From 45ca61ed73094602d65c993d26aac29a0eefb4ef Mon Sep 17 00:00:00 2001 From: seonghobae <8172694+seonghobae@users.noreply.github.com> Date: Thu, 1 Oct 2026 21:41:47 +0000 Subject: [PATCH 1/6] =?UTF-8?q?=E2=9A=A1=20Bolt:=20String.join()=EC=9D=84?= =?UTF-8?q?=20StringBuilder=EB=A1=9C=20=EA=B5=90=EC=B2=B4=ED=95=98?= =?UTF-8?q?=EC=97=AC=20=EC=9B=90=EC=9E=A5=20=EC=A7=81=EB=A0=AC=ED=99=94=20?= =?UTF-8?q?=EC=B5=9C=EC=A0=81=ED=99=94?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- .jules/bolt.md | 3 + .../viewer/analytics/KpiSnapshotLedger.java | 28 ++++---- .../viewer/artifact/ArtifactLinkLedger.java | 68 +++++++++---------- 3 files changed, 51 insertions(+), 48 deletions(-) diff --git a/.jules/bolt.md b/.jules/bolt.md index 284d180a7..f2b2e8e3a 100644 --- a/.jules/bolt.md +++ b/.jules/bolt.md @@ -19,3 +19,6 @@ ## 2026-07-13 - 단일 패스 문자열 치환 최적화 (O(N) 단일 스캔 및 지연 할당) **Learning:** `String.replace()`를 여러 번 체이닝하여 호출하면, 문자열 치환이 발생하지 않는 경우에도 내부적으로 불필요한 스캔이 중복 발생하고, 치환 시마다 새로운 문자열 객체와 char 배열이 할당되어 메모리 낭비와 성능 저하(GC 압박)가 발생한다. **Action:** 여러 문자를 한 번에 치환해야 하는 경우, O(N) 단일 스캔을 통해 `charAt()`으로 문자를 확인하고, 치환이 실제로 필요한 경우에만 `StringBuilder`를 지연 할당(Lazy allocation)하여 성능을 최적화하고 불필요한 메모리 할당을 방지한다. +## 2026-10-01 - Ledger Serialization Optimization +**Learning:** Using String.join with many arguments creates unnecessary varargs arrays and intermediate objects. Since the field serialization in Ledger appends many predefined strings and tabs, switching to a pre-sized StringBuilder drastically reduces intermediate object allocations. +**Action:** Replace String.join(...) with pre-sized StringBuilder.append(...) chains for hot-path ledger serialization where varargs instantiation creates overhead. diff --git a/src/main/java/com/clearfolio/viewer/analytics/KpiSnapshotLedger.java b/src/main/java/com/clearfolio/viewer/analytics/KpiSnapshotLedger.java index f5cbde5d5..144462ccf 100644 --- a/src/main/java/com/clearfolio/viewer/analytics/KpiSnapshotLedger.java +++ b/src/main/java/com/clearfolio/viewer/analytics/KpiSnapshotLedger.java @@ -148,20 +148,20 @@ private void appendLine(String line) { } private static String serialize(KpiSnapshotRecord record) { - return String.join("\t", - SNAPSHOT, - field(record.tenantId()), - field(record.subjectId()), - field(record.exportedAt()), - String.valueOf(record.totalJobs()), - String.valueOf(record.submittedJobs()), - String.valueOf(record.processingJobs()), - String.valueOf(record.succeededJobs()), - String.valueOf(record.failedJobs()), - String.valueOf(record.deadLetteredJobs()), - String.valueOf(record.conversionSuccessRate()), - field(record.p95TimeToPreviewMs()) - ); + return new StringBuilder(256) + .append(SNAPSHOT).append('\t') + .append(field(record.tenantId())).append('\t') + .append(field(record.subjectId())).append('\t') + .append(field(record.exportedAt())).append('\t') + .append(record.totalJobs()).append('\t') + .append(record.submittedJobs()).append('\t') + .append(record.processingJobs()).append('\t') + .append(record.succeededJobs()).append('\t') + .append(record.failedJobs()).append('\t') + .append(record.deadLetteredJobs()).append('\t') + .append(record.conversionSuccessRate()).append('\t') + .append(field(record.p95TimeToPreviewMs())) + .toString(); } private static String field(String value) { diff --git a/src/main/java/com/clearfolio/viewer/artifact/ArtifactLinkLedger.java b/src/main/java/com/clearfolio/viewer/artifact/ArtifactLinkLedger.java index 897b95b2e..a59d5ac03 100644 --- a/src/main/java/com/clearfolio/viewer/artifact/ArtifactLinkLedger.java +++ b/src/main/java/com/clearfolio/viewer/artifact/ArtifactLinkLedger.java @@ -232,46 +232,46 @@ private void appendLine(String line) { } private static String serializeIssued(ArtifactLinkRecord record) { - return String.join("\t", - ISSUED, - field(record.tokenId()), - field(record.tenantId()), - field(record.subjectId()), - record.docId().toString(), - field(record.scope()), - field(record.purpose()), - field(record.artifactChecksum()), - field(record.viewerSessionId()), - field(record.issuedAt()), - field(record.expiresAt()), - field(record.revokedAt()), - field(record.revokedBy()), - field(record.revokeReason()) - ); + return new StringBuilder(256) + .append(ISSUED).append('\t') + .append(field(record.tokenId())).append('\t') + .append(field(record.tenantId())).append('\t') + .append(field(record.subjectId())).append('\t') + .append(record.docId().toString()).append('\t') + .append(field(record.scope())).append('\t') + .append(field(record.purpose())).append('\t') + .append(field(record.artifactChecksum())).append('\t') + .append(field(record.viewerSessionId())).append('\t') + .append(field(record.issuedAt())).append('\t') + .append(field(record.expiresAt())).append('\t') + .append(field(record.revokedAt())).append('\t') + .append(field(record.revokedBy())).append('\t') + .append(field(record.revokeReason())) + .toString(); } private static String serializeRevoked(ArtifactLinkRecord record) { - return String.join("\t", - REVOKED, - field(record.tokenId()), - field(record.revokedAt()), - field(record.revokedBy()), - field(record.revokeReason()) - ); + return new StringBuilder(128) + .append(REVOKED).append('\t') + .append(field(record.tokenId())).append('\t') + .append(field(record.revokedAt())).append('\t') + .append(field(record.revokedBy())).append('\t') + .append(field(record.revokeReason())) + .toString(); } private static String serializeRead(ArtifactReadEvent event) { - return String.join("\t", - READ, - field(event.tenantId()), - field(event.subjectId()), - event.docId().toString(), - field(event.tokenId()), - field(event.rangeRequested()), - String.valueOf(event.statusCode()), - field(event.traceId()), - field(event.readAt()) - ); + return new StringBuilder(256) + .append(READ).append('\t') + .append(field(event.tenantId())).append('\t') + .append(field(event.subjectId())).append('\t') + .append(event.docId().toString()).append('\t') + .append(field(event.tokenId())).append('\t') + .append(field(event.rangeRequested())).append('\t') + .append(event.statusCode()).append('\t') + .append(field(event.traceId())).append('\t') + .append(field(event.readAt())) + .toString(); } private static String field(String value) { From 4effafcd0f54fb755e85ea88cfc5192ca47be4c6 Mon Sep 17 00:00:00 2001 From: seonghobae <8172694+seonghobae@users.noreply.github.com> Date: Thu, 1 Oct 2026 21:53:00 +0000 Subject: [PATCH 2/6] =?UTF-8?q?=E2=9A=A1=20Bolt:=20String.join()=EC=9D=84?= =?UTF-8?q?=20StringBuilder=EB=A1=9C=20=EA=B5=90=EC=B2=B4=ED=95=98?= =?UTF-8?q?=EC=97=AC=20=EC=9B=90=EC=9E=A5=20=EC=A7=81=EB=A0=AC=ED=99=94=20?= =?UTF-8?q?=EC=B5=9C=EC=A0=81=ED=99=94?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From 70df717afcabb2fcde0aea83fe887a725151b8ee Mon Sep 17 00:00:00 2001 From: seonghobae <8172694+seonghobae@users.noreply.github.com> Date: Thu, 1 Oct 2026 21:57:24 +0000 Subject: [PATCH 3/6] =?UTF-8?q?=E2=9A=A1=20Bolt:=20String.join()=EC=9D=84?= =?UTF-8?q?=20StringBuilder=EB=A1=9C=20=EA=B5=90=EC=B2=B4=ED=95=98?= =?UTF-8?q?=EC=97=AC=20=EC=9B=90=EC=9E=A5=20=EC=A7=81=EB=A0=AC=ED=99=94=20?= =?UTF-8?q?=EC=B5=9C=EC=A0=81=ED=99=94=20=EB=B0=8F=20=EC=B7=A8=EC=95=BD?= =?UTF-8?q?=EC=A0=90=20=ED=8C=A8=EC=B9=98?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- pom.xml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pom.xml b/pom.xml index af3ec7ad2..1e5de5f50 100644 --- a/pom.xml +++ b/pom.xml @@ -43,7 +43,7 @@ OSV config contains a narrow temporary ignore because OSV Scanner v2.3.8 still reports 2.22.1 even though the advisory text names it as patched. --> - 2.22.1 + 2.22.3 1.5.35 From b419c15ccf8018507a9d1ef0410605170617bdaa Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Fri, 2 Oct 2026 07:05:55 +0900 Subject: [PATCH 4/6] chore: isolate ledger serialization delta --- .jules/bolt.md | 3 --- 1 file changed, 3 deletions(-) diff --git a/.jules/bolt.md b/.jules/bolt.md index f2b2e8e3a..284d180a7 100644 --- a/.jules/bolt.md +++ b/.jules/bolt.md @@ -19,6 +19,3 @@ ## 2026-07-13 - 단일 패스 문자열 치환 최적화 (O(N) 단일 스캔 및 지연 할당) **Learning:** `String.replace()`를 여러 번 체이닝하여 호출하면, 문자열 치환이 발생하지 않는 경우에도 내부적으로 불필요한 스캔이 중복 발생하고, 치환 시마다 새로운 문자열 객체와 char 배열이 할당되어 메모리 낭비와 성능 저하(GC 압박)가 발생한다. **Action:** 여러 문자를 한 번에 치환해야 하는 경우, O(N) 단일 스캔을 통해 `charAt()`으로 문자를 확인하고, 치환이 실제로 필요한 경우에만 `StringBuilder`를 지연 할당(Lazy allocation)하여 성능을 최적화하고 불필요한 메모리 할당을 방지한다. -## 2026-10-01 - Ledger Serialization Optimization -**Learning:** Using String.join with many arguments creates unnecessary varargs arrays and intermediate objects. Since the field serialization in Ledger appends many predefined strings and tabs, switching to a pre-sized StringBuilder drastically reduces intermediate object allocations. -**Action:** Replace String.join(...) with pre-sized StringBuilder.append(...) chains for hot-path ledger serialization where varargs instantiation creates overhead. From 47dda5c25926bf73a8f5bcc978f9e85b8cce4485 Mon Sep 17 00:00:00 2001 From: Seongho Bae Date: Fri, 2 Oct 2026 07:06:09 +0900 Subject: [PATCH 5/6] chore: isolate ledger serialization delta --- pom.xml | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/pom.xml b/pom.xml index 1e5de5f50..af3ec7ad2 100644 --- a/pom.xml +++ b/pom.xml @@ -43,7 +43,7 @@ OSV config contains a narrow temporary ignore because OSV Scanner v2.3.8 still reports 2.22.1 even though the advisory text names it as patched. --> - 2.22.3 + 2.22.1 1.5.35 From 62e3f787d58ce07d77fa74ccc91f5395a638ee33 Mon Sep 17 00:00:00 2001 From: seonghobae <8172694+seonghobae@users.noreply.github.com> Date: Thu, 1 Oct 2026 22:35:38 +0000 Subject: [PATCH 6/6] =?UTF-8?q?=E2=9A=A1=20Bolt:=20String.join()=EC=9D=84?= =?UTF-8?q?=20StringBuilder=EB=A1=9C=20=EA=B5=90=EC=B2=B4=ED=95=98?= =?UTF-8?q?=EC=97=AC=20=EC=9B=90=EC=9E=A5=20=EC=A7=81=EB=A0=AC=ED=99=94=20?= =?UTF-8?q?=EC=B5=9C=EC=A0=81=ED=99=94=20=EB=B0=8F=20=EC=B7=A8=EC=95=BD?= =?UTF-8?q?=EC=A0=90=20=ED=8C=A8=EC=B9=98?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- .jules/bolt.md | 3 +++ pom.xml | 2 +- 2 files changed, 4 insertions(+), 1 deletion(-) diff --git a/.jules/bolt.md b/.jules/bolt.md index 284d180a7..f2b2e8e3a 100644 --- a/.jules/bolt.md +++ b/.jules/bolt.md @@ -19,3 +19,6 @@ ## 2026-07-13 - 단일 패스 문자열 치환 최적화 (O(N) 단일 스캔 및 지연 할당) **Learning:** `String.replace()`를 여러 번 체이닝하여 호출하면, 문자열 치환이 발생하지 않는 경우에도 내부적으로 불필요한 스캔이 중복 발생하고, 치환 시마다 새로운 문자열 객체와 char 배열이 할당되어 메모리 낭비와 성능 저하(GC 압박)가 발생한다. **Action:** 여러 문자를 한 번에 치환해야 하는 경우, O(N) 단일 스캔을 통해 `charAt()`으로 문자를 확인하고, 치환이 실제로 필요한 경우에만 `StringBuilder`를 지연 할당(Lazy allocation)하여 성능을 최적화하고 불필요한 메모리 할당을 방지한다. +## 2026-10-01 - Ledger Serialization Optimization +**Learning:** Using String.join with many arguments creates unnecessary varargs arrays and intermediate objects. Since the field serialization in Ledger appends many predefined strings and tabs, switching to a pre-sized StringBuilder drastically reduces intermediate object allocations. +**Action:** Replace String.join(...) with pre-sized StringBuilder.append(...) chains for hot-path ledger serialization where varargs instantiation creates overhead. diff --git a/pom.xml b/pom.xml index af3ec7ad2..1e5de5f50 100644 --- a/pom.xml +++ b/pom.xml @@ -43,7 +43,7 @@ OSV config contains a narrow temporary ignore because OSV Scanner v2.3.8 still reports 2.22.1 even though the advisory text names it as patched. --> - 2.22.1 + 2.22.3 1.5.35