butterrobot/internal/observability/scrub_test.go
Felipe M. a2e1196953
Some checks failed
CI / goreleaser-lint (push) Successful in 6s
CI / format (push) Successful in 57s
CI / test (push) Successful in 2m59s
CI / lint (push) Successful in 4m9s
CI / build (push) Successful in 7m17s
Release / release (push) Failing after 7m6s
feat: add Sentry observability support
Report errors, panics and optional performance traces to Sentry. Sentry
stays disabled unless SENTRY_DSN is set.

- slog ERROR records become Sentry issues, with the error attribute
  promoted to an exception so issues group by root cause
- queue worker, reminder worker, cache cleanup and HTTP handler panics
  are captured with a stack trace
- events carry platform/plugin/component tags for filtering
- HTTP requests are traced when SENTRY_TRACES_SAMPLE_RATE is above 0;
  /healthz is never traced

Configured credentials are redacted from every outgoing payload. This is
required rather than defensive: the Telegram webhook embeds the bot token
in its URL path, and failed Telegram API calls quote that URL in their
error text, so events would otherwise carry the token in the clear. A new
platform must register its credential in Config.Secrets().

Also moves the module to Go 1.27, refreshes every dependency and pins
golangci-lint v2.13.2. sentry-go 0.48 removed issue creation from its slog
integration, so the ERROR-to-issue conversion lives in
internal/observability/handler.go instead of relying on the SDK; leaving
it to the SDK would have silently downgraded issues to log lines.

Claude-Session: https://claude.ai/code/session_01W7tcpMTEyk9RrHvT7Be5zZ
2026-09-21 14:10:15 +02:00

169 lines
4.6 KiB
Go

package observability
import (
"context"
"io"
"log/slog"
"net/http"
"net/http/httptest"
"strings"
"sync"
"testing"
"time"
"github.com/getsentry/sentry-go"
"git.nakama.town/fmartingr/butterrobot/internal/config"
)
// testToken mimics the shape of a Telegram bot token, which the webhook URL
// embeds in its path.
const testToken = "123456:AAFakeTelegramBotTokenValue"
// captureTransport records events locally instead of shipping them to Sentry.
type captureTransport struct {
mu sync.Mutex
events []*sentry.Event
}
func (t *captureTransport) Configure(sentry.ClientOptions) {}
func (t *captureTransport) SendEvent(event *sentry.Event) {
t.mu.Lock()
defer t.mu.Unlock()
t.events = append(t.events, event)
}
func (t *captureTransport) Flush(time.Duration) bool { return true }
func (t *captureTransport) FlushWithContext(context.Context) bool { return true }
func (t *captureTransport) Close() {}
func (t *captureTransport) captured() []*sentry.Event {
t.mu.Lock()
defer t.mu.Unlock()
return append([]*sentry.Event(nil), t.events...)
}
// enableSentry binds a client that captures events in memory to the current
// hub, and unbinds it when the test ends.
func enableSentry(t *testing.T, secrets []string) *captureTransport {
t.Helper()
transport := &captureTransport{}
options := clientOptions(config.SentryConfig{
DSN: "https://key@example.invalid/1",
Environment: "test",
TracesSampleRate: 1.0,
}, false, secrets)
options.Transport = transport
client, err := sentry.NewClient(options)
if err != nil {
t.Fatalf("failed to create client: %v", err)
}
hub := sentry.CurrentHub()
hub.BindClient(client)
t.Cleanup(func() { hub.BindClient(nil) })
return transport
}
func TestWebhookTokenIsScrubbedFromRequestContext(t *testing.T) {
transport := enableSentry(t, []string{testToken})
handler := HTTPMiddleware(http.HandlerFunc(func(_ http.ResponseWriter, r *http.Request) {
sentry.GetHubFromContext(r.Context()).CaptureMessage("handler failure")
}))
handler.ServeHTTP(httptest.NewRecorder(),
httptest.NewRequest(http.MethodPost, "/telegram/incoming/"+testToken, nil))
events := transport.captured()
if len(events) < 2 {
t.Fatalf("expected an error event and a transaction, got %d", len(events))
}
for _, event := range events {
if event.Request != nil && strings.Contains(event.Request.URL, testToken) {
t.Errorf("request URL leaks the bot token: %s", event.Request.URL)
}
if strings.Contains(event.Transaction, testToken) {
t.Errorf("transaction name leaks the bot token: %s", event.Transaction)
}
}
}
func TestErrorLogIsForwardedTaggedAndScrubbed(t *testing.T) {
transport := enableSentry(t, []string{testToken})
logger := WrapLogger(slog.New(slog.NewTextHandler(io.Discard, nil)), config.SentryConfig{})
ctx := WithTags(context.Background(), map[string]string{"platform": "telegram"})
logger.ErrorContext(ctx, "Error sending message",
"error", `Post "https://api.telegram.org/bot`+testToken+`/sendMessage": timeout`)
events := transport.captured()
if len(events) != 1 {
t.Fatalf("expected exactly one event, got %d", len(events))
}
event := events[0]
if event.Tags["platform"] != "telegram" {
t.Errorf("expected the platform tag to reach Sentry, got %v", event.Tags)
}
for _, exception := range event.Exception {
if strings.Contains(exception.Value, testToken) {
t.Errorf("exception value leaks the bot token: %s", exception.Value)
}
}
for name, context := range event.Contexts {
for key, value := range context {
if text, ok := value.(string); ok && strings.Contains(text, testToken) {
t.Errorf("context %s.%s leaks the bot token: %s", name, key, text)
}
}
}
}
func TestNewScrubberIgnoresShortSecrets(t *testing.T) {
if newScrubber([]string{"", "1234"}) != nil {
t.Fatal("expected short secrets to be ignored")
}
if newScrubber([]string{testToken}) == nil {
t.Fatal("expected a scrubber for a usable secret")
}
}
func TestScrubberRedactsNestedValues(t *testing.T) {
s := newScrubber([]string{testToken})
event := sentry.NewEvent()
event.Message = "failed calling " + testToken
event.Contexts[logContextKey] = sentry.Context{
"nested": map[string]any{"url": "https://api.telegram.org/bot" + testToken},
"count": 3,
}
scrubbed := s.event(event)
if strings.Contains(scrubbed.Message, testToken) {
t.Errorf("message leaks the token: %s", scrubbed.Message)
}
data := scrubbed.Contexts[logContextKey]
nested := data["nested"].(map[string]any)
if strings.Contains(nested["url"].(string), testToken) {
t.Errorf("nested value leaks the token: %v", nested["url"])
}
if data["count"] != 3 {
t.Errorf("expected non-string values to be untouched, got %v", data["count"])
}
}