Files
next.orly.dev/pkg/database/save-event.go
mleku 6b98c23606
Some checks failed
Go / build-and-release (push) Has been cancelled
add first draft graph query implementation
2025-12-04 09:28:13 +00:00

445 lines
14 KiB
Go

//go:build !(js && wasm)
package database
import (
"bytes"
"context"
"errors"
"fmt"
"strings"
"github.com/dgraph-io/badger/v4"
"lol.mleku.dev/chk"
"lol.mleku.dev/log"
"next.orly.dev/pkg/database/indexes"
"next.orly.dev/pkg/database/indexes/types"
"git.mleku.dev/mleku/nostr/encoders/event"
"git.mleku.dev/mleku/nostr/encoders/filter"
"git.mleku.dev/mleku/nostr/encoders/hex"
"git.mleku.dev/mleku/nostr/encoders/kind"
"git.mleku.dev/mleku/nostr/encoders/tag"
)
var (
// ErrOlderThanExisting is returned when a candidate event is older than an existing replaceable/addressable event.
ErrOlderThanExisting = errors.New("older than existing event")
// ErrMissingDTag is returned when a parameterized replaceable event lacks the required 'd' tag.
ErrMissingDTag = errors.New("event is missing a d tag identifier")
)
func (d *D) GetSerialsFromFilter(f *filter.F) (
sers types.Uint40s, err error,
) {
// Try p-tag graph optimization first
if CanUsePTagGraph(f) {
log.D.F("GetSerialsFromFilter: trying p-tag graph optimization")
if sers, err = d.QueryPTagGraph(f); err == nil && len(sers) >= 0 {
log.D.F("GetSerialsFromFilter: p-tag graph optimization returned %d serials", len(sers))
return
}
// Fall through to traditional indexes on error
log.D.F("GetSerialsFromFilter: p-tag graph optimization failed, falling back to traditional indexes: %v", err)
err = nil
}
var idxs []Range
if idxs, err = GetIndexesFromFilter(f); chk.E(err) {
return
}
// Pre-allocate slice with estimated capacity to reduce reallocations
sers = make(
types.Uint40s, 0, len(idxs)*100,
) // Estimate 100 serials per index
for _, idx := range idxs {
var s types.Uint40s
if s, err = d.GetSerialsByRange(idx); chk.E(err) {
continue
}
sers = append(sers, s...)
}
return
}
// WouldReplaceEvent checks if the provided event would replace existing events
// based on Nostr's replaceable or parameterized replaceable semantics. It
// returns true if the candidate is newer-or-equal than existing events.
// If an existing event is newer, it returns (false, nil, ErrOlderThanExisting).
// If no conflicts exist, it returns (false, nil, nil).
func (d *D) WouldReplaceEvent(ev *event.E) (bool, types.Uint40s, error) {
// Only relevant for replaceable or parameterized replaceable kinds
if !(kind.IsReplaceable(ev.Kind) || kind.IsParameterizedReplaceable(ev.Kind)) {
return false, nil, nil
}
var f *filter.F
if kind.IsReplaceable(ev.Kind) {
f = &filter.F{
Authors: tag.NewFromBytesSlice(ev.Pubkey),
Kinds: kind.NewS(kind.New(ev.Kind)),
}
} else {
// parameterized replaceable requires 'd' tag
dTag := ev.Tags.GetFirst([]byte("d"))
if dTag == nil {
return false, nil, ErrMissingDTag
}
f = &filter.F{
Authors: tag.NewFromBytesSlice(ev.Pubkey),
Kinds: kind.NewS(kind.New(ev.Kind)),
Tags: tag.NewS(
tag.NewFromAny("d", dTag.Value()),
),
}
}
sers, err := d.GetSerialsFromFilter(f)
if chk.E(err) {
return false, nil, err
}
if len(sers) == 0 {
return false, nil, nil
}
// Determine if any existing event is newer than the candidate
shouldReplace := true
for _, s := range sers {
oldEv, ferr := d.FetchEventBySerial(s)
if chk.E(ferr) {
continue
}
if ev.CreatedAt < oldEv.CreatedAt {
shouldReplace = false
break
}
}
if shouldReplace {
return true, nil, nil
}
return false, nil, ErrOlderThanExisting
}
// SaveEvent saves an event to the database, generating all the necessary indexes.
func (d *D) SaveEvent(c context.Context, ev *event.E) (
replaced bool, err error,
) {
if ev == nil {
err = errors.New("nil event")
return
}
// Reject ephemeral events (kinds 20000-29999) - they should never be stored
if ev.Kind >= 20000 && ev.Kind <= 29999 {
err = errors.New("blocked: ephemeral events should not be stored")
return
}
// Validate kind 3 (follow list) events have at least one p tag
// This prevents storing malformed follow lists that may come from buggy relays
if ev.Kind == 3 {
hasPTag := false
tagCount := 0
if ev.Tags != nil {
tagCount = ev.Tags.Len()
for _, tag := range *ev.Tags {
if tag != nil && tag.Len() >= 2 {
key := tag.Key()
if len(key) == 1 && key[0] == 'p' {
hasPTag = true
break
}
}
}
}
if !hasPTag {
log.W.F("SaveEvent: rejecting kind 3 event without p tags from pubkey %x (total tags: %d, event ID: %x)",
ev.Pubkey, tagCount, ev.ID)
err = errors.New("blocked: kind 3 follow list events must have at least one p tag")
return
}
}
// check if the event already exists
var ser *types.Uint40
if ser, err = d.GetSerialById(ev.ID); err == nil && ser != nil {
err = errors.New("blocked: event already exists: " + hex.Enc(ev.ID[:]))
return
}
// If the error is "id not found", we can proceed with saving the event
if err != nil && strings.Contains(err.Error(), "id not found in database") {
// Reset error since this is expected for new events
err = nil
} else if err != nil {
// For any other error, return it
// log.E.F("error checking if event exists: %s", err)
return
}
// Check if the event has been deleted before allowing resubmission
if err = d.CheckForDeleted(ev, nil); err != nil {
// log.I.F(
// "SaveEvent: rejecting resubmission of deleted event ID=%s: %v",
// hex.Enc(ev.ID), err,
// )
err = fmt.Errorf("blocked: %s", err.Error())
return
}
// check for replacement - only validate, don't delete old events
if kind.IsReplaceable(ev.Kind) || kind.IsParameterizedReplaceable(ev.Kind) {
var werr error
if replaced, _, werr = d.WouldReplaceEvent(ev); werr != nil {
if errors.Is(werr, ErrOlderThanExisting) {
if kind.IsReplaceable(ev.Kind) {
err = errors.New("blocked: event is older than existing replaceable event")
} else {
err = errors.New("blocked: event is older than existing addressable event")
}
return
}
if errors.Is(werr, ErrMissingDTag) {
// keep behavior consistent with previous implementation
err = ErrMissingDTag
return
}
// any other error
return
}
// Note: replaced flag is kept for compatibility but old events are no longer deleted
}
// Get the next sequence number for the event
var serial uint64
if serial, err = d.seq.Next(); chk.E(err) {
return
}
// Generate all indexes for the event
var idxs [][]byte
if idxs, err = GetIndexesForEvent(ev, serial); chk.E(err) {
return
}
// Collect all pubkeys for graph: author + p-tags
// Store with direction indicator: author (0) vs p-tag (1)
type pubkeyWithDirection struct {
serial *types.Uint40
isAuthor bool
}
pubkeysForGraph := make(map[string]pubkeyWithDirection)
// Add author pubkey
var authorSerial *types.Uint40
if authorSerial, err = d.GetOrCreatePubkeySerial(ev.Pubkey); chk.E(err) {
return
}
pubkeysForGraph[hex.Enc(ev.Pubkey)] = pubkeyWithDirection{
serial: authorSerial,
isAuthor: true,
}
// Extract p-tag pubkeys using GetAll
pTags := ev.Tags.GetAll([]byte("p"))
for _, pTag := range pTags {
if pTag.Len() >= 2 {
// Get pubkey from p-tag, handling both binary and hex storage formats
// ValueHex() returns hex regardless of internal storage format
var ptagPubkey []byte
if ptagPubkey, err = hex.Dec(string(pTag.ValueHex())); err == nil && len(ptagPubkey) == 32 {
pkHex := hex.Enc(ptagPubkey)
// Skip if already added as author
if _, exists := pubkeysForGraph[pkHex]; !exists {
var ptagSerial *types.Uint40
if ptagSerial, err = d.GetOrCreatePubkeySerial(ptagPubkey); chk.E(err) {
return
}
pubkeysForGraph[pkHex] = pubkeyWithDirection{
serial: ptagSerial,
isAuthor: false,
}
}
}
}
}
// log.T.F(
// "SaveEvent: generated %d indexes for event %x (kind %d)", len(idxs),
// ev.ID, ev.Kind,
// )
// Create serial resolver for compact encoding
resolver := NewDatabaseSerialResolver(d, d.serialCache)
// Serialize event in compact format using serial references
// This dramatically reduces storage by replacing 32-byte IDs/pubkeys with 5-byte serials
compactData, compactErr := MarshalCompactEvent(ev, resolver)
// Calculate legacy size for comparison (for metrics tracking)
// We marshal to get accurate size comparison
legacyBuf := new(bytes.Buffer)
ev.MarshalBinary(legacyBuf)
legacySize := legacyBuf.Len()
if compactErr != nil {
// Fall back to legacy format if compact encoding fails
log.W.F("SaveEvent: compact encoding failed, using legacy format: %v", compactErr)
compactData = legacyBuf.Bytes()
} else {
// Track storage savings
TrackCompactSaving(legacySize, len(compactData))
log.T.F("SaveEvent: compact %d bytes vs legacy %d bytes (saved %d bytes, %.1f%%)",
len(compactData), legacySize, legacySize-len(compactData),
float64(legacySize-len(compactData))/float64(legacySize)*100.0)
}
// Start a transaction to save the event and all its indexes
err = d.Update(
func(txn *badger.Txn) (err error) {
// Pre-allocate key buffer to avoid allocations in loop
ser := new(types.Uint40)
if err = ser.Set(serial); chk.E(err) {
return
}
// Save each index
for _, key := range idxs {
if err = txn.Set(key, nil); chk.E(err) {
return
}
}
// Store the SerialEventId mapping (serial -> full 32-byte event ID)
// This is required for reconstructing compact events
if err = d.StoreEventIdSerial(txn, serial, ev.ID); chk.E(err) {
return
}
// Cache the event ID mapping
d.serialCache.CacheEventId(serial, ev.ID)
// Store compact event with cmp prefix
// Format: cmp|serial|compact_event_data
// This is the only storage format - legacy evt/sev/aev/rev prefixes
// are handled by migration and no longer written for new events
cmpKeyBuf := new(bytes.Buffer)
if err = indexes.CompactEventEnc(ser).MarshalWrite(cmpKeyBuf); chk.E(err) {
return
}
if err = txn.Set(cmpKeyBuf.Bytes(), compactData); chk.E(err) {
return
}
// Create graph edges between event and all related pubkeys
// This creates bidirectional edges: event->pubkey and pubkey->event
// Include the event kind and direction for efficient graph queries
eventKind := new(types.Uint16)
eventKind.Set(ev.Kind)
for _, pkInfo := range pubkeysForGraph {
// Determine direction for forward edge (event -> pubkey perspective)
directionForward := new(types.Letter)
// Determine direction for reverse edge (pubkey -> event perspective)
directionReverse := new(types.Letter)
if pkInfo.isAuthor {
// Event author relationship
directionForward.Set(types.EdgeDirectionAuthor) // 0: author
directionReverse.Set(types.EdgeDirectionAuthor) // 0: is author of event
} else {
// P-tag relationship
directionForward.Set(types.EdgeDirectionPTagOut) // 1: event references pubkey (outbound)
directionReverse.Set(types.EdgeDirectionPTagIn) // 2: pubkey is referenced (inbound)
}
// Create event -> pubkey edge (with kind and direction)
epgKeyBuf := new(bytes.Buffer)
if err = indexes.EventPubkeyGraphEnc(ser, pkInfo.serial, eventKind, directionForward).MarshalWrite(epgKeyBuf); chk.E(err) {
return
}
// Make a copy of the key bytes to avoid buffer reuse issues in txn
epgKey := make([]byte, epgKeyBuf.Len())
copy(epgKey, epgKeyBuf.Bytes())
if err = txn.Set(epgKey, nil); chk.E(err) {
return
}
// Create pubkey -> event edge (reverse, with kind and direction for filtering)
pegKeyBuf := new(bytes.Buffer)
if err = indexes.PubkeyEventGraphEnc(pkInfo.serial, eventKind, directionReverse, ser).MarshalWrite(pegKeyBuf); chk.E(err) {
return
}
if err = txn.Set(pegKeyBuf.Bytes(), nil); chk.E(err) {
return
}
}
// Create event-to-event graph edges for e-tags
// This enables thread traversal and finding replies/reactions to events
eTags := ev.Tags.GetAll([]byte("e"))
for _, eTag := range eTags {
if eTag.Len() >= 2 {
// Get event ID from e-tag, handling both binary and hex storage formats
var targetEventID []byte
if targetEventID, err = hex.Dec(string(eTag.ValueHex())); err != nil || len(targetEventID) != 32 {
continue
}
// Look up the target event's serial (if it exists in our database)
var targetSerial *types.Uint40
if targetSerial, err = d.GetSerialById(targetEventID); err != nil {
// Target event not in our database - skip edge creation
// This is normal for replies to events we don't have
err = nil
continue
}
// Create forward edge: source event -> target event (outbound e-tag)
directionOut := new(types.Letter)
directionOut.Set(types.EdgeDirectionETagOut)
eegKeyBuf := new(bytes.Buffer)
if err = indexes.EventEventGraphEnc(ser, targetSerial, eventKind, directionOut).MarshalWrite(eegKeyBuf); chk.E(err) {
return
}
// Make a copy of the key bytes to avoid buffer reuse issues in txn
eegKey := make([]byte, eegKeyBuf.Len())
copy(eegKey, eegKeyBuf.Bytes())
if err = txn.Set(eegKey, nil); chk.E(err) {
return
}
// Create reverse edge: target event -> source event (inbound e-tag)
directionIn := new(types.Letter)
directionIn.Set(types.EdgeDirectionETagIn)
geeKeyBuf := new(bytes.Buffer)
if err = indexes.GraphEventEventEnc(targetSerial, eventKind, directionIn, ser).MarshalWrite(geeKeyBuf); chk.E(err) {
return
}
if err = txn.Set(geeKeyBuf.Bytes(), nil); chk.E(err) {
return
}
}
}
return
},
)
if err != nil {
return
}
// Process deletion events to actually delete the referenced events
if ev.Kind == kind.Deletion.K {
if err = d.ProcessDelete(ev, nil); chk.E(err) {
log.W.F("failed to process deletion for event %x: %v", ev.ID, err)
// Don't return error - the deletion event was saved successfully
err = nil
}
}
// Invalidate query cache since a new event was stored
// This ensures subsequent queries will see the new event
if d.queryCache != nil {
d.queryCache.Invalidate()
// log.T.F("SaveEvent: invalidated query cache")
}
return
}