picoclaw/pkg/channels/buzz/seen_store.go

package buzz

import (
	"encoding/json"
	"errors"
	"os"
	"path/filepath"
	"sync"

	"github.com/nbd-wtf/go-nostr"

	"github.com/sipeed/picoclaw/pkg/config"
)

// recentIDWindow is how many recently-handled event IDs we keep per channel to
// survive the inclusive-Since boundary (NIP-01 redelivers events with
// created_at >= Since) and relays that silently ignore the Since filter.
// A few hundred is ample: it covers the burst of events at the high-water
// timestamp plus any stragglers a misbehaving relay re-sends.
const recentIDWindow = 256

// channelHighWater records the highest seen event created_at and a bounded
// ring of recently handled event IDs for a single Buzz channel ("h" tag).
type channelHighWater struct {
	MaxCreatedAt int64    `json:"max_created_at"`
	RecentIDs    []string `json:"recent_ids,omitempty"`
}

// seenStore persists per-channel high-water state so a service restart does
// not redeliver already-handled kind:9 events to HandleInboundContext.
//
// The store tracks, per "h" tag value:
//   - MaxCreatedAt: the largest event.CreatedAt we have handled, used to set a
//     Since filter on resubscribe so the relay skips most stored history.
//   - RecentIDs: a bounded ring of the most recently handled event IDs. This
//     is the authoritative dedup guard: Since is inclusive in NIP-01, so events
//     at exactly the high-water timestamp are redelivered, and some relays
//     ignore Since entirely. The ring absorbs both cases without unbounded
//     growth.
//
// The store is JSON-backed under ~/.picoclaw/buzz/seen-<channel-name>.json,
// matching the persistence pattern used by the wecom reqID store.
type seenStore struct {
	mu        sync.Mutex
	path      string
	highWater map[string]*channelHighWater
}

// newSeenStore opens (or creates) the persistent seen-state store at path.
// If path is empty it defaults to ~/.picoclaw/buzz/seen-<defaultName>.json.
// A missing file is not an error; the store starts empty.
func newSeenStore(path, defaultName string) *seenStore {
	if path == "" {
		path = defaultSeenStorePath(defaultName)
	}
	s := &seenStore{
		path:      path,
		highWater: make(map[string]*channelHighWater),
	}
	_ = s.load()
	return s
}

func defaultSeenStorePath(channelName string) string {
	home, err := os.UserHomeDir()
	if err != nil || home == "" {
		return filepath.Join(os.TempDir(), "picoclaw-buzz-seen-"+channelName+".json")
	}
	name := channelName
	if name == "" {
		name = config.ChannelBuzz
	}
	return filepath.Join(home, ".picoclaw", "buzz", "seen-"+name+".json")
}

// highWaterMark returns the maximum seen CreatedAt across all channels, or
// zero if nothing has been recorded. Callers use it to set the subscription
// Since filter so the relay skips events older than what we have handled.
func (s *seenStore) highWaterMark() nostr.Timestamp {
	s.mu.Lock()
	defer s.mu.Unlock()
	var max int64
	for _, hw := range s.highWater {
		if hw.MaxCreatedAt > max {
			max = hw.MaxCreatedAt
		}
	}
	return nostr.Timestamp(max)
}

// seen reports whether an event has already been handled for chatID. Events
// older than the channel's high-water timestamp are necessarily replayed
// history. Events at the inclusive high-water boundary require an ID lookup.
func (s *seenStore) seen(chatID, eventID string, createdAt nostr.Timestamp) bool {
	if eventID == "" {
		return false
	}
	s.mu.Lock()
	defer s.mu.Unlock()
	hw, ok := s.highWater[chatID]
	if !ok {
		return false
	}
	if int64(createdAt) < hw.MaxCreatedAt {
		return true
	}
	for _, id := range hw.RecentIDs {
		if id == eventID {
			return true
		}
	}
	return false
}

// markHandled records that eventID (with createdAt) has been handled for
// chatID and persists the store. It updates MaxCreatedAt when createdAt is
// newer and trims the RecentIDs ring to recentIDWindow entries.
func (s *seenStore) markHandled(chatID, eventID string, createdAt nostr.Timestamp) error {
	if chatID == "" || eventID == "" {
		return nil
	}
	s.mu.Lock()
	defer s.mu.Unlock()
	hw, ok := s.highWater[chatID]
	if !ok {
		hw = &channelHighWater{}
		s.highWater[chatID] = hw
	}
	ts := int64(createdAt)
	if ts > hw.MaxCreatedAt {
		hw.MaxCreatedAt = ts
	}
	// Avoid duplicate entries in the ring.
	for _, id := range hw.RecentIDs {
		if id == eventID {
			return s.saveLocked()
		}
	}
	hw.RecentIDs = append(hw.RecentIDs, eventID)
	if len(hw.RecentIDs) > recentIDWindow {
		// Drop the oldest entries, keeping only the most recent window.
		drop := len(hw.RecentIDs) - recentIDWindow
		hw.RecentIDs = hw.RecentIDs[drop:]
	}
	return s.saveLocked()
}

// load reads the persisted high-water state. A missing file is not an error.
func (s *seenStore) load() error {
	s.mu.Lock()
	defer s.mu.Unlock()

	data, err := os.ReadFile(s.path)
	if err != nil {
		if errors.Is(err, os.ErrNotExist) {
			return nil
		}
		return err
	}
	var hw map[string]*channelHighWater
	if err := json.Unmarshal(data, &hw); err != nil {
		return err
	}
	s.highWater = hw
	if s.highWater == nil {
		s.highWater = make(map[string]*channelHighWater)
	}
	return nil
}

// saveLocked writes the high-water state to disk. Caller must hold s.mu.
func (s *seenStore) saveLocked() error {
	if err := os.MkdirAll(filepath.Dir(s.path), 0o700); err != nil {
		return err
	}
	data, err := json.MarshalIndent(s.highWater, "", "  ")
	if err != nil {
		return err
	}
	tmp := s.path + ".tmp"
	if err := os.WriteFile(tmp, data, 0o600); err != nil {
		return err
	}
	return os.Rename(tmp, s.path)
}