Files
audita/internal/core/redaction/redaction.go

124 lines
3.2 KiB
Go

// Package redaction sanitizes diagnostic representations without modifying domain data.
package redaction
import (
"bytes"
"encoding/json"
"fmt"
"sort"
"strings"
"gopkg.in/yaml.v3"
)
// Redactor replaces configured values in a single pass, longest matches first.
// Its zero value is usable and leaves strings unchanged.
type Redactor struct{ replacer *strings.Replacer }
func New(secrets []string) Redactor {
values := map[string]bool{}
for _, secret := range secrets {
if secret == "" {
continue
}
values[secret] = true
// Provider content may itself be serialized JSON inside a string field.
encoded, _ := json.Marshal(secret)
values[string(encoded[1:len(encoded)-1])] = true
}
keys := make([]string, 0, len(values))
for value := range values {
keys = append(keys, value)
}
sort.Slice(keys, func(i, j int) bool {
if len(keys[i]) == len(keys[j]) {
return keys[i] < keys[j]
}
return len(keys[i]) > len(keys[j])
})
pairs := make([]string, 0, 2*len(keys))
for _, key := range keys {
pairs = append(pairs, key, "[REDACTED]")
}
return Redactor{replacer: strings.NewReplacer(pairs...)}
}
func (r Redactor) String(value string) string {
if r.replacer == nil {
return value
}
return r.replacer.Replace(value)
}
// JSON serializes first, then redacts only decoded string values and keys.
// Numbers retain their exact representation. Failure never returns original data.
func (r Redactor) JSON(value any) ([]byte, error) {
raw, err := json.Marshal(value)
if err != nil {
return nil, fmt.Errorf("serialize diagnostic: %s", r.String(err.Error()))
}
decoder := json.NewDecoder(bytes.NewReader(raw))
decoder.UseNumber()
var decoded any
if err := decoder.Decode(&decoded); err != nil {
return nil, fmt.Errorf("decode diagnostic: %s", r.String(err.Error()))
}
safe, err := r.walk(decoded)
if err != nil {
return nil, err
}
return json.MarshalIndent(safe, "", " ")
}
func (r Redactor) walk(value any) (any, error) {
switch v := value.(type) {
case string:
return r.String(v), nil
case []any:
for i := range v {
var err error
v[i], err = r.walk(v[i])
if err != nil {
return nil, err
}
}
case map[string]any:
safe := make(map[string]any, len(v))
for key, item := range v {
safeKey := r.String(key)
if _, exists := safe[safeKey]; exists {
return nil, fmt.Errorf("redaction would produce duplicate object keys")
}
safeItem, err := r.walk(item)
if err != nil {
return nil, err
}
safe[safeKey] = safeItem
}
return safe, nil
}
return value, nil
}
// YAML preserves scalar types while sanitizing string nodes, including map keys.
func (r Redactor) YAML(raw []byte) ([]byte, error) {
var node yaml.Node
if err := yaml.Unmarshal(raw, &node); err != nil {
return nil, fmt.Errorf("decode diagnostic YAML: %s", r.String(err.Error()))
}
var visit func(*yaml.Node)
visit = func(n *yaml.Node) {
if n.Kind == yaml.ScalarNode && n.Tag == "!!str" {
n.Value = r.String(n.Value)
}
n.HeadComment = r.String(n.HeadComment)
n.LineComment = r.String(n.LineComment)
n.FootComment = r.String(n.FootComment)
for _, child := range n.Content {
visit(child)
}
}
visit(&node)
return yaml.Marshal(&node)
}