trip, we set v back to the original value to
   442  	// preserve the original RW flags contained in reflect.Value.
   443  	v0 := v
   444  	haveAddr := false
   445  
   446  	// If v is a named type and is addressable,
   447  	// start with its address, so that if the type has pointer methods,
   448  	// we find them.
   449  	if v.Kind() != reflect.Pointer && v.Type().Name() != "" && v.CanAddr() {
   450  		haveAddr = true
   451  		v = v.Addr()
   452  	}
   453  	for {
   454  		// Load value from interface, but only if the result will be
   455  		// usefully addressable.
   456  		if v.Kind() == reflect.Interface && !v.IsNil() {
   457  			e := v.Elem()
   458  			if e.Kind() == reflect.Pointer && !e.IsNil() && (!decodingNull || e.Elem().Kind() == reflect.Pointer) {
   459  				haveAddr = false
   460  				v = e
   461  				continue
   462  			}
   463  		}
   464  
   465  		if v.Kind() != reflect.Pointer {
   466  			break
   467  		}
   468  
   469  		if decodingNull && v.CanSet() {
   470  			break
   471  		}
   472  
   473  		// Prevent infinite loop if v is an interface pointing to its own address:
   474  		//     var v any
   475  		//     v = &v
   476  		if v.Elem().Kind() == reflect.Interface && v.Elem().Elem().Equal(v) {
   477  			v = v.Elem()
   478  			break
   479  		}
   480  		if v.IsNil() {
   481  			v.Set(reflect.New(v.Type().Elem()))
   482  		}
   483  		if v.Type().NumMethod() > 0 && v.CanInterface() {
   484  			if u, ok := reflect.TypeAssert[Unmarshaler](v); ok {
   485  				return u, nil, reflect.Value{}
   486  			}
   487  			if !decodingNull {
   488  				if u, ok := reflect.TypeAssert[encoding.TextUnmarshaler](v); ok {
   489  					return nil, u, reflect.Value{}
   490  				}
   491  			}
   492  		}
   493  
   494  		if haveAddr {
   495  			v = v0 // restore original value after round-trip Value.Addr().Elem()
   496  			haveAddr = false
   497  		} else {
   498  			v = v.Elem()
   499  		}
   500  	}
   501  	return nil, nil, v
   502  }
   503  
   504  // array consumes an array from d.data[d.off-1:], decoding into v.
   505  // The first byte of the array ('[') has been read already.
   506  func (d *decodeState) array(v reflect.Value) error {
   507  	// Check for unmarshaler.
   508  	u, ut, pv := indirect(v, false)
   509  	if u != nil {
   510  		start := d.readIndex()
   511  		d.skip()
   512  		return u.UnmarshalJSON(d.data[start:d.off])
   513  	}
   514  	if ut != nil {
   515  		d.saveError(&UnmarshalTypeError{Value: "array", Type: v.Type(), Offset: int64(d.off)})
   516  		d.skip()
   517  		return nil
   518  	}
   519  	v = pv
   520  
   521  	// Check type of target.
   522  	switch v.Kind() {
   523  	case reflect.Interface:
   524  		if v.NumMethod() == 0 {
   525  			// Decoding into nil interface? Switch to non-reflect code.
   526  			ai := d.arrayInterface()
   527  			v.Set(reflect.ValueOf(ai))
   528  			return nil
   529  		}
   530  		// Otherwise it's invalid.
   531  		fallthrough
   532  	default:
   533  		d.saveError(&UnmarshalTypeError{Value: "array", Type: v.Type(), Offset: int64(d.off)})
   534  		d.skip()
   535  		return nil
   536  	case reflect.Array, reflect.Slice:
   537  		break
   538  	}
   539  
   540  	i := 0
   541  	for {
   542  		// Look ahead for ] - can only happen on first iteration.
   543  		d.scanWhile(scanSkipSpace)
   544  		if d.opcode == scanEndArray {
   545  			break
   546  		}
   547  
   548  		// Expand slice length, growing the slice if necessary.
   549  		if v.Kind() == reflect.Slice {
   550  			if i >= v.Cap() {
   551  				v.Grow(1)
   552  			}
   553  			if i >= v.Len() {
   554  				v.SetLen(i + 1)
   555  			}
   556  		}
   557  
   558  		if i < v.Len() {
   559  			// Decode into element.
   560  			if err := d.value(v.Index(i)); err != nil {
   561  				return err
   562  			}
   563  		} else {
   564  			// Ran out of fixed array: skip.
   565  			if err := d.value(reflect.Value{}); err != nil {
   566  				return err
   567  			}
   568  		}
   569  		i++
   570  
   571  		// Next token must be , or ].
   572  		if d.opcode == scanSkipSpace {
   573  			d.scanWhile(scanSkipSpace)
   574  		}
   575  		if d.opcode == scanEndArray {
   576  			break
   577  		}
   578  		if d.opcode != scanArrayValue {
   579  			panic(phasePanicMsg)
   580  		}
   581  	}
   582  
   583  	if i < v.Len() {
   584  		if v.Kind() == reflect.Array {
   585  			for ; i < v.Len(); i++ {
   586  				v.Index(i).SetZero() // zero remainder of array
   587  			}
   588  		} else {
   589  			v.SetLen(i) // truncate the slice
   590  		}
   591  	}
   592  	if i == 0 && v.Kind() == reflect.Slice {
   593  		v.Set(reflect.MakeSlice(v.Type(), 0, 0))
   594  	}
   595  	return nil
   596  }
   597  
   598  var nullLiteral = []byte("null")
   599  var textUnmarshalerType = reflect.TypeFor[encoding.TextUnmarshaler]()
   600  
   601  // object consumes an object from d.data[d.off-1:], decoding into v.
   602  // The first byte ('{') of the object has been read already.
   603  func (d *decodeState) object(v reflect.Value) error {
   604  	// Check for unmarshaler.
   605  	u, ut, pv := indirect(v, false)
   606  	if u != nil {
   607  		start := d.readIndex()
   608  		d.skip()
   609  		return u.UnmarshalJSON(d.data[start:d.off])
   610  	}
   611  	if ut != nil {
   612  		d.saveError(&UnmarshalTypeError{Value: "object", Type: v.Type(), Offset: int64(d.off)})
   613  		d.skip()
   614  		return nil
   615  	}
   616  	v = pv
   617  	t := v.Type()
   618  
   619  	// Decoding into nil interface? Switch to non-reflect code.
   620  	if v.Kind() == reflect.Interface && v.NumMethod() == 0 {
   621  		oi := d.objectInterface()
   622  		v.Set(reflect.ValueOf(oi))
   623  		return nil
   624  	}
   625  
   626  	var fields structFields
   627  
   628  	// Check type of target:
   629  	//   struct or
   630  	//   map[T1]T2 where T1 is string, an integer type,
   631  	//             or an encoding.TextUnmarshaler
   632  	switch v.Kind() {
   633  	case reflect.Map:
   634  		// Map key must either have string kind, have an integer kind,
   635  		// or be an encoding.TextUnmarshaler.
   636  		switch t.Key().Kind() {
   637  		case reflect.String,
   638  			reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64,
   639  			reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr:
   640  		default:
   641  			if !reflect.PointerTo(t.Key()).Implements(textUnmarshalerType) {
   642  				d.saveError(&UnmarshalTypeError{Value: "object", Type: t, Offset: int64(d.off)})
   643  				d.skip()
   644  				return nil
   645  			}
   646  		}
   647  		if v.IsNil() {
   648  			v.Set(reflect.MakeMap(t))
   649  		}
   650  	case reflect.Struct:
   651  		fields = cachedTypeFields(t)
   652  		// ok
   653  	default:
   654  		d.saveError(&UnmarshalTypeError{Value: "object", Type: t, Offset: int64(d.off)})
   655  		d.skip()
   656  		return nil
   657  	}
   658  
   659  	var mapElem reflect.Value
   660  	var origErrorContext errorContext
   661  	if d.errorContext != nil {
   662  		origErrorContext = *d.errorContext
   663  	}
   664  
   665  	for {
   666  		// Read opening " of string key or closing }.
   667  		d.scanWhile(scanSkipSpace)
   668  		if d.opcode == scanEndObject {
   669  			// closing } - can only happen on first iteration.
   670  			break
   671  		}
   672  		if d.opcode != scanBeginLiteral {
   673  			panic(phasePanicMsg)
   674  		}
   675  
   676  		// Read key.
   677  		start := d.readIndex()
   678  		d.rescanLiteral()
   679  		item := d.data[start:d.readIndex()]
   680  		key, ok := unquoteBytes(item)
   681  		if !ok {
   682  			panic(phasePanicMsg)
   683  		}
   684  
   685  		// Figure out field corresponding to key.
   686  		var subv reflect.Value
   687  		destring := false // whether the value is wrapped in a string to be decoded first
   688  
   689  		if v.Kind() == reflect.Map {
   690  			elemType := t.Elem()
   691  			if !mapElem.IsValid() {
   692  				mapElem = reflect.New(elemType).Elem()
   693  			} else {
   694  				mapElem.SetZero()
   695  			}
   696  			subv = mapElem
   697  		} else {
   698  			f := fields.byExactName[string(key)]
   699  			if f == nil {
   700  				f = fields.byFoldedName[string(foldName(key))]
   701  			}
   702  			if f != nil {
   703  				subv = v
   704  				destring = f.quoted
   705  				if d.errorContext == nil {
   706  					d.errorContext = new(errorContext)
   707  				}
   708  				for i, ind := range f.index {
   709  					if subv.Kind() == reflect.Pointer {
   710  						if subv.IsNil() {
   711  							// If a struct embeds a pointer to an unexported type,
   712  							// it is not possible to set a newly allocated value
   713  							// since the field is unexported.
   714  							//
   715  							// See https://golang.org/issue/21357
   716  							if !subv.CanSet() {
   717  								d.saveError(fmt.Errorf("json: cannot set embedded pointer to unexported struct: %v", subv.Type().Elem()))
   718  								// Invalidate subv to ensure d.value(subv) skips over
   719  								// the JSON value without assigning it to subv.
   720  								subv = reflect.Value{}
   721  								destring = false
   722  								break
   723  							}
   724  							subv.Set(reflect.New(subv.Type().Elem()))
   725  						}
   726  						subv = subv.Elem()
   727  					}
   728  					if i < len(f.index)-1 {
   729  						d.errorContext.FieldStack = append(
   730  							d.errorContext.FieldStack,
   731  							subv.Type().Field(ind).Name,
   732  						)
   733  					}
   734  					subv = subv.Field(ind)
   735  				}
   736  				d.errorContext.Struct = t
   737  				d.errorContext.FieldStack = append(d.errorContext.FieldStack, f.name)
   738  			} else if d.disallowUnknownFields {
   739  				d.saveError(fmt.Errorf("json: unknown field %q", key))
   740  			}
   741  		}
   742  
   743  		// Read : before value.
   744  		if d.opcode == scanSkipSpace {
   745  			d.scanWhile(scanSkipSpace)
   746  		}
   747  		if d.opcode != scanObjectKey {
   748  			panic(phasePanicMsg)
   749  		}
   750  		d.scanWhile(scanSkipSpace)
   751  
   752  		if destring {
   753  			switch qv := d.valueQuoted().(type) {
   754  			case nil:
   755  				if err := d.literalStore(nullLiteral, subv, false); err != nil {
   756  					return err
   757  				}
   758  			case string:
   759  				if err := d.literalStore([]byte(qv), subv, true); err != nil {
   760  					return err
   761  				}
   762  			default:
   763  				d.saveError(fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal unquoted value into %v", subv.Type()))
   764  			}
   765  		} else {
   766  			if err := d.value(subv); err != nil {
   767  				return err
   768  			}
   769  		}
   770  
   771  		// Write value back to map;
   772  		// if using struct, subv points into struct already.
   773  		if v.Kind() == reflect.Map {
   774  			kt := t.Key()
   775  			var kv reflect.Value
   776  			if reflect.PointerTo(kt).Implements(textUnmarshalerType) {
   777  				kv = reflect.New(kt)
   778  				if err := d.literalStore(item, kv, true); err != nil {
   779  					return err
   780  				}
   781  				kv = kv.Elem()
   782  			} else {
   783  				switch kt.Kind() {
   784  				case reflect.String:
   785  					kv = reflect.New(kt).Elem()
   786  					kv.SetString(string(key))
   787  				case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
   788  					s := string(key)
   789  					n, err := strconv.ParseInt(s, 10, 64)
   790  					if err != nil || kt.OverflowInt(n) {
   791  						d.saveError(&UnmarshalTypeError{Value: "number " + s, Type: kt, Offset: int64(start + 1)})
   792  						break
   793  					}
   794  					kv = reflect.New(kt).Elem()
   795  					kv.SetInt(n)
   796  				case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr:
   797  					s := string(key)
   798  					n, err := strconv.ParseUint(s, 10, 64)
   799  					if err != nil || kt.OverflowUint(n) {
   800  						d.saveError(&UnmarshalTypeError{Value: "number " + s, Type: kt, Offset: int64(start + 1)})
   801  						break
   802  					}
   803  					kv = reflect.New(kt).Elem()
   804  					kv.SetUint(n)
   805  				default:
   806  					panic("json: Unexpected key type") // should never occur
   807  				}
   808  			}
   809  			if kv.IsValid() {
   810  				v.SetMapIndex(kv, subv)
   811  			}
   812  		}
   813  
   814  		// Next token must be , or }.
   815  		if d.opcode == scanSkipSpace {
   816  			d.scanWhile(scanSkipSpace)
   817  		}
   818  		if d.errorContext != nil {
   819  			// Reset errorContext to its original state.
   820  			// Keep the same underlying array for FieldStack, to reuse the
   821  			// space and avoid unnecessary allocs.
   822  			d.errorContext.FieldStack = d.errorContext.FieldStack[:len(origErrorContext.FieldStack)]
   823  			d.errorContext.Struct = origErrorContext.Struct
   824  		}
   825  		if d.opcode == scanEndObject {
   826  			break
   827  		}
   828  		if d.opcode != scanObjectValue {
   829  			panic(phasePanicMsg)
   830  		}
   831  	}
   832  	return nil
   833  }
   834  
   835  // convertNumber converts the number literal s to a float64 or a Number
   836  // depending on the setting of d.useNumber.
   837  func (d *decodeState) convertNumber(s string) (any, error) {
   838  	if d.useNumber {
   839  		return Number(s), nil
   840  	}
   841  	f, err := strconv.ParseFloat(s, 64)
   842  	if err != nil {
   843  		return nil, &UnmarshalTypeError{Value: "number " + s, Type: reflect.TypeFor[float64](), Offset: int64(d.off)}
   844  	}
   845  	return f, nil
   846  }
   847  
   848  var numberType = reflect.TypeFor[Number]()
   849  
   850  // literalStore decodes a literal stored in item into v.
   851  //
   852  // fromQuoted indicates whether this literal came from unwrapping a
   853  // string from the ",string" struct tag option. this is used only to
   854  // produce more helpful error messages.
   855  func (d *decodeState) literalStore(item []byte, v reflect.Value, fromQuoted bool) error {
   856  	// Check for unmarshaler.
   857  	if len(item) == 0 {
   858  		// Empty string given.
   859  		d.saveError(fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type()))
   860  		return nil
   861  	}
   862  	isNull := item[0] == 'n' // null
   863  	u, ut, pv := indirect(v, isNull)
   864  	if u != nil {
   865  		return u.UnmarshalJSON(item)
   866  	}
   867  	if ut != nil {
   868  		if item[0] != '"' {
   869  			if fromQuoted {
   870  				d.saveError(fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type()))
   871  				return nil
   872  			}
   873  			val := "number"
   874  			switch item[0] {
   875  			case 'n':
   876  				val = "null"
   877  			case 't', 'f':
   878  				val = "bool"
   879  			}
   880  			d.saveError(&UnmarshalTypeError{Value: val, Type: v.Type(), Offset: int64(d.readIndex())})
   881  			return nil
   882  		}
   883  		s, ok := unquoteBytes(item)
   884  		if !ok {
   885  			if fromQuoted {
   886  				return fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type())
   887  			}
   888  			panic(phasePanicMsg)
   889  		}
   890  		return ut.UnmarshalText(s)
   891  	}
   892  
   893  	v = pv
   894  
   895  	switch c := item[0]; c {
   896  	case 'n': // null
   897  		// The main parser checks that only true and false can reach here,
   898  		// but if this was a quoted string input, it could be anything.
   899  		if fromQuoted && string(item) != "null" {
   900  			d.saveError(fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type()))
   901  			break
   902  		}
   903  		switch v.Kind() {
   904  		case reflect.Interface, reflect.Pointer, reflect.Map, reflect.Slice:
   905  			v.SetZero()
   906  			// otherwise, ignore null for primitives/string
   907  		}
   908  	case 't', 'f': // true, false
   909  		value := item[0] == 't'
   910  		// The main parser checks that only true and false can reach here,
   911  		// but if this was a quoted string input, it could be anything.
   912  		if fromQuoted && string(item) != "true" && string(item) != "false" {
   913  			d.saveError(fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type()))
   914  			break
   915  		}
   916  		switch v.Kind() {
   917  		default:
   918  			if fromQuoted {
   919  				d.saveError(fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type()))
   920  			} else {
   921  				d.saveError(&UnmarshalTypeError{Value: "bool", Type: v.Type(), Offset: int64(d.readIndex())})
   922  			}
   923  		case reflect.Bool:
   924  			v.SetBool(value)
   925  		case reflect.Interface:
   926  			if v.NumMethod() == 0 {
   927  				v.Set(reflect.ValueOf(value))
   928  			} else {
   929  				d.saveError(&UnmarshalTypeError{Value: "bool", Type: v.Type(), Offset: int64(d.readIndex())})
   930  			}
   931  		}
   932  
   933  	case '"': // string
   934  		s, ok := unquoteBytes(item)
   935  		if !ok {
   936  			if fromQuoted {
   937  				return fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type())
   938  			}
   939  			panic(phasePanicMsg)
   940  		}
   941  		switch v.Kind() {
   942  		default:
   943  			d.saveError(&UnmarshalTypeError{Value: "string", Type: v.Type(), Offset: int64(d.readIndex())})
   944  		case reflect.Slice:
   945  			if v.Type().Elem().Kind() != reflect.Uint8 {
   946  				d.saveError(&UnmarshalTypeError{Value: "string", Type: v.Type(), Offset: int64(d.readIndex())})
   947  				break
   948  			}
   949  			b := make([]byte, base64.StdEncoding.DecodedLen(len(s)))
   950  			n, err := base64.StdEncoding.Decode(b, s)
   951  			if err != nil {
   952  				d.saveError(err)
   953  				break
   954  			}
   955  			v.SetBytes(b[:n])
   956  		case reflect.String:
   957  			t := string(s)
   958  			if v.Type() == numberType && !isValidNumber(t) {
   959  				return fmt.Errorf("json: invalid number literal, trying to unmarshal %q into Number", item)
   960  			}
   961  			v.SetString(t)
   962  		case reflect.Interface:
   963  			if v.NumMethod() == 0 {
   964  				v.Set(reflect.ValueOf(string(s)))
   965  			} else {
   966  				d.saveError(&UnmarshalTypeError{Value: "string", Type: v.Type(), Offset: int64(d.readIndex())})
   967  			}
   968  		}
   969  
   970  	default: // number
   971  		if c != '-' && (c < '0' || c > '9') {
   972  			if fromQuoted {
   973  				return fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type())
   974  			}
   975  			panic(phasePanicMsg)
   976  		}
   977  		switch v.Kind() {
   978  		default:
   979  			if v.Kind() == reflect.String && v.Type() == numberType {
   980  				// s must be a valid number, because it's
   981  				// already been tokenized.
   982  				v.SetString(string(item))
   983  				break
   984  			}
   985  			if fromQuoted {
   986  				return fmt.Errorf("json: invalid use of ,string struct tag, trying to unmarshal %q into %v", item, v.Type())
   987  			}
   988  			d.saveError(&UnmarshalTypeError{Value: "number", Type: v.Type(), Offset: int64(d.readIndex())})
   989  		case reflect.Interface:
   990  			n, err := d.convertNumber(string(item))
   991  			if err != nil {
   992  				d.saveError(err)
   993  				break
   994  			}
   995  			if v.NumMethod() != 0 {
   996  				d.saveError(&UnmarshalTypeError{Value: "number", Type: v.Type(), Offset: int64(d.readIndex())})
   997  				break
   998  			}
   999  			v.Set(reflect.ValueOf(n))
  1000  
  1001  		case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
  1002  			n, err := strconv.ParseInt(string(item), 10, 64)
  1003  			if err != nil || v.OverflowInt(n) {
  1004  				d.saveError(&UnmarshalTypeError{Value: "number " + string(item), Type: v.Type(), Offset: int64(d.readIndex())})
  1005  				break
  1006  			}
  1007  			v.SetInt(n)
  1008  
  1009  		case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64, reflect.Uintptr:
  1010  			n, err := strconv.ParseUint(string(item), 10, 64)
  1011  			if err != nil || v.OverflowUint(n) {
  1012  				d.saveError(&UnmarshalTypeError{Value: "number " + string(item), Type: v.Type(), Offset: int64(d.readIndex())})
  1013  				break
  1014  			}
  1015  			v.SetUint(n)
  1016  
  1017  		case reflect.Float32, reflect.Float64:
  1018  			n, err := strconv.ParseFloat(string(item), v.Type().Bits())
  1019  			if err != nil || v.OverflowFloat(n) {
  1020  				d.saveError(&UnmarshalTypeError{Value: "number " + string(item), Type: v.Type(), Offset: int64(d.readIndex())})
  1021  				break
  1022  			}
  1023  			v.SetFloat(n)
  1024  		}
  1025  	}
  1026  	return nil
  1027  }
  1028  
  1029  // The xxxInterface routines build up a value to be stored
  1030  // in an empty interface. They are not strictly necessary,
  1031  // but they avoid the weight of reflection in this common case.
  1032  
  1033  // valueInterface is like value but returns any.
  1034  func (d *decodeState) valueInterface() (val any) {
  1035  	switch d.opcode {
  1036  	default:
  1037  		panic(phasePanicMsg)
  1038  	case scanBeginArray:
  1039  		val = d.arrayInterface()
  1040  		d.scanNext()
  1041  	case scanBeginObject:
  1042  		val = d.objectInterface()
  1043  		d.scanNext()
  1044  	case scanBeginLiteral:
  1045  		val = d.literalInterface()
  1046  	}
  1047  	return
  1048  }
  1049  
  1050  // arrayInterface is like array but returns []any.
  1051  func (d *decodeState) arrayInterface() []any {
  1052  	var v = make([]any, 0)
  1053  	for {
  1054  		// Look ahead for ] - can only happen on first iteration.
  1055  		d.scanWhile(scanSkipSpace)
  1056  		if d.opcode == scanEndArray {
  1057  			break
  1058  		}
  1059  
  1060  		v = append(v, d.valueInterface())
  1061  
  1062  		// Next token must be , or ].
  1063  		if d.opcode == scanSkipSpace {
  1064  			d.scanWhile(scanSkipSpace)
  1065  		}
  1066  		if d.opcode == scanEndArray {
  1067  			break
  1068  		}
  1069  		if d.opcode != scanArrayValue {
  1070  			panic(phasePanicMsg)
  1071  		}
  1072  	}
  1073  	return v
  1074  }
  1075  
  1076  // objectInterface is like object but returns map[string]any.
  1077  func (d *decodeState) objectInterface() map[string]any {
  1078  	m := make(map[string]any)
  1079  	for {
  1080  		// Read opening " of string key or closing }.
  1081  		d.scanWhile(scanSkipSpace)
  1082  		if d.opcode == scanEndObject {
  1083  			// closing } - can only happen on first iteration.
  1084  			break
  1085  		}
  1086  		if d.opcode != scanBeginLiteral {
  1087  			panic(phasePanicMsg)
  1088  		}
  1089  
  1090  		// Read string key.
  1091  		start := d.readIndex()
  1092  		d.rescanLiteral()
  1093  		item := d.data[start:d.readIndex()]
  1094  		key, ok := unquote(item)
  1095  		if !ok {
  1096  			panic(phasePanicMsg)
  1097  		}
  1098  
  1099  		// Read : before value.
  1100  		if d.opcode == scanSkipSpace {
  1101  			d.scanWhile(scanSkipSpace)
  1102  		}
  1103  		if d.opcode != scanObjectKey {
  1104  			panic(phasePanicMsg)
  1105  		}
  1106  		d.scanWhile(scanSkipSpace)
  1107  
  1108  		// Read value.
  1109  		m[key] = d.valueInterface()
  1110  
  1111  		// Next token must be , or }.
  1112  		if d.opcode == scanSkipSpace {
  1113  			d.scanWhile(scanSkipSpace)
  1114  		}
  1115  		if d.opcode == scanEndObject {
  1116  			break
  1117  		}
  1118  		if d.opcode != scanObjectValue {
  1119  			panic(phasePanicMsg)
  1120  		}
  1121  	}
  1122  	return m
  1123  }
  1124  
  1125  // literalInterface consumes and returns a literal from d.data[d.off-1:] and
  1126  // it reads the following byte ahead. The first byte of the literal has been
  1127  // read already (that's how the caller knows it's a literal).
  1128  func (d *decodeState) literalInterface() any {
  1129  	// All bytes inside literal return scanContinue op code.
  1130  	start := d.readIndex()
  1131  	d.rescanLiteral()
  1132  
  1133  	item := d.data[start:d.readIndex()]
  1134  
  1135  	switch c := item[0]; c {
  1136  	case 'n': // null
  1137  		return nil
  1138  
  1139  	case 't', 'f': // true, false
  1140  		return c == 't'
  1141  
  1142  	case '"': // string
  1143  		s, ok := unquote(item)
  1144  		if !ok {
  1145  			panic(phasePanicMsg)
  1146  		}
  1147  		return s
  1148  
  1149  	default: // number
  1150  		if c != '-' && (c < '0' || c > '9') {
  1151  			panic(phasePanicMsg)
  1152  		}
  1153  		n, err := d.convertNumber(string(item))
  1154  		if err != nil {
  1155  			d.saveError(err)
  1156  		}
  1157  		return n
  1158  	}
  1159  }
  1160  
  1161  // getu4 decodes \uXXXX from the beginning of s, returning the hex value,
  1162  // or it returns -1.
  1163  func getu4(s []byte) rune {
  1164  	if len(s) < 6 || s[0] != '\\' || s[1] != 'u' {
  1165  		return -1
  1166  	}
  1167  	var r rune
  1168  	for _, c := range s[2:6] {
  1169  		switch {
  1170  		case '0' <= c && c <= '9':
  1171  			c = c - '0'
  1172  		case 'a' <= c && c <= 'f':
  1173  			c = c - 'a' + 10
  1174  		case 'A' <= c && c <= 'F':
  1175  			c = c - 'A' + 10
  1176  		default:
  1177  			return -1
  1178  		}
  1179  		r = r*16 + rune(c)
  1180  	}
  1181  	return r
  1182  }
  1183  
  1184  // unquote converts a quoted JSON string literal s into an actual string t.
  1185  // The rules are different than for Go, so cannot use strconv.Unquote.
  1186  func unquote(s []byte) (t string, ok bool) {
  1187  	s, ok = unquoteBytes(s)
  1188  	t = string(s)
  1189  	return
  1190  }
  1191  
  1192  func unquoteBytes(s []byte) (t []byte, ok bool) {
  1193  	if len(s) < 2 || s[0] != '"' || s[len(s)-1] != '"' {
  1194  		return
  1195  	}
  1196  	s = s[1 : len(s)-1]
  1197  
  1198  	// Check for unusual characters. If there are none,
  1199  	// then no unquoting is needed, so return a slice of the
  1200  	// original bytes.
  1201  	r := 0
  1202  	for r < len(s) {
  1203  		c := s[r]
  1204  		if c == '\\' || c == '"' || c < ' ' {
  1205  			break
  1206  		}
  1207  		rr, size := utf8.DecodeRune(s[r:])
  1208  		if rr == utf8.RuneError && size == 1 {
  1209  			break
  1210  		}
  1211  		r += size
  1212  	}
  1213  	if r == len(s) {
  1214  		return s, true
  1215  	}
  1216  
  1217  	b := make([]byte, len(s)+2*utf8.UTFMax)
  1218  	w := copy(b, s[0:r])
  1219  	for r < len(s) {
  1220  		// Out of room? Can only happen if s is full of
  1221  		// malformed UTF-8 and we're replacing each
  1222  		// byte with RuneError.
  1223  		if w >= len(b)-2*utf8.UTFMax {
  1224  			nb := make([]byte, (len(b)+utf8.UTFMax)*2)
  1225  			copy(nb, b[0:w])
  1226  			b = nb
  1227  		}
  1228  		switch c := s[r]; {
  1229  		case c == '\\':
  1230  			r++
  1231  			if r >= len(s) {
  1232  				return
  1233  			}
  1234  			switch s[r] {
  1235  			default:
  1236  				return
  1237  			case '"', '\\', '/', '\'':
  1238  				b[w] = s[r]
  1239  				r++
  1240  				w++
  1241  			case 'b':
  1242  				b[w] = '\b'
  1243  				r++
  1244  				w++
  1245  			case 'f':
  1246  				b[w] = '\f'
  1247  				r++
  1248  				w++
  1249  			case 'n':
  1250  				b[w] = '\n'
  1251  				r++
  1252  				w++
  1253  			case 'r':
  1254  				b[w] = '\r'
  1255  				r++
  1256  				w++
  1257  			case 't':
  1258  				b[w] = '\t'
  1259  				r++
  1260  				w++
  1261  			case 'u':
  1262  				r--
  1263  				rr := getu4(s[r:])
  1264  				if rr < 0 {
  1265  					return
  1266  				}
  1267  				r += 6
  1268  				if utf16.IsSurrogate(rr) {
  1269  					rr1 := getu4(s[r:])
  1270  					if dec := utf16.DecodeRune(rr, rr1); dec != unicode.ReplacementChar {
  1271  						// A valid pair; consume.
  1272  						r += 6
  1273  						w += utf8.EncodeRune(b[w:], dec)
  1274  						break
  1275  					}
  1276  					// Invalid surrogate; fall back to replacement rune.
  1277  					rr = unicode.ReplacementChar
  1278  				}
  1279  				w += utf8.EncodeRune(b[w:], rr)
  1280  			}
  1281  
  1282  		// Quote, control characters are invalid.
  1283  		case c == '"', c < ' ':
  1284  			return
  1285  
  1286  		// ASCII
  1287  		case c < utf8.RuneSelf:
  1288  			b[w] = c
  1289  			r++
  1290  			w++
  1291  
  1292  		// Coerce to well-formed UTF-8.
  1293  		default:
  1294  			rr, size := utf8.DecodeRune(s[r:])
  1295  			r += size
  1296  			w += utf8.EncodeRune(b[w:], rr)
  1297  		}
  1298  	}
  1299  	return b[0:w], true
  1300  }
  1301  

View as plain text