add: implement auto-partitioning function for `matches` table, integr… by 4RH1T3CT0R7 · Pull Request #1 · 4RH1T3CT0R7/TJudge

4RH1T3CT0R7 · 2026-02-15T15:12:15Z

…ate partition creation at app startup, and improve team membership checks and panic recovery logic

…support for animations, and refine accessibility and transitions

Copilot

Pull request overview

This PR improves runtime resilience and security by adding automatic Postgres partition creation for the matches table at startup, tightening authorization checks for program uploads, and enhancing error/edge-case handling in worker, WebSocket, and domain services.

Changes:

Add a Postgres function + startup hook to auto-create monthly matches partitions.
Enforce team membership for multipart program uploads and add related handler tests.
Improve operational robustness: recover worker panics with respawn logic, add WebSocket broadcast timeout behavior, and introduce a dead-letter queue for malformed queue items.

Reviewed changes

Copilot reviewed 35 out of 36 changed files in this pull request and generated 8 comments.

Show a summary per file

File	Description
web/src/pages/TournamentDetail.tsx	Forces WebSocket hook to be enabled for tournament updates.
web/src/pages/GameView.tsx	Sanitizes markdown link URLs before rendering.
web/src/pages/GameDetail.tsx	Sanitizes markdown link URLs before rendering.
migrations/000024_add_auto_partition_function.up.sql	Adds `create_matches_partition_if_needed()` and runs it once.
migrations/000024_add_auto_partition_function.down.sql	Drops the auto-partition function.
internal/infrastructure/db/db.go	Adds `EnsureMatchPartitions()` startup helper.
cmd/api/main.go	Calls `EnsureMatchPartitions()` during API startup; wires `teamRepo` into `ProgramHandler`.
cmd/worker/main.go	Calls `EnsureMatchPartitions()` during worker startup.
internal/infrastructure/db/match_repository.go	Removes N+1 match fetching by loading all tournament matches in one query.
internal/infrastructure/db/tournament_repository.go	Expands tournament update query to include additional fields.
internal/api/handlers/program.go	Adds `TeamMembershipChecker` and enforces membership on file upload.
internal/api/handlers/program_test.go	Updates handler construction + adds tests for membership enforcement.
internal/domain/team/service.go	Prevents deleting the last-member team during an active tournament.
internal/domain/team/service_test.go	Adds coverage for active-tournament delete restriction.
internal/worker/pool.go	Adds panic recovery with logging + respawn; ensures scaling respects MinWorkers after panic.
internal/websocket/hub.go	Adds a 1s timeout retry before dropping broadcast messages.
internal/infrastructure/queue/queue.go	On unmarshal failure, pushes the raw item to a dead-letter queue and logs details.
internal/domain/rating/elo.go	Clamps Elo rating floor at 0.
internal/domain/rating/elo_test.go	Updates tests to reflect rating floor behavior.
internal/domain/auth/service.go	Adds dummy bcrypt compare on user-not-found and enforces email uniqueness in profile update.
internal/domain/auth/service_test.go	Adds/updates tests for email uniqueness checks.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

Copilot · 2026-02-15T15:20:08Z

internal/worker/pool.go

 	p.wg.Add(1)
 	go func() {
 		defer p.wg.Done()
 		defer p.totalWorkers.Add(-1)
+		defer func() {
+			if r := recover(); r != nil {
+				p.log.Error("Worker panic recovered",
+					zap.Int32("worker_id", current),
+					zap.Any("panic", r),
+					zap.String("stack", string(debug.Stack())),
+				)
+				// Respawn worker if pool is still running and below minimum capacity
+				if p.ctx.Err() == nil {
+					time.AfterFunc(time.Second, func() {
+						if p.ctx.Err() != nil {
+							return // Pool was stopped, do not respawn
+						}
+						if int(p.totalWorkers.Load()) < p.config.MinWorkers {
+							p.log.Info("Respawning worker after panic",
+								zap.Int32("current_workers", p.totalWorkers.Load()),
+								zap.Int("min_workers", p.config.MinWorkers),
+							)
+							p.spawnWorker()
+						}
+					})
+				}


When a worker panics, the goroutine exits without removing its workerCancel from p.workerCancels. Because the respawn path appends a new cancel, repeated panics can grow workerCancels unbounded and make scale-down accounting inaccurate. Consider tracking cancels by worker ID (map) or removing the cancel in a deferred cleanup, and ensure workerCancel() is called on worker exit to release context resources.

Copilot · 2026-02-15T15:20:08Z

internal/websocket/hub.go

+		// Channel full, try with timeout before dropping
+		timer := time.NewTimer(time.Second)
+		defer timer.Stop()
+		select {
+		case h.broadcast <- message:
+		case <-timer.C:
+			h.log.Error("Broadcast channel full, message dropped after 1s timeout",
+				zap.String("tournament_id", tournamentID.String()),
+				zap.String("type", messageType),
+			)
+		}


Broadcast now blocks up to 1s when the broadcast channel is full. This can tie up caller goroutines under load and amplify backpressure (potentially impacting HTTP request latency if broadcasts happen inline). If dropping is acceptable, consider keeping it non-blocking (or using a much shorter timeout / metrics-based shedding) and emitting a counter/metric when drops occur.

Suggested change

// Channel full, try with timeout before dropping

timer := time.NewTimer(time.Second)

defer timer.Stop()

select {

case h.broadcast <- message:

case <-timer.C:

h.log.Error("Broadcast channel full, message dropped after 1s timeout",

zap.String("tournament_id", tournamentID.String()),

zap.String("type", messageType),

)

}

// Channel full, drop message immediately to avoid blocking caller

h.log.Error("Broadcast channel full, message dropped",

zap.String("tournament_id", tournamentID.String()),

zap.String("type", messageType),

)

Copilot · 2026-02-15T15:20:09Z

internal/infrastructure/queue/queue.go

+		qm.log.Error("Failed to unmarshal match, moved to dead-letter queue",
+			zap.Error(err),
+			zap.String("raw_data", result[1]),


Logging raw_data from Redis on JSON unmarshal failure can significantly increase log volume and may unintentionally record large/untrusted payloads. Consider logging a truncated prefix (or a hash/length), and relying on the dead-letter queue entry for full inspection.

Suggested change

qm.log.Error("Failed to unmarshal match, moved to dead-letter queue",

zap.Error(err),

zap.String("raw_data", result[1]),

// Log only a truncated prefix of the raw data to avoid large/untrusted payloads in logs.

rawData := result[1]

const maxLoggedRawDataLen = 512

truncatedRawData := rawData

if len(truncatedRawData) > maxLoggedRawDataLen {

truncatedRawData = truncatedRawData[:maxLoggedRawDataLen]

}

qm.log.Error("Failed to unmarshal match, moved to dead-letter queue",

zap.Error(err),

zap.String("raw_data_truncated", truncatedRawData),

zap.Int("raw_data_length", len(rawData)),

Copilot · 2026-02-15T15:20:09Z

migrations/000024_add_auto_partition_function.up.sql

+    current_partition := 'matches_' || to_char(current_month, 'YYYY_MM');
+    IF NOT EXISTS (
+        SELECT 1 FROM pg_class WHERE relname = current_partition
+    ) THEN


The partition existence check uses pg_class WHERE relname = ..., which is schema-agnostic and doesn’t verify that the found relation is actually a partition of matches. In environments with multiple schemas or an unrelated table with the same name, this can skip creating the needed partition. Consider using to_regclass() with an explicit schema (or joining pg_class to pg_namespace and verifying pg_inherits) to make the check unambiguous.

Copilot · 2026-02-15T15:20:09Z

internal/api/handlers/program.go


 // NewProgramHandler создаёт новый program handler
 func NewProgramHandler(
 	programRepo ProgramRepository,
 	tournamentRepo TournamentParticipantAdder,
 	matchScheduler MatchScheduler,
 	gameLookup GameLookup,
 	matchChecker MatchExistenceChecker,
 	roundChecker RoundCompletionChecker,
+	teamChecker TeamMembershipChecker,
 	uploadDir string,
 	log *logger.Logger,
 ) *ProgramHandler {


NewProgramHandler now requires an extra teamChecker argument, but there is still at least one call site using the old signature (tests/benchmark/api_benchmark_test.go:127), which will break compilation of the test package. Update that call site (and any others) to pass a TeamMembershipChecker (or nil if appropriate).

Copilot · 2026-02-15T15:20:09Z

internal/infrastructure/db/db.go

+// EnsureMatchPartitions создаёт партиции таблицы matches для текущего и следующего месяца.
+// Вызывается при старте приложения для гарантии наличия партиций.
+func (db *DB) EnsureMatchPartitions(ctx context.Context) error {
+	_, err := db.ExecContext(ctx, "SELECT create_matches_partition_if_needed()")
+	if err != nil {
+		db.log.Warn("Failed to ensure match partitions (function may not exist yet)", zap.Error(err))
+		return nil // Non-fatal: partitions may already exist or function not yet migrated
+	}
+	db.log.Info("Match partitions verified")
+	return nil


EnsureMatchPartitions logs and returns nil for all errors, which can silently mask real failures (e.g., permissions/DDL errors) and still lead to runtime INSERT failures once partitions are missing. Consider only suppressing the specific “function does not exist” case, and return other errors so callers can decide whether startup should fail.

Copilot · 2026-02-15T15:20:10Z

cmd/worker/main.go

+	_ = database.EnsureMatchPartitions(context.Background())
+


Startup calls EnsureMatchPartitions with context.Background() and ignores the returned error. If the DB stalls or DDL fails, this can hang startup or hide a critical operational issue (missing partitions). Consider using a bounded startup context (timeout) and handling/logging the returned error explicitly.

Suggested change

_ = database.EnsureMatchPartitions(context.Background())

partitionsCtx, partitionsCancel := context.WithTimeout(context.Background(), 30*time.Second)

defer partitionsCancel()

if err := database.EnsureMatchPartitions(partitionsCtx); err != nil {

log.Fatal("Failed to ensure match partitions", zap.Error(err))

}

Copilot · 2026-02-15T15:20:10Z

cmd/api/main.go

 	}

+	// Обеспечиваем наличие партиций таблицы matches
+	_ = database.EnsureMatchPartitions(context.Background())


Startup calls EnsureMatchPartitions with context.Background() and discards the returned error. If partition creation fails (permissions, lock contention, etc.) you may only discover it later via INSERT failures. Consider using a timeout context and handling the error (even if you choose to continue startup).

Suggested change

_ = database.EnsureMatchPartitions(context.Background())

matchPartitionsCtx, matchPartitionsCancel := context.WithTimeout(context.Background(), 30*time.Second)

defer matchPartitionsCancel()

if err := database.EnsureMatchPartitions(matchPartitionsCtx); err != nil {

log.Error("Failed to ensure match partitions", zap.Error(err))

}

add: implement auto-partitioning function for matches table, integr…

2a5613f

…ate partition creation at app startup, and improve team membership checks and panic recovery logic

4RH1T3CT0R7 requested a review from Copilot February 15, 2026 15:12

Copilot started reviewing on behalf of 4RH1T3CT0R7 February 15, 2026 15:12 View session

add: implement useEscapeKey hook for modal handling, reduce motion …

5d1a674

…support for animations, and refine accessibility and transitions

4RH1T3CT0R7 self-assigned this Feb 15, 2026

4RH1T3CT0R7 merged commit d45992e into main Feb 15, 2026
9 checks passed

4RH1T3CT0R7 deleted the review/comprehensive-audit branch February 15, 2026 15:18

Copilot AI reviewed Feb 15, 2026

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Comments

add: implement auto-partitioning function for `matches` table, integr…#1

add: implement auto-partitioning function for `matches` table, integr…#1
4RH1T3CT0R7 merged 2 commits intomainfrom
review/comprehensive-audit

4RH1T3CT0R7 commented Feb 15, 2026

Uh oh!

Uh oh!

Copilot AI left a comment

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Copilot AI Feb 15, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

-		qm.log.Error("Failed to unmarshal match, moved to dead-letter queue",
-			zap.Error(err),
-			zap.String("raw_data", result[1]),
+		// Log only a truncated prefix of the raw data to avoid large/untrusted payloads in logs.
+		rawData := result[1]
+		const maxLoggedRawDataLen = 512
+		truncatedRawData := rawData
+		if len(truncatedRawData) > maxLoggedRawDataLen {
+			truncatedRawData = truncatedRawData[:maxLoggedRawDataLen]
+		}
+		qm.log.Error("Failed to unmarshal match, moved to dead-letter queue",
+			zap.Error(err),
+			zap.String("raw_data_truncated", truncatedRawData),
+			zap.Int("raw_data_length", len(rawData)),

-	_ = database.EnsureMatchPartitions(context.Background())
+	partitionsCtx, partitionsCancel := context.WithTimeout(context.Background(), 30*time.Second)
+	defer partitionsCancel()
+	if err := database.EnsureMatchPartitions(partitionsCtx); err != nil {
+		log.Fatal("Failed to ensure match partitions", zap.Error(err))
+	}

-	_ = database.EnsureMatchPartitions(context.Background())
+	matchPartitionsCtx, matchPartitionsCancel := context.WithTimeout(context.Background(), 30*time.Second)
+	defer matchPartitionsCancel()
+	if err := database.EnsureMatchPartitions(matchPartitionsCtx); err != nil {
+		log.Error("Failed to ensure match partitions", zap.Error(err))
+	}

Comments

Conversation

4RH1T3CT0R7 commented Feb 15, 2026

Uh oh!

Uh oh!

Copilot AI left a comment

Choose a reason for hiding this comment

Pull request overview

Reviewed changes

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Copilot AI Feb 15, 2026

Choose a reason for hiding this comment

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant