hush/vendor/google.golang.org/protobuf/internal/descfmt/stringer.go
jx12n 4d9a26498e hush: one-time secret links the server cannot read
Paste a secret, get a link, send it. The first person to open it and press
Reveal sees the secret; the link dies at that moment. The recipient needs a
browser and nothing else — no account, no client, no installed tooling.

The server cannot read what it stores. AES-256-GCM happens in the browser and
the key lives in the URL fragment, which browsers never transmit, so hushd
holds ciphertext and no key material. That is a property of where the key sits
rather than a promise about our conduct, which is why there is deliberately no
endpoint accepting a plaintext secret and no server-side-encryption fallback:
two guarantees behind one URL would be worse than one honest guarantee.

Three decisions carry the design:

  * GET /s/{id} touches NO storage, not even to check existence. Slack, Teams,
    WhatsApp, iMessage and Outlook Safe Links all fetch a URL before a human
    sees it, so destroying on GET would destroy most secrets in transit and the
    recipient's "already used" would be indistinguishable from interception.
    Only POST /reveal consumes. Bot user-agent detection is an arms race;
    removing the side effect from GET is not. Pinned by
    TestGettingTheRevealPageNeverConsumesTheSecret.
  * Destruction is one Redis GETDEL, which is atomic. GET-then-DEL has a window
    where two simultaneous readers both win, and for a one-time secret that
    window is the product. The store contract demands atomicity and the same
    concurrency test runs against both implementations.
  * Missing, already-revealed, expired and evicted are ONE indistinguishable
    410. Separating them would confirm to a prober that a given link was real.

The secret id IS the capability, so secret.ID is a struct whose every
accidental path — %v, %s, String(), slog, json.Marshal — emits a redacted
handle or refuses, and the raw value needs an explicit Value(). The first
version tried to prevent leaks by implementing no String() at all; its own test
caught that Go's fmt prints unexported fields anyway, so forbidding the method
had removed the control rather than the leak.

Operationally: structured JSON on stdout in the fleet's wire format, which
Vector already collects with no annotation; six hush_* metrics on the chassis
registry with no id, IP or path in any label; five alert rules wired into
vmalert. The public Ingress enumerates /, /s/ and /api/ so /metrics, /healthz
and /readyz share the port but are unreachable from the internet — no
basic-auth middleware to maintain and get wrong.

Dependencies are vendored because go-chassis is private: the Woodpecker test
step and the in-cluster Kaniko build both run -mod=vendor with GOPROXY=off and
hold no git credential.

cmd/hush-mcp is a stdio MCP server doing the same client-side crypto locally,
so using hush from an agent preserves the same guarantee as using it from a
browser.
2026-09-03 00:08:38 -06:00

415 lines
12 KiB
Go
Raw Permalink Blame History

This file contains invisible Unicode characters

This file contains invisible Unicode characters that are indistinguishable to humans but may be processed differently by a computer. If you think that this is intentional, you can safely ignore this warning. Use the Escape button to reveal them.

// Copyright 2018 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
// Package descfmt provides functionality to format descriptors.
package descfmt
import (
"fmt"
"io"
"reflect"
"strconv"
"strings"
"google.golang.org/protobuf/internal/detrand"
"google.golang.org/protobuf/internal/pragma"
"google.golang.org/protobuf/reflect/protoreflect"
)
type list interface {
Len() int
pragma.DoNotImplement
}
func FormatList(s fmt.State, r rune, vs list) {
io.WriteString(s, formatListOpt(vs, true, r == 'v' && (s.Flag('+') || s.Flag('#'))))
}
func formatListOpt(vs list, isRoot, allowMulti bool) string {
start, end := "[", "]"
if isRoot {
var name string
switch vs.(type) {
case protoreflect.Names:
name = "Names"
case protoreflect.FieldNumbers:
name = "FieldNumbers"
case protoreflect.FieldRanges:
name = "FieldRanges"
case protoreflect.EnumRanges:
name = "EnumRanges"
case protoreflect.FileImports:
name = "FileImports"
case protoreflect.Descriptor:
name = reflect.ValueOf(vs).MethodByName("Get").Type().Out(0).Name() + "s"
default:
name = reflect.ValueOf(vs).Elem().Type().Name()
}
start, end = name+"{", "}"
}
var ss []string
switch vs := vs.(type) {
case protoreflect.Names:
for i := 0; i < vs.Len(); i++ {
ss = append(ss, fmt.Sprint(vs.Get(i)))
}
return start + joinStrings(ss, false) + end
case protoreflect.FieldNumbers:
for i := 0; i < vs.Len(); i++ {
ss = append(ss, fmt.Sprint(vs.Get(i)))
}
return start + joinStrings(ss, false) + end
case protoreflect.FieldRanges:
for i := 0; i < vs.Len(); i++ {
r := vs.Get(i)
if r[0]+1 == r[1] {
ss = append(ss, fmt.Sprintf("%d", r[0]))
} else {
ss = append(ss, fmt.Sprintf("%d:%d", r[0], r[1])) // enum ranges are end exclusive
}
}
return start + joinStrings(ss, false) + end
case protoreflect.EnumRanges:
for i := 0; i < vs.Len(); i++ {
r := vs.Get(i)
if r[0] == r[1] {
ss = append(ss, fmt.Sprintf("%d", r[0]))
} else {
ss = append(ss, fmt.Sprintf("%d:%d", r[0], int64(r[1])+1)) // enum ranges are end inclusive
}
}
return start + joinStrings(ss, false) + end
case protoreflect.FileImports:
for i := 0; i < vs.Len(); i++ {
var rs records
rv := reflect.ValueOf(vs.Get(i))
rs.Append(rv, []methodAndName{
{rv.MethodByName("Path"), "Path"},
{rv.MethodByName("Package"), "Package"},
{rv.MethodByName("IsPublic"), "IsPublic"},
{rv.MethodByName("IsWeak"), "IsWeak"},
}...)
ss = append(ss, "{"+rs.Join()+"}")
}
return start + joinStrings(ss, allowMulti) + end
default:
_, isEnumValue := vs.(protoreflect.EnumValueDescriptors)
for i := 0; i < vs.Len(); i++ {
m := reflect.ValueOf(vs).MethodByName("Get")
v := m.Call([]reflect.Value{reflect.ValueOf(i)})[0].Interface()
ss = append(ss, formatDescOpt(v.(protoreflect.Descriptor), false, allowMulti && !isEnumValue, nil))
}
return start + joinStrings(ss, allowMulti && isEnumValue) + end
}
}
type methodAndName struct {
method reflect.Value
name string
}
func FormatDesc(s fmt.State, r rune, t protoreflect.Descriptor) {
io.WriteString(s, formatDescOpt(t, true, r == 'v' && (s.Flag('+') || s.Flag('#')), nil))
}
func InternalFormatDescOptForTesting(t protoreflect.Descriptor, isRoot, allowMulti bool, record func(string)) string {
return formatDescOpt(t, isRoot, allowMulti, record)
}
func formatDescOpt(t protoreflect.Descriptor, isRoot, allowMulti bool, record func(string)) string {
rv := reflect.ValueOf(t)
rt := rv.MethodByName("ProtoType").Type().In(0)
start, end := "{", "}"
if isRoot {
start = rt.Name() + "{"
}
_, isFile := t.(protoreflect.FileDescriptor)
rs := records{
allowMulti: allowMulti,
record: record,
}
if t.IsPlaceholder() {
if isFile {
rs.Append(rv, []methodAndName{
{rv.MethodByName("Path"), "Path"},
{rv.MethodByName("Package"), "Package"},
{rv.MethodByName("IsPlaceholder"), "IsPlaceholder"},
}...)
} else {
rs.Append(rv, []methodAndName{
{rv.MethodByName("FullName"), "FullName"},
{rv.MethodByName("IsPlaceholder"), "IsPlaceholder"},
}...)
}
} else {
switch {
case isFile:
rs.Append(rv, methodAndName{rv.MethodByName("Syntax"), "Syntax"})
case isRoot:
rs.Append(rv, []methodAndName{
{rv.MethodByName("Syntax"), "Syntax"},
{rv.MethodByName("FullName"), "FullName"},
}...)
default:
rs.Append(rv, methodAndName{rv.MethodByName("Name"), "Name"})
}
switch t := t.(type) {
case protoreflect.FieldDescriptor:
accessors := []methodAndName{
{rv.MethodByName("Number"), "Number"},
{rv.MethodByName("Cardinality"), "Cardinality"},
{rv.MethodByName("Kind"), "Kind"},
{rv.MethodByName("HasJSONName"), "HasJSONName"},
{rv.MethodByName("JSONName"), "JSONName"},
{rv.MethodByName("HasPresence"), "HasPresence"},
{rv.MethodByName("IsExtension"), "IsExtension"},
{rv.MethodByName("IsPacked"), "IsPacked"},
{rv.MethodByName("IsWeak"), "IsWeak"},
{rv.MethodByName("IsList"), "IsList"},
{rv.MethodByName("IsMap"), "IsMap"},
{rv.MethodByName("MapKey"), "MapKey"},
{rv.MethodByName("MapValue"), "MapValue"},
{rv.MethodByName("HasDefault"), "HasDefault"},
{rv.MethodByName("Default"), "Default"},
{rv.MethodByName("ContainingOneof"), "ContainingOneof"},
{rv.MethodByName("ContainingMessage"), "ContainingMessage"},
{rv.MethodByName("Message"), "Message"},
{rv.MethodByName("Enum"), "Enum"},
}
for _, s := range accessors {
switch s.name {
case "MapKey":
if k := t.MapKey(); k != nil {
rs.recs = append(rs.recs, [2]string{"MapKey", k.Kind().String()})
}
case "MapValue":
if v := t.MapValue(); v != nil {
switch v.Kind() {
case protoreflect.EnumKind:
rs.AppendRecs("MapValue", [2]string{"MapValue", string(v.Enum().FullName())})
case protoreflect.MessageKind, protoreflect.GroupKind:
rs.AppendRecs("MapValue", [2]string{"MapValue", string(v.Message().FullName())})
default:
rs.AppendRecs("MapValue", [2]string{"MapValue", v.Kind().String()})
}
}
case "ContainingOneof":
if od := t.ContainingOneof(); od != nil {
rs.AppendRecs("ContainingOneof", [2]string{"Oneof", string(od.Name())})
}
case "ContainingMessage":
if t.IsExtension() {
rs.AppendRecs("ContainingMessage", [2]string{"Extendee", string(t.ContainingMessage().FullName())})
}
case "Message":
if !t.IsMap() {
rs.Append(rv, s)
}
default:
rs.Append(rv, s)
}
}
case protoreflect.OneofDescriptor:
var ss []string
fs := t.Fields()
for i := 0; i < fs.Len(); i++ {
ss = append(ss, string(fs.Get(i).Name()))
}
if len(ss) > 0 {
rs.AppendRecs("Fields", [2]string{"Fields", "[" + joinStrings(ss, false) + "]"})
}
case protoreflect.FileDescriptor:
rs.Append(rv, []methodAndName{
{rv.MethodByName("Path"), "Path"},
{rv.MethodByName("Package"), "Package"},
{rv.MethodByName("Imports"), "Imports"},
{rv.MethodByName("Messages"), "Messages"},
{rv.MethodByName("Enums"), "Enums"},
{rv.MethodByName("Extensions"), "Extensions"},
{rv.MethodByName("Services"), "Services"},
}...)
case protoreflect.MessageDescriptor:
rs.Append(rv, []methodAndName{
{rv.MethodByName("IsMapEntry"), "IsMapEntry"},
{rv.MethodByName("Fields"), "Fields"},
{rv.MethodByName("Oneofs"), "Oneofs"},
{rv.MethodByName("ReservedNames"), "ReservedNames"},
{rv.MethodByName("ReservedRanges"), "ReservedRanges"},
{rv.MethodByName("RequiredNumbers"), "RequiredNumbers"},
{rv.MethodByName("ExtensionRanges"), "ExtensionRanges"},
{rv.MethodByName("Messages"), "Messages"},
{rv.MethodByName("Enums"), "Enums"},
{rv.MethodByName("Extensions"), "Extensions"},
}...)
case protoreflect.EnumDescriptor:
rs.Append(rv, []methodAndName{
{rv.MethodByName("Values"), "Values"},
{rv.MethodByName("ReservedNames"), "ReservedNames"},
{rv.MethodByName("ReservedRanges"), "ReservedRanges"},
{rv.MethodByName("IsClosed"), "IsClosed"},
}...)
case protoreflect.EnumValueDescriptor:
rs.Append(rv, []methodAndName{
{rv.MethodByName("Number"), "Number"},
}...)
case protoreflect.ServiceDescriptor:
rs.Append(rv, []methodAndName{
{rv.MethodByName("Methods"), "Methods"},
}...)
case protoreflect.MethodDescriptor:
rs.Append(rv, []methodAndName{
{rv.MethodByName("Input"), "Input"},
{rv.MethodByName("Output"), "Output"},
{rv.MethodByName("IsStreamingClient"), "IsStreamingClient"},
{rv.MethodByName("IsStreamingServer"), "IsStreamingServer"},
}...)
}
if m := rv.MethodByName("GoType"); m.IsValid() {
rs.Append(rv, methodAndName{m, "GoType"})
}
}
return start + rs.Join() + end
}
type records struct {
recs [][2]string
allowMulti bool
// record is a function that will be called for every Append() or
// AppendRecs() call, to be used for testing with the
// InternalFormatDescOptForTesting function.
record func(string)
}
func (rs *records) AppendRecs(fieldName string, newRecs [2]string) {
if rs.record != nil {
rs.record(fieldName)
}
rs.recs = append(rs.recs, newRecs)
}
func (rs *records) Append(v reflect.Value, accessors ...methodAndName) {
for _, a := range accessors {
if rs.record != nil {
rs.record(a.name)
}
var rv reflect.Value
if a.method.IsValid() {
rv = a.method.Call(nil)[0]
}
if v.Kind() == reflect.Struct && !rv.IsValid() {
rv = v.FieldByName(a.name)
}
if !rv.IsValid() {
panic(fmt.Sprintf("unknown accessor: %v.%s", v.Type(), a.name))
}
if _, ok := rv.Interface().(protoreflect.Value); ok {
rv = rv.MethodByName("Interface").Call(nil)[0]
if !rv.IsNil() {
rv = rv.Elem()
}
}
// Ignore zero values.
var isZero bool
switch rv.Kind() {
case reflect.Interface, reflect.Slice:
isZero = rv.IsNil()
case reflect.Bool:
isZero = rv.Bool() == false
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
isZero = rv.Int() == 0
case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
isZero = rv.Uint() == 0
case reflect.String:
isZero = rv.String() == ""
}
if n, ok := rv.Interface().(list); ok {
isZero = n.Len() == 0
}
if isZero {
continue
}
// Format the value.
var s string
v := rv.Interface()
switch v := v.(type) {
case list:
s = formatListOpt(v, false, rs.allowMulti)
case protoreflect.FieldDescriptor, protoreflect.OneofDescriptor, protoreflect.EnumValueDescriptor, protoreflect.MethodDescriptor:
s = string(v.(protoreflect.Descriptor).Name())
case protoreflect.Descriptor:
s = string(v.FullName())
case string:
s = strconv.Quote(v)
case []byte:
s = fmt.Sprintf("%q", v)
default:
s = fmt.Sprint(v)
}
rs.recs = append(rs.recs, [2]string{a.name, s})
}
}
func (rs *records) Join() string {
var ss []string
// In single line mode, simply join all records with commas.
if !rs.allowMulti {
for _, r := range rs.recs {
ss = append(ss, r[0]+formatColon(0)+r[1])
}
return joinStrings(ss, false)
}
// In allowMulti line mode, align single line records for more readable output.
var maxLen int
flush := func(i int) {
for _, r := range rs.recs[len(ss):i] {
ss = append(ss, r[0]+formatColon(maxLen-len(r[0]))+r[1])
}
maxLen = 0
}
for i, r := range rs.recs {
if isMulti := strings.Contains(r[1], "\n"); isMulti {
flush(i)
ss = append(ss, r[0]+formatColon(0)+strings.Join(strings.Split(r[1], "\n"), "\n\t"))
} else if maxLen < len(r[0]) {
maxLen = len(r[0])
}
}
flush(len(rs.recs))
return joinStrings(ss, true)
}
func formatColon(padding int) string {
// Deliberately introduce instability into the debug output to
// discourage users from performing string comparisons.
// This provides us flexibility to change the output in the future.
if detrand.Bool() {
return ":" + strings.Repeat(" ", 1+padding) // use non-breaking spaces (U+00a0)
} else {
return ":" + strings.Repeat(" ", 1+padding) // use regular spaces (U+0020)
}
}
func joinStrings(ss []string, isMulti bool) string {
if len(ss) == 0 {
return ""
}
if isMulti {
return "\n\t" + strings.Join(ss, "\n\t") + "\n"
}
return strings.Join(ss, ", ")
}