routewatch/internal/server/handlers_test.go

package server

import (
	"context"
	"encoding/json"
	"net/http"
	"net/http/httptest"
	"runtime"
	"slices"
	"testing"
	"time"

	"git.eeqj.de/sneak/routewatch/internal/config"
	"git.eeqj.de/sneak/routewatch/internal/database"
	"git.eeqj.de/sneak/routewatch/internal/logger"
	"git.eeqj.de/sneak/routewatch/internal/metrics"
	"git.eeqj.de/sneak/routewatch/internal/streamer"
	"github.com/google/uuid"
)

// blockingStatsDB embeds database.Store (left nil) and overrides only
// GetStatsContext, which blocks until release is closed. The stats handlers
// call it in a goroutine; every other Store method is unused on the timeout
// path and would panic if called.
type blockingStatsDB struct {
	database.Store
	release chan struct{}
}

func (d blockingStatsDB) GetStatsContext(_ context.Context) (database.Stats, error) {
	<-d.release

	return database.Stats{}, nil
}

// TestStatsHandlersDoNotLeakOnTimeout drives each stats handler repeatedly with
// a request whose context times out before the database responds, then releases
// the blocked queries and asserts the goroutine count returns to its starting
// value. Before the fix the per-request goroutine sent on an unbuffered channel
// that nothing received once the timeout won, so it blocked forever and every
// poll leaked one goroutine.
func TestStatsHandlersDoNotLeakOnTimeout(t *testing.T) {
	release := make(chan struct{})
	db := blockingStatsDB{release: release}
	s := New(db, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})

	handlers := map[string]http.HandlerFunc{
		"status.json": s.handleStatusJSON(),
		"stats":       s.handleStats(),
	}

	baseline := settledGoroutineCount()

	const (
		iterations     = 20
		requestTimeout = 50 * time.Millisecond
	)
	for _, handler := range handlers {
		for range iterations {
			ctx, cancel := context.WithTimeout(context.Background(), requestTimeout)
			req := httptest.NewRequest(http.MethodGet, "/", nil).WithContext(ctx)
			handler(httptest.NewRecorder(), req)
			cancel()
		}
	}

	// Let the blocked queries finish; with buffered channels each goroutine's
	// send now succeeds and the goroutine exits.
	close(release)

	if !waitForGoroutines(baseline) {
		t.Fatalf("goroutines did not return to baseline %d, got %d",
			baseline, runtime.NumGoroutine())
	}
}

// TestStatsHandlersAnswerFromMemory checks that both stats handlers answer 200
// with the live route counts and the prefix distribution while the database is
// closed, so that any query would fail: the request path reads them from
// memory. The prefix distribution query it used to run read every live route
// and, on a large database, took the whole 4-second deadline, so
// /api/v1/stats answered 500 (https://git.eeqj.de/sneak/routewatch/issues/30).
// The oldest and newest route times still come from one-row lookups at the ends
// of an index; with the database closed they are left out of the answer.
func TestStatsHandlersAnswerFromMemory(t *testing.T) {
	db, err := database.New(&config.Config{StateDir: t.TempDir()}, logger.New())
	if err != nil {
		t.Fatalf("database.New: %v", err)
	}

	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
	if err := db.UpsertLiveRouteBatch([]*database.LiveRoute{
		{
			ID: uuid.New(), Prefix: "198.51.100.0/24", MaskLength: 24, IPVersion: 4,
			OriginASN: 64500, PeerIP: "192.0.2.1", ASPath: []int{64500}, NextHop: "192.0.2.1",
			LastUpdated: ts,
		},
		{
			ID: uuid.New(), Prefix: "2001:db8::/32", MaskLength: 32, IPVersion: 6,
			OriginASN: 64501, PeerIP: "2001:db8::1", ASPath: []int{64501}, NextHop: "2001:db8::1",
			LastUpdated: ts,
		},
	}); err != nil {
		t.Fatalf("UpsertLiveRouteBatch: %v", err)
	}
	if err := db.Close(); err != nil {
		t.Fatalf("Close: %v", err)
	}

	s := New(db, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})
	handlers := map[string]http.HandlerFunc{
		"status.json": s.handleStatusJSON(),
		"stats":       s.handleStats(),
	}

	for name, handler := range handlers {
		rec := httptest.NewRecorder()
		handler(rec, httptest.NewRequest(http.MethodGet, "/", nil))
		if rec.Code != http.StatusOK {
			t.Errorf("%s: status %d, want %d; body %s", name, rec.Code, http.StatusOK, rec.Body)

			continue
		}

		var body struct {
			Data struct {
				IPv4Routes             int                           `json:"ipv4_routes"`
				IPv6Routes             int                           `json:"ipv6_routes"`
				IPv4PrefixDistribution []database.PrefixDistribution `json:"ipv4_prefix_distribution"`
				IPv6PrefixDistribution []database.PrefixDistribution `json:"ipv6_prefix_distribution"`
			} `json:"data"`
		}
		if err := json.Unmarshal(rec.Body.Bytes(), &body); err != nil {
			t.Fatalf("%s: decoding the answer: %v", name, err)
		}

		if body.Data.IPv4Routes != 1 || body.Data.IPv6Routes != 1 {
			t.Errorf("%s: routes = (v4 %d, v6 %d), want (1, 1)", name, body.Data.IPv4Routes, body.Data.IPv6Routes)
		}
		wantV4 := []database.PrefixDistribution{{MaskLength: 24, Count: 1}}
		if !slices.Equal(body.Data.IPv4PrefixDistribution, wantV4) {
			t.Errorf("%s: IPv4 distribution = %v, want %v", name, body.Data.IPv4PrefixDistribution, wantV4)
		}
		wantV6 := []database.PrefixDistribution{{MaskLength: 32, Count: 1}}
		if !slices.Equal(body.Data.IPv6PrefixDistribution, wantV6) {
			t.Errorf("%s: IPv6 distribution = %v, want %v", name, body.Data.IPv6PrefixDistribution, wantV6)
		}
	}
}

// TestPrefixDetailHandlersParsePrefix checks that both prefix pages find a
// stored route whatever valid text form the URL writes its prefix in, and
// answer 400 for a prefix that does not parse. The tables store the form
// net/netip prints. Before the fix the URL text was looked up as typed, so
// "2001:DB8:0:0::/32" was not found although "2001:db8::/32" was stored, and
// "foo/24" answered 500 (https://git.eeqj.de/sneak/routewatch/issues/71).
func TestPrefixDetailHandlersParsePrefix(t *testing.T) {
	db, err := database.New(&config.Config{StateDir: t.TempDir()}, logger.New())
	if err != nil {
		t.Fatalf("database.New: %v", err)
	}
	t.Cleanup(func() { _ = db.Close() })

	ts := time.Date(2026, 1, 2, 3, 4, 5, 0, time.UTC)
	if err := db.UpsertLiveRouteBatch([]*database.LiveRoute{
		{
			ID: uuid.New(), Prefix: "192.0.2.0/24", MaskLength: 24, IPVersion: 4,
			OriginASN: 64500, PeerIP: "198.51.100.1", ASPath: []int{64500}, NextHop: "198.51.100.1",
			LastUpdated: ts,
		},
		{
			ID: uuid.New(), Prefix: "2001:db8::/32", MaskLength: 32, IPVersion: 6,
			OriginASN: 64501, PeerIP: "2001:db8::1", ASPath: []int{64501}, NextHop: "2001:db8::1",
			LastUpdated: ts,
		},
	}); err != nil {
		t.Fatalf("UpsertLiveRouteBatch: %v", err)
	}

	s := New(db, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})

	tests := []struct {
		path string
		want int
	}{
		{"/api/v1/prefix/foo/24", http.StatusBadRequest},
		{"/prefix/foo/24", http.StatusBadRequest},
		{"/api/v1/prefix/2001:DB8:0:0::/32", http.StatusOK},
		{"/prefix/2001:DB8:0:0::/32", http.StatusOK},
		{"/api/v1/prefix/192.0.2.1/24", http.StatusOK},
		{"/prefix/192.0.2.1/24", http.StatusOK},
	}

	for _, tt := range tests {
		rec := httptest.NewRecorder()
		s.router.ServeHTTP(rec, httptest.NewRequest(http.MethodGet, tt.path, nil))
		if rec.Code != tt.want {
			t.Errorf("%s: status %d, want %d; body %s", tt.path, rec.Code, tt.want, rec.Body)
		}
	}
}

// shortRequestDeadline is the deadline the tests below give a request. The
// server's 30-second limit and the stats handlers' 4-second limit inherit it,
// so the tests do not wait for either.
const shortRequestDeadline = 50 * time.Millisecond

// TestRequestPastTimeLimitAnswers504 sends a request through the server's
// router to a page and to a JSON route whose handlers wait for the time limit
// to pass and return without answering. The page answers 504; the JSON route
// answers 504 with the JSON error body. Before the fix both answered 408,
// written by the time limit middleware from its own goroutine while the
// handler could still be writing.
func TestRequestPastTimeLimitAnswers504(t *testing.T) {
	s := New(nil, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})

	waitForTimeLimit := func(_ http.ResponseWriter, r *http.Request) {
		<-r.Context().Done()
	}
	// Registered the way setupRoutes registers pages and JSON routes.
	s.router.Get("/slow-page", waitForTimeLimit)
	s.router.Get("/slow.json", JSONValidationMiddleware(http.HandlerFunc(waitForTimeLimit)).ServeHTTP)

	tests := []struct {
		path         string
		wantJSONBody bool
	}{
		{path: "/slow-page"},
		{path: "/slow.json", wantJSONBody: true},
	}

	for _, tt := range tests {
		ctx, cancel := context.WithTimeout(context.Background(), shortRequestDeadline)
		rec := httptest.NewRecorder()
		s.router.ServeHTTP(rec, httptest.NewRequest(http.MethodGet, tt.path, nil).WithContext(ctx))
		cancel()

		if rec.Code != http.StatusGatewayTimeout {
			t.Errorf("%s: status %d, want %d; body %s", tt.path, rec.Code, http.StatusGatewayTimeout, rec.Body)
		}
		if tt.wantJSONBody {
			checkJSONErrorCode(t, tt.path, rec, http.StatusGatewayTimeout)
		}
	}
}

// TestStatsHandlersAnswer504WhenDatabaseReadRunsOut calls each stats handler
// with a store that does not answer before the handler's 4-second limit runs
// out, and checks that the handler itself answers 504 with the JSON error
// body. Before the fix status.json answered 408, and /api/v1/stats wrote
// nothing, which the JSON check on its route turned into 500.
func TestStatsHandlersAnswer504WhenDatabaseReadRunsOut(t *testing.T) {
	release := make(chan struct{})
	defer close(release)
	db := blockingStatsDB{release: release}
	s := New(db, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})

	handlers := map[string]http.HandlerFunc{
		"status.json": s.handleStatusJSON(),
		"stats":       s.handleStats(),
	}

	for name, handler := range handlers {
		ctx, cancel := context.WithTimeout(context.Background(), shortRequestDeadline)
		rec := httptest.NewRecorder()
		handler(rec, httptest.NewRequest(http.MethodGet, "/", nil).WithContext(ctx))
		cancel()

		if rec.Code != http.StatusGatewayTimeout {
			t.Errorf("%s: status %d, want %d; body %s", name, rec.Code, http.StatusGatewayTimeout, rec.Body)
		}
		checkJSONErrorCode(t, name, rec, http.StatusGatewayTimeout)
	}
}

// slowDB embeds database.Store (left nil) and overrides the reads the routes
// in TestRoutesAnswer504WhenDatabaseReadIsCutOff make. Each read waits until
// its context ends and returns the context's error, as a slow database read
// cut off by a time limit does.
type slowDB struct {
	database.Store
}

func (slowDB) GetStatsContext(ctx context.Context) (database.Stats, error) {
	<-ctx.Done()

	return database.Stats{}, ctx.Err()
}

func (slowDB) GetIPInfoContext(ctx context.Context, _ string) (*database.IPInfo, error) {
	<-ctx.Done()

	return nil, ctx.Err()
}

func (slowDB) GetASDetailsContext(ctx context.Context, _ int) (*database.ASN, []database.LiveRoute, error) {
	<-ctx.Done()

	return nil, nil, ctx.Err()
}

func (slowDB) GetPrefixDetailsContext(ctx context.Context, _ string) ([]database.LiveRoute, error) {
	<-ctx.Done()

	return nil, ctx.Err()
}

func (slowDB) GetRandomPrefixesByLengthContext(ctx context.Context, _, _, _ int) ([]database.LiveRoute, error) {
	<-ctx.Done()

	return nil, ctx.Err()
}

// TestRoutesAnswer504WhenDatabaseReadIsCutOff sends a request through the
// server's router for each route that reads the database, with a store whose
// reads fail when the request's time limit passes. Each route answers 504, the
// JSON routes with the JSON error body. Before the fix they answered 408 or,
// when the handler wrote first, its own error: 500, or 404 on the IP lookups.
// The health check is left out: it answers 503 when its own 2-second ping runs
// out.
func TestRoutesAnswer504WhenDatabaseReadIsCutOff(t *testing.T) {
	s := New(slowDB{}, streamer.New(logger.New(), metrics.New()), logger.New(), &config.Config{})

	tests := []struct {
		path         string
		wantJSONBody bool
	}{
		{path: "/as/64500"},
		{path: "/prefix/192.0.2.0/24"},
		{path: "/prefixlength/24"},
		{path: "/prefixlength6/48"},
		{path: "/status.json", wantJSONBody: true},
		{path: "/ip/192.0.2.1", wantJSONBody: true},
		{path: "/api/v1/stats", wantJSONBody: true},
		{path: "/api/v1/ip/192.0.2.1", wantJSONBody: true},
		{path: "/api/v1/as/64500", wantJSONBody: true},
		{path: "/api/v1/prefix/192.0.2.0/24", wantJSONBody: true},
	}

	for _, tt := range tests {
		ctx, cancel := context.WithTimeout(context.Background(), shortRequestDeadline)
		rec := httptest.NewRecorder()
		s.router.ServeHTTP(rec, httptest.NewRequest(http.MethodGet, tt.path, nil).WithContext(ctx))
		cancel()

		if rec.Code != http.StatusGatewayTimeout {
			t.Errorf("%s: status %d, want %d; body %s", tt.path, rec.Code, http.StatusGatewayTimeout, rec.Body)
		}
		if tt.wantJSONBody {
			checkJSONErrorCode(t, tt.path, rec, http.StatusGatewayTimeout)
		}
	}
}

// checkJSONErrorCode checks that rec holds the JSON error body with the given
// code.
func checkJSONErrorCode(t *testing.T, name string, rec *httptest.ResponseRecorder, want int) {
	t.Helper()

	var body struct {
		Status string `json:"status"`
		Error  struct {
			Code int `json:"code"`
		} `json:"error"`
	}
	if err := json.Unmarshal(rec.Body.Bytes(), &body); err != nil {
		t.Errorf("%s: decoding the answer %q: %v", name, rec.Body, err)

		return
	}
	if body.Status != "error" || body.Error.Code != want {
		t.Errorf("%s: answer %s, want status \"error\" and error code %d", name, rec.Body, want)
	}
}

// settledGoroutineCount lets transient goroutines finish, then reports the
// current count.
func settledGoroutineCount() int {
	prev := runtime.NumGoroutine()
	for range 20 {
		time.Sleep(10 * time.Millisecond)
		cur := runtime.NumGoroutine()
		if cur == prev {
			return cur
		}
		prev = cur
	}

	return prev
}

// waitForGoroutines waits until the goroutine count drops to target or below.
func waitForGoroutines(target int) bool {
	for range 100 {
		if runtime.NumGoroutine() <= target {
			return true
		}
		time.Sleep(10 * time.Millisecond)
	}

	return false
}