field declaration. An empty tag string is equivalent to an absent tag.
The tags are made visible through a <a href="/pkg/reflect/#StructTag">reflection interface</a>
and take part in <a href="#Type_identity">type identity</a> for structs
but are otherwise ignored.
</p>

<pre>
struct {
	x, y float64 ""  // an empty tag string is like an absent tag
	name string  "any string is permitted as a tag"

	// If the first bit is set, then all following bits constitute a two's
	// complement encoded number in big-endian byte order.
	if len(b) > 0 && b[0]&0x80 != 0 {
		// Handling negative numbers relies on the following identity:
		//	-a-1 == ^a
		//
		// If the number is negative, we use an inversion mask to invert the
		// data bytes and treat the value as an unsigned number.
		var inv byte // 0x00 if positive or zero, 0xff if negative

field declaration. An empty tag string is equivalent to an absent tag.
The tags are made visible through a <a href="/pkg/reflect/#StructTag">reflection interface</a>
and take part in <a href="#Type_identity">type identity</a> for structs
but are otherwise ignored.
</p>

<pre>
struct {
	x, y float64 ""  // an empty tag string is like an absent tag
	name string  "any string is permitted as a tag"

C source.

Cgo first invokes gcc -E -dM on the preamble, in order to find out
about simple #defines for constants and the like. These are recorded
for later use.

Next, cgo needs to identify the kinds for each identifier. For the
identifiers C.foo, cgo generates this C program:

	<preamble>
	#line 1 "not-declared"
	void __cgo_f_1_1(void) { __typeof__(foo) *__cgo_undefined__1; }
	#line 1 "not-type"

)

func bytediff(a, b []byte) string {
	const (
		uniqueA  = "-  "
		uniqueB  = "+  "
		identity = "   "
	)
	var ss []string
	sa := strings.Split(strings.TrimSpace(hex.Dump(a)), "\n")
	sb := strings.Split(strings.TrimSpace(hex.Dump(b)), "\n")
	for len(sa) > 0 && len(sb) > 0 {
		if sa[0] == sb[0] {
			ss = append(ss, identity+sa[0])
		} else {
			ss = append(ss, uniqueA+sa[0])
			ss = append(ss, uniqueB+sb[0])
		}

	for _, n := range f.Name {
		if n.Kind == "func" {
			functions[n.Go] = false
		}
	}

	// Now that we have all the name types filled in,
	// scan through the Refs to identify the ones that
	// are trying to do a ,err call. Also check that
	// functions are only used in calls.
	for _, r := range f.Ref {
		if r.Name.IsConst() && r.Name.Const == "" {

// such as supported string encodings, support for sub-second timestamps,
// or support for sparse files.
//
// The Writer currently provides no support for sparse files.
type Format int

// Constants to identify various tar formats.
const (
	// Deliberately hide the meaning of constants from public API.
	_ Format = (1 << iota) / 4 // Sequence of 0, 0, 1, 2, 4, 8, etc...

	// FormatUnknown indicates that the format is unknown.

// This assumes that 0.0 headers have already been converted to 0.1 headers
// by the PAX header parsing logic.
func (tr *Reader) readGNUSparsePAXHeaders(hdr *Header) (sparseDatas, error) {
	// Identify the version of GNU headers.
	var is1x0 bool
	major, minor := hdr.PAXRecords[paxGNUSparseMajor], hdr.PAXRecords[paxGNUSparseMinor]
	switch {
	case major == "0" && (minor == "0" || minor == "1"):
		is1x0 = false