package host

import (
	"bytes"
	"context"
	"log"
	"net/http"
	"net/http/httptest"
	"reflect"
	"runtime"
	"strings"
	"sync"
	"testing"
	"time"

	"github.com/carlosframework/platform/internal/activator"
	"github.com/carlosframework/platform/internal/registry"
	"github.com/carlosframework/platform/internal/release"
)

// The stamp is the whole licence for skipping the sweep's grace: it is written
// only after the adopting pass is finished with the host. A drift whose route
// is still stamped with the old version is an adoption in flight, and one with
// no row at all is not a route this box adopts.
func TestConvergeAdoptedCyclesOnlyWhenTheStampAgrees(t *testing.T) {
	rows := map[string]*registry.Route{
		"done.test":     {Host: "done.test", Version: "v2"},
		"inflight.test": {Host: "inflight.test", Version: "v1"},
		"named.test":    {Host: "named.test", Version: "v2", Unit: "console.service"},
		"unitback.test": {Host: "unitback.test", Version: "v2", Backing: registry.BackingUnit},
	}
	drifted := func() map[string]string {
		return map[string]string{
			"done.test": "v2", "inflight.test": "v2", "named.test": "v2",
			"unitback.test": "v2", "norow.test": "v2",
		}
	}
	var cycled []string
	n := convergeAdopted(drifted,
		func(h string) (*registry.Route, error) { return rows[h], nil },
		func(h, ver string) bool { cycled = append(cycled, h+"@"+ver); return true }, 10)
	if want := []string{"done.test@v2"}; !reflect.DeepEqual(cycled, want) {
		t.Fatalf("cycled = %v, want %v", cycled, want)
	}
	if n != 1 {
		t.Fatalf("convergeAdopted = %d, want 1", n)
	}
}

// The ordinary tick, where nothing has shipped, must cost no registry read:
// this leg runs every two seconds on every box.
func TestConvergeAdoptedReadsNoRouteWhenNothingDrifted(t *testing.T) {
	n := convergeAdopted(func() map[string]string { return nil },
		func(h string) (*registry.Route, error) {
			t.Fatalf("read the route for %s with nothing drifted", h)
			return nil, nil
		},
		func(h, ver string) bool {
			t.Fatalf("cycled %s with nothing drifted", h)
			return false
		}, 10)
	if n != 0 {
		t.Fatalf("convergeAdopted = %d, want 0", n)
	}
}

// The leg runs on its own loop, never waiting on the fast tick: a tick that
// takes a full reconcile and adopt pass can run tens of seconds, and a
// stamped adoption must not wait behind it. Here the loop keeps calling the
// leg on its own cadence, and stops when told.
func TestConvergeLoopRunsTheLegOnItsOwnCadence(t *testing.T) {
	t.Parallel()
	calls := make(chan struct{}, 100)
	stop := make(chan struct{})
	done := make(chan struct{})
	go func() {
		runConvergeLoop(func() { calls <- struct{}{} }, 5*time.Millisecond, stop)
		close(done)
	}()
	for i := 0; i < 3; i++ {
		select {
		case <-calls:
		case <-time.After(10 * time.Second):
			t.Fatalf("the loop ran the leg %d times, want it to keep running", i)
		}
	}
	close(stop)
	select {
	case <-done:
	case <-time.After(10 * time.Second):
		t.Fatal("the loop did not stop")
	}
}

// The production sequence behind discussion #231, with the real pass on both
// sides. The agent cannot write the version store, so the ROOT pass adopts a
// new release for a warm exec instance: it stages, repoints and stamps, and —
// having no handle on the agent's children — cycles nothing. The agent's own
// pass then finds the host up to date and never reaches its restart callback.
// Before the converge leg, that left the child on the old build until the
// activator sweep's 90s grace ran out.
func TestRootAdoptionOfAWarmExecInstanceIsFinishedByTheAgent(t *testing.T) {
	env := newFastFixture(t)
	ctx := context.Background()
	plat := runtime.GOOS + "-" + runtime.GOARCH
	host := "warm.test"
	ship := func(ver string) {
		t.Helper()
		m := release.Manifest{
			App: "fastapp", Version: ver, Kind: "binary",
			Artifacts: map[string]release.Artifact{plat: {}},
		}
		if err := release.Ship(ctx, env.store, "bab", m, map[string][]byte{plat: []byte("build " + ver)}); err != nil {
			t.Fatalf("Ship %s: %v", ver, err)
		}
		if err := release.Promote(ctx, env.store, env.priv, "bab", "fastapp", "edge", ver); err != nil {
			t.Fatalf("Promote %s: %v", ver, err)
		}
	}
	// rootRestartRoute with systemctl stubbed out: same dispatch, bounce nil.
	rootPass := func() {
		t.Helper()
		restart := makeRestartRoute(env.reg, func(string) {}, unitOnly(func(string) error { return nil }), nil)
		if err := release.Pass(ctx, env.reg, env.store, env.pub, env.optDir, restart, release.BakeDir(env.dataDir)); err != nil {
			t.Fatalf("root pass: %v", err)
		}
	}
	if err := env.reg.Upsert(registry.Route{
		Host: host, Kind: "instance", App: "fastapp", Channel: "edge",
		Status: "ready", Hibernate: true,
	}); err != nil {
		t.Fatalf("upsert: %v", err)
	}

	ship("v1")
	rootPass()
	spawned := "v1" // the awake child's build: frozen at exec, as the activator records it

	ship("v2")
	rootPass()
	if v, err := release.LiveVersion(env.optDir, host); err != nil || v != "v2" {
		t.Fatalf("live = %q, %v after the root pass; want v2", v, err)
	}
	if rt, err := env.reg.Get(host); err != nil || rt == nil || rt.Version != "v2" {
		t.Fatalf("route = %+v, %v after the root pass; want it stamped v2", rt, err)
	}

	// The agent's tick. Its pass has a bounce to offer and never gets to use
	// it: the host is already up to date.
	var bounced, cycled []string
	restart := makeRestartRoute(env.reg, func(string) {}, unitOnly(func(string) error { return nil }),
		func(h string) bool { bounced = append(bounced, h); return true })
	d := env.deps(nil)
	d.restart = restart
	d.passOpts = fastPassOpts(env.dataDir, func(string, string, error) {}, restart, nil)
	d.fullPass = func(ctx context.Context) {
		if err := release.Pass(ctx, d.reg, d.store, d.pub, d.optDir, d.restart, d.passOpts...); err != nil {
			t.Fatalf("agent pass: %v", err)
		}
	}
	leg := func() {
		convergeAdopted(
			func() map[string]string {
				live, _ := release.LiveVersion(env.optDir, host)
				if live == "" || live == spawned {
					return nil
				}
				return map[string]string{host: live}
			},
			env.reg.Get,
			func(h, ver string) bool { cycled = append(cycled, h+"@"+ver); spawned = ver; return true },
			adoptedCyclesPerTick)
	}
	now := time.Unix(1754640000, 0)
	fastTick(ctx, d, now) // seeds the serial
	if err := release.BumpSerial(ctx, env.store, now); err != nil {
		t.Fatalf("BumpSerial: %v", err)
	}
	fastTick(ctx, d, now.Add(10*time.Second)) // the serial move: the agent's full pass
	leg()                                     // the converge loop's next run

	if len(bounced) != 0 {
		t.Fatalf("the agent's pass bounced %v — it found the host already adopted, so this test no longer models the gap", bounced)
	}
	if want := []string{host + "@v2"}; !reflect.DeepEqual(cycled, want) {
		t.Fatalf("cycled = %v, want %v: the warm child is still on v1 under a symlink and a stamp that both say v2", cycled, want)
	}
	leg()
	if len(cycled) != 1 {
		t.Fatalf("cycled again on the next tick: %v", cycled)
	}
}

// One pass can adopt every warm instance of an app inside a few seconds.
// Stopping them all on one tick would blip every host of the app at the same
// moment; the leg spends its budget in host order and leaves the rest for the
// ticks that follow. A drifted host it has to wait on costs the others nothing.
func TestConvergeAdoptedCyclesOnePerTickInHostOrder(t *testing.T) {
	spawned := map[string]string{"a.test": "v1", "b.test": "v1", "c.test": "v1", "wait.test": "v1"}
	stamp := map[string]string{"a.test": "v2", "b.test": "v2", "c.test": "v2", "wait.test": "v1"}
	drifted := func() map[string]string {
		out := map[string]string{}
		for h, v := range spawned {
			if v != "v2" {
				out[h] = "v2"
			}
		}
		return out
	}
	get := func(h string) (*registry.Route, error) { return &registry.Route{Host: h, Version: stamp[h]}, nil }
	var cycled []string
	converge := func(h, ver string) bool { cycled = append(cycled, h); spawned[h] = ver; return true }

	for tick := 1; tick <= 4; tick++ {
		if n := convergeAdopted(drifted, get, converge, adoptedCyclesPerTick); tick <= 3 && n != 1 {
			t.Fatalf("tick %d cycled %d, want 1", tick, n)
		}
	}
	if want := []string{"a.test", "b.test", "c.test"}; !reflect.DeepEqual(cycled, want) {
		t.Fatalf("cycled = %v, want %v: one per tick, in host order, never the host whose stamp does not agree", cycled, want)
	}
}

// versionedBackend is activator.Backend with the one thing convergence needs
// and fakeWakeBackend does not model: a child that knows which build it was
// exec'd from, read off the SAME live symlink the release pass writes. It
// binds no socket — the caller supplies the proxy — so it runs anywhere.
type versionedBackend struct {
	fakeWakeBackend
	optDir string

	mu     sync.Mutex
	starts []string // the build each spawn exec'd, in order
}

type flagProc struct {
	mu      sync.Mutex
	stopped bool
}

func (p *flagProc) Stop(context.Context) error {
	p.mu.Lock()
	p.stopped = true
	p.mu.Unlock()
	return nil
}

func (p *flagProc) Exited() bool {
	p.mu.Lock()
	defer p.mu.Unlock()
	return p.stopped
}

func (b *versionedBackend) LiveVersion(host string) (string, error) {
	return release.LiveVersion(b.optDir, host)
}

func (b *versionedBackend) Start(_ context.Context, rt registry.Route) (activator.Proc, string, error) {
	ver, err := release.LiveVersion(b.optDir, rt.Host)
	if err != nil {
		return nil, "", err
	}
	b.mu.Lock()
	b.starts = append(b.starts, ver)
	b.mu.Unlock()
	return &flagProc{}, ver, nil
}

func (b *versionedBackend) started() []string {
	b.mu.Lock()
	defer b.mu.Unlock()
	return append([]string(nil), b.starts...)
}

// The whole of discussion #231 with nothing stood in for: the real release
// pass as root runs it, a real activator owning the child, and the leg exactly
// as fastConvergeLoop builds it (convergeLegFor). The sweeper is never
// started, so the ONLY thing that can cycle the instance is the leg — a 90s
// grace cannot rescue this test.
func TestConvergeLegCyclesARootAdoptedInstanceOnARealActivator(t *testing.T) {
	t.Parallel()
	env := newFastFixture(t)
	ctx := context.Background()
	plat := runtime.GOOS + "-" + runtime.GOARCH
	host := "warm.test"
	rt := registry.Route{
		Host: host, Kind: "instance", App: "fastapp", Channel: "edge",
		Status: "ready", Hibernate: true,
	}
	if err := env.reg.Upsert(rt); err != nil {
		t.Fatalf("upsert: %v", err)
	}
	adopt := func(ver string) {
		t.Helper()
		m := release.Manifest{
			App: "fastapp", Version: ver, Kind: "binary",
			Artifacts: map[string]release.Artifact{plat: {}},
		}
		if err := release.Ship(ctx, env.store, "bab", m, map[string][]byte{plat: []byte("build " + ver)}); err != nil {
			t.Fatalf("Ship %s: %v", ver, err)
		}
		if err := release.Promote(ctx, env.store, env.priv, "bab", "fastapp", "edge", ver); err != nil {
			t.Fatalf("Promote %s: %v", ver, err)
		}
		// The root pass: rootRestartRoute's dispatch (bounce nil), systemctl stubbed.
		restart := makeRestartRoute(env.reg, func(string) {}, unitOnly(func(string) error { return nil }), nil)
		if err := release.Pass(ctx, env.reg, env.store, env.pub, env.optDir, restart, release.BakeDir(env.dataDir)); err != nil {
			t.Fatalf("root pass for %s: %v", ver, err)
		}
	}
	backend := &versionedBackend{optDir: env.optDir}
	svc := activator.New(activator.Config{Bucket: "test-bucket", Idle: time.Hour}, backend, env.reg)
	t.Cleanup(func() {
		sctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
		defer cancel()
		svc.Shutdown(sctx)
	})
	request := func() {
		t.Helper()
		rec := httptest.NewRecorder()
		proxy := http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) { w.WriteHeader(http.StatusNoContent) })
		if !svc.ServeHTTP(rec, httptest.NewRequest("GET", "http://"+host+"/", nil), rt, proxy) {
			t.Fatal("the activator did not take the request")
		}
		if rec.Code != http.StatusNoContent {
			t.Fatalf("request = %d, want 204", rec.Code)
		}
	}
	leg := convergeLegFor(svc, env.reg)
	if leg == nil {
		t.Fatal("convergeLegFor returned no leg for a box with an activator")
	}

	// "Was it stopped?" is asked of the spawn count, after a request: Owns
	// stays true while an entry drains, so it cannot tell a spared child from
	// one being stopped. A request against a stopped instance spawns again.
	adopt("v1")
	request() // wakes the instance onto v1
	leg()
	request()
	if got, want := backend.started(), []string{"v1"}; !reflect.DeepEqual(got, want) {
		t.Fatalf("spawned builds = %v, want %v: the leg stopped an instance whose build is the adopted one", got, want)
	}

	adopt("v2") // root: staged, repointed, stamped — and nothing cycled
	if !svc.Owns(host) {
		t.Fatal("the root pass stopped the agent's child — this test no longer models the gap")
	}
	leg()
	deadline := time.Now().Add(10 * time.Second)
	for svc.Owns(host) {
		if time.Now().After(deadline) {
			t.Fatal("the warm instance is still running v1 after the leg: it was adopted onto v2 and nothing cycled it")
		}
		time.Sleep(5 * time.Millisecond)
	}

	request() // the next request wakes it onto v2
	if got, want := backend.started(), []string{"v1", "v2"}; !reflect.DeepEqual(got, want) {
		t.Fatalf("spawned builds = %v, want %v", got, want)
	}
	leg()
	request()
	if got, want := backend.started(), []string{"v1", "v2"}; !reflect.DeepEqual(got, want) {
		t.Fatalf("spawned builds = %v, want %v: the leg cycled the instance again after it had converged", got, want)
	}
}

func TestConvergeLegForABoxWithNoActivator(t *testing.T) {
	if convergeLegFor(nil, nil) != nil {
		t.Fatal("convergeLegFor built a leg with no activator to ask")
	}
}

// The adopt passes' restart dispatch must be built on adoptBounceFn, not
// bounceFn. With the converge leg in place an instance can already be on the
// new build when a slower pass reaches its own restart; the plain Bounce would
// stop that replacement for nothing. A stale child still gets its cycle.
func TestAdoptRestartDispatchSparesAConvergedChild(t *testing.T) {
	env := newFastFixture(t)
	host := "warm.test"
	rt := registry.Route{
		Host: host, Kind: "instance", App: "fastapp", Channel: "edge",
		Status: "ready", Hibernate: true,
	}
	if err := env.reg.Upsert(rt); err != nil {
		t.Fatalf("upsert: %v", err)
	}
	env.stageLive(t, host, "v1")
	backend := &versionedBackend{optDir: env.optDir}
	svc := activator.New(activator.Config{Bucket: "test-bucket", Idle: time.Hour}, backend, env.reg)
	t.Cleanup(func() {
		sctx, cancel := context.WithTimeout(context.Background(), 10*time.Second)
		defer cancel()
		svc.Shutdown(sctx)
	})
	wake := func() {
		t.Helper()
		rec := httptest.NewRecorder()
		proxy := http.HandlerFunc(func(w http.ResponseWriter, _ *http.Request) { w.WriteHeader(http.StatusNoContent) })
		if !svc.ServeHTTP(rec, httptest.NewRequest("GET", "http://"+host+"/", nil), rt, proxy) || rec.Code != http.StatusNoContent {
			t.Fatalf("wake request = %d", rec.Code)
		}
	}
	restart := makeRestartRoute(env.reg, func(string) {}, unitOnly(func(u string) error {
		t.Fatalf("named-unit arm fired for %q on a route that names no unit", u)
		return nil
	}), adoptBounceFn(svc))

	// Asked of the spawn count, after a request — see the note in
	// TestConvergeLegCyclesARootAdoptedInstanceOnARealActivator.
	wake() // exec'd from v1, and live/<host> names v1
	if err := restart(host); err != nil {
		t.Fatalf("restart: %v", err)
	}
	wake()
	if got, want := backend.started(), []string{"v1"}; !reflect.DeepEqual(got, want) {
		t.Fatalf("spawned builds = %v, want %v: the adopt dispatch stopped a child already running the build live/<host> names", got, want)
	}

	env.stageLive(t, host, "v2") // the symlink moves under the running child
	if err := restart(host); err != nil {
		t.Fatalf("restart: %v", err)
	}
	wake()
	if got, want := backend.started(), []string{"v1", "v2"}; !reflect.DeepEqual(got, want) {
		t.Fatalf("spawned builds = %v, want %v: the adopt dispatch left a stale child running", got, want)
	}
}

// A leg past slowLegAfter says so in the journal, with what it was doing; a
// quick one says nothing, because the fast loop runs every two seconds.
func TestLogSlowSpeaksOnlyWhenSlow(t *testing.T) {
	var buf bytes.Buffer
	prev, flags := log.Writer(), log.Flags()
	log.SetOutput(&buf)
	log.SetFlags(0)
	t.Cleanup(func() { log.SetOutput(prev); log.SetFlags(flags) })

	logSlow("fast sweep", time.Now(), "3 routes")
	if buf.Len() != 0 {
		t.Fatalf("a quick leg logged %q", buf.String())
	}
	logSlow("serial pass stages", time.Now().Add(-2*slowLegAfter), "reconcile 9.0s, adopt 1.0s")
	got := buf.String()
	for _, want := range []string{"serial pass stages took 10", "(reconcile 9.0s, adopt 1.0s)"} {
		if !strings.Contains(got, want) {
			t.Fatalf("slow-leg line %q lacks %q", got, want)
		}
	}
}

// The point of the leg's own loop: it acts while the fast tick is stuck in a
// long full pass. The pass is held open here until the leg has cycled the
// drifted instance, which it could never do if it still rode the tick.
func TestConvergeLoopActsWhileTheFullPassIsBusy(t *testing.T) {
	env := newFastFixture(t)
	d := env.deps(nil)
	inPass, unblock := make(chan struct{}), make(chan struct{})
	d.fullPass = func(context.Context) { close(inPass); <-unblock }

	cycled := make(chan string, 1)
	leg := func() {
		convergeAdopted(
			func() map[string]string { return map[string]string{"warm.test": "v2"} },
			func(h string) (*registry.Route, error) { return &registry.Route{Host: h, Version: "v2"}, nil },
			func(h, ver string) bool {
				select {
				case cycled <- h + "@" + ver:
				default:
				}
				return true
			},
			adoptedCyclesPerTick)
	}
	ctx := context.Background()
	now := time.Unix(1754640000, 0)
	fastTick(ctx, d, now) // seeds the serial
	if err := release.BumpSerial(ctx, env.store, now); err != nil {
		t.Fatalf("BumpSerial: %v", err)
	}
	tickDone := make(chan struct{})
	go func() { fastTick(ctx, d, now.Add(10*time.Second)); close(tickDone) }()
	<-inPass // the tick is now stuck in its full pass
	// Started only now, so anything it cycles was cycled DURING the pass.
	stop := make(chan struct{})
	defer close(stop)
	go runConvergeLoop(leg, 5*time.Millisecond, stop)

	select {
	case got := <-cycled:
		if got != "warm.test@v2" {
			t.Fatalf("cycled %q", got)
		}
	case <-time.After(10 * time.Second):
		t.Fatal("the converge leg waited for the full pass")
	}
	close(unblock)
	<-tickDone
}