290 lines
9.0 KiB
Go
290 lines
9.0 KiB
Go
package webui
|
|
|
|
import (
|
|
"context"
|
|
"encoding/base64"
|
|
"encoding/json"
|
|
"log/slog"
|
|
"net/http"
|
|
|
|
"github.com/samcharles93/archie-core/internal/domain/storecontract"
|
|
"github.com/samcharles93/archie-core/internal/events"
|
|
"github.com/samcharles93/archie-core/internal/logging"
|
|
)
|
|
|
|
// sseBacklogPageSize bounds one EventsSince fetch during catch-up.
|
|
const sseBacklogPageSize = 200
|
|
|
|
// sseNoiseKinds are event kinds published for internal wiring rather than
|
|
// for a dashboard viewer: turn_completed fires on every chat turn so
|
|
// input-driven curators can wake on it (see events.KindTurnCompleted), and
|
|
// carries no task correlation (TaskID is always 0, Detail a raw session ID)
|
|
// -- on a dashboard where "Live activity" is otherwise almost entirely
|
|
// task lifecycle events, it drowned the signal it sits next to. Curators
|
|
// read the Bus directly and are unaffected by excluding a kind here, at the
|
|
// SSE boundary, rather than at publish.
|
|
var sseNoiseKinds = map[string]bool{
|
|
events.KindTurnCompleted: true,
|
|
}
|
|
|
|
// sseVisible reports whether e should reach an SSE client.
|
|
func sseVisible(e events.Event) bool {
|
|
return !sseNoiseKinds[e.Kind]
|
|
}
|
|
|
|
// handleSSE serves task history, latest resource snapshots, and optional
|
|
// live logs over one connection. Append-only topics resume from independent
|
|
// cursors; state topics replay their latest value on each connection.
|
|
func (s *Server) handleSSE(w http.ResponseWriter, r *http.Request) {
|
|
cursors := sseCursors(r)
|
|
stream, ok := newSSEStream(s.Store, s.Log, w, cursors.Tasks)
|
|
if !ok {
|
|
http.Error(w, "streaming unsupported", http.StatusInternalServerError)
|
|
return
|
|
}
|
|
stream.logsSince = cursors.Logs
|
|
|
|
// Register before reading the backlog. An event published between the
|
|
// backlog read and subscription would otherwise be lost permanently.
|
|
conn, snapshot, stale, unregister := s.registerSSEConn()
|
|
defer unregister()
|
|
var logs <-chan logging.Entry
|
|
if r.URL.Query().Get("topics") == "logs" && s.LogFeed != nil {
|
|
logs = s.LogFeed.Subscribe(r.Context())
|
|
}
|
|
// Headers make an idle stream live before any resource changes.
|
|
w.WriteHeader(http.StatusOK)
|
|
stream.fl.Flush()
|
|
|
|
if !stream.catchUp(r.Context(), "") {
|
|
return
|
|
}
|
|
for _, update := range snapshot {
|
|
if !stream.sendLive(update) {
|
|
return
|
|
}
|
|
}
|
|
if r.URL.Query().Get("topics") == "logs" && !stream.sendLogSnapshot(s.LogFeed) {
|
|
return
|
|
}
|
|
stream.drain(r.Context(), conn, stale, logs)
|
|
}
|
|
|
|
// sseSince is the task-only cursor view retained for legacy clients/tests.
|
|
// Native reconnects prefer Last-Event-ID; deliberate reconnects use ?since=.
|
|
// An invalid or pre-cursor integer degrades to replay from the beginning.
|
|
func sseSince(r *http.Request) string { return sseCursors(r).Tasks }
|
|
|
|
type streamCursors struct {
|
|
Tasks string `json:"tasks"`
|
|
Logs int64 `json:"logs"`
|
|
}
|
|
|
|
// EventSource only sends Last-Event-ID on reconnects it owns. A deliberate
|
|
// reconnect (entering/leaving Logs) supplies the same token in ?since=.
|
|
func sseCursors(r *http.Request) streamCursors {
|
|
value := r.URL.Query().Get("since")
|
|
if header := r.Header.Get("Last-Event-ID"); header != "" {
|
|
value = header
|
|
}
|
|
var cursors streamCursors
|
|
if data, err := base64.RawURLEncoding.DecodeString(value); err == nil {
|
|
_ = json.Unmarshal(data, &cursors)
|
|
} else if _, _, ok := storecontract.ParseEventCursor(value); ok {
|
|
cursors.Tasks = value // legacy task cursor
|
|
}
|
|
if _, _, ok := storecontract.ParseEventCursor(cursors.Tasks); !ok {
|
|
cursors.Tasks = ""
|
|
}
|
|
if cursors.Logs < 0 {
|
|
cursors.Logs = 0
|
|
}
|
|
return cursors
|
|
}
|
|
|
|
func (c streamCursors) id() string {
|
|
encoded, _ := json.Marshal(c)
|
|
return base64.RawURLEncoding.EncodeToString(encoded)
|
|
}
|
|
|
|
// sseStream renders one client's event-stream response: an SSE frame per
|
|
// event, an SSE comment when the persisted backlog can't be read, and the
|
|
// running "since" cursor that deduplicates the backlog against live
|
|
// broadcasts covering the same event.
|
|
type sseStream struct {
|
|
store storecontract.TaskStore
|
|
log *slog.Logger
|
|
w http.ResponseWriter
|
|
fl http.Flusher
|
|
since string
|
|
logsSince int64
|
|
}
|
|
|
|
// newSSEStream builds a stream over w, reporting false if w cannot be
|
|
// flushed incrementally -- SSE does not work without that.
|
|
func newSSEStream(st storecontract.TaskStore, log *slog.Logger, w http.ResponseWriter, since string) (*sseStream, bool) {
|
|
fl, ok := w.(http.Flusher)
|
|
if !ok {
|
|
return nil, false
|
|
}
|
|
w.Header().Set("Content-Type", "text/event-stream")
|
|
w.Header().Set("Cache-Control", "no-cache")
|
|
return &sseStream{store: st, log: log, w: w, fl: fl, since: since}, true
|
|
}
|
|
|
|
// send writes one event as an SSE frame. It reports false on a write
|
|
// error, meaning the connection is gone and the caller must stop.
|
|
func (s *sseStream) send(e events.Event) bool {
|
|
return s.writeLive(liveUpdate{topic: "tasks", data: e}, streamCursors{Tasks: storecontract.EventCursor(e.At, e.ID), Logs: s.logsSince}.id())
|
|
}
|
|
|
|
func (s *sseStream) sendLogSnapshot(feed *logging.Feed) bool {
|
|
if feed == nil {
|
|
return s.sendLive(liveUpdate{topic: "logs-status", data: map[string]bool{"available": false}})
|
|
}
|
|
for _, entry := range feed.Snapshot() {
|
|
if !s.sendLog(entry) {
|
|
return false
|
|
}
|
|
}
|
|
return true
|
|
}
|
|
|
|
func (s *sseStream) sendLog(entry logging.Entry) bool {
|
|
if entry.ID <= s.logsSince {
|
|
return true
|
|
}
|
|
if !s.writeLive(liveUpdate{topic: "logs", data: entry}, streamCursors{Tasks: s.since, Logs: entry.ID}.id()) {
|
|
return false
|
|
}
|
|
s.logsSince = entry.ID
|
|
return true
|
|
}
|
|
|
|
func (s *sseStream) sendLive(update liveUpdate) bool { return s.writeLive(update, "") }
|
|
|
|
func (s *sseStream) writeLive(update liveUpdate, id string) bool {
|
|
body, err := marshalLiveFrame(update)
|
|
if err != nil {
|
|
return false
|
|
}
|
|
prefix := ""
|
|
if id != "" {
|
|
prefix = "id: " + id + "\n"
|
|
}
|
|
if _, err := s.w.Write([]byte(prefix + "data: " + string(body) + "\n\n")); err != nil {
|
|
return false
|
|
}
|
|
s.fl.Flush()
|
|
return true
|
|
}
|
|
|
|
// writeErrorComment reports a backlog fetch failure as an SSE comment.
|
|
// Comments are invisible to EventSource's own event parsing, but writing
|
|
// one and flushing before the handler returns is what actually reaches the
|
|
// client: it ends the response, which is what drives EventSource's
|
|
// reconnect.
|
|
func (s *sseStream) writeErrorComment(err error) {
|
|
_, _ = s.w.Write([]byte(":error " + err.Error() + "\n\n"))
|
|
s.fl.Flush()
|
|
}
|
|
|
|
// catchUp sends every persisted event after s.since, advancing s.since as
|
|
// it goes, until it reaches target ("" means "drain everything currently
|
|
// available"), runs out of backlog, or fails. It reports false on failure
|
|
// -- a fetch error or a dead connection -- meaning the caller must stop.
|
|
func (s *sseStream) catchUp(ctx context.Context, target string) bool {
|
|
for {
|
|
before := s.since
|
|
backlog, err := s.store.EventsSince(ctx, s.since, sseBacklogPageSize)
|
|
if err != nil {
|
|
s.log.Error("sse backlog fetch failed", "error", err, "since", s.since)
|
|
s.writeErrorComment(err)
|
|
return false
|
|
}
|
|
reachedTarget, ok := s.sendPage(backlog, target)
|
|
if !ok {
|
|
return false
|
|
}
|
|
if reachedTarget || len(backlog) < sseBacklogPageSize || s.since == before {
|
|
return true
|
|
}
|
|
}
|
|
}
|
|
|
|
// sendPage sends every event in backlog newer than s.since, advancing it as
|
|
// it goes. reachedTarget reports whether target was reached mid-page --
|
|
// stopping there is correct, since nothing beyond it is needed yet -- and
|
|
// ok reports whether every send succeeded.
|
|
//
|
|
// since advances past a filtered-out event (sseVisible false) exactly as it
|
|
// would for a sent one: EventsSince only ever returns events after since, so
|
|
// leaving since behind a filtered event would make the next catchUp refetch
|
|
// it forever.
|
|
func (s *sseStream) sendPage(backlog []events.Event, target string) (reachedTarget, ok bool) {
|
|
for _, e := range backlog {
|
|
cursor := storecontract.EventCursor(e.At, e.ID)
|
|
if cursor <= s.since {
|
|
continue
|
|
}
|
|
if sseVisible(e) && !s.send(e) {
|
|
return false, false
|
|
}
|
|
s.since = cursor
|
|
if target != "" && cursor >= target {
|
|
return true, true
|
|
}
|
|
}
|
|
return false, true
|
|
}
|
|
|
|
// drain relays broadcast events after the initial catch-up. A broadcast can
|
|
// reach the client before its event lands in the durable store, so each one
|
|
// first fills any persisted gap ahead of it, then is delivered directly if
|
|
// catch-up did not already cover it -- both paths go through the same
|
|
// since-based deduplication.
|
|
func (s *sseStream) drain(ctx context.Context, conn <-chan liveUpdate, stale <-chan struct{}, logs <-chan logging.Entry) {
|
|
for {
|
|
select {
|
|
case <-ctx.Done():
|
|
return
|
|
case <-stale:
|
|
return
|
|
case entry, ok := <-logs:
|
|
if !ok {
|
|
logs = nil
|
|
continue
|
|
}
|
|
if !s.sendLog(entry) {
|
|
return
|
|
}
|
|
case update := <-conn:
|
|
if update.topic == "tasks" {
|
|
e, ok := update.data.(events.Event)
|
|
if ok && !s.relayTask(ctx, e) {
|
|
return
|
|
}
|
|
} else if !s.sendLive(update) {
|
|
return
|
|
}
|
|
}
|
|
}
|
|
}
|
|
|
|
func (s *sseStream) relayTask(ctx context.Context, e events.Event) bool {
|
|
cursor := storecontract.EventCursor(e.At, e.ID)
|
|
if cursor <= s.since {
|
|
return true
|
|
}
|
|
if !s.catchUp(ctx, cursor) {
|
|
return false
|
|
}
|
|
if cursor > s.since {
|
|
if sseVisible(e) && !s.send(e) {
|
|
return false
|
|
}
|
|
s.since = cursor
|
|
}
|
|
return true
|
|
}
|