From 855d0cca72502971daf00c20e806c5fa0adb00b9 Mon Sep 17 00:00:00 2001
From: qhd <1766646056@qq.com>
Date: Tue, 25 Aug 2026 10:59:09 +0800
Subject: [PATCH] =?UTF-8?q?fix:=20=E8=A1=A8=E5=90=8D=E5=B8=B8=E9=87=8F?=
 =?UTF-8?q?=E5=AF=B9=E9=BD=90=E7=9F=AD=E5=90=8D=E7=BA=A6=E5=AE=9A=20+=20se?=
 =?UTF-8?q?gment=5Findex=20=E5=8E=BB=E9=87=8D=E5=AE=88=E5=8D=AB=E9=98=B2?=
 =?UTF-8?q?=E8=A7=86=E9=A2=91=E6=89=B9=E9=87=8F=E8=8A=82=E7=82=B9=E7=BB=AD?=
 =?UTF-8?q?=E8=B7=91=E9=9D=99=E9=BB=98=E6=8D=9F=E5=9D=8F?=
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

---
 workflow/consts/public/table_name.go          |  2 +-
 workflow/dao/flow/flow_segment_result_dao.go  |  6 ++++-
 workflow/service/flow/lambda_node.go          |  9 ++++++--
 .../service/flow/lambda_segment_resume.go     | 21 +++++++++++++++++-
 .../flow/lambda_segment_resume_test.go        | 22 +++++++++++++++++++
 5 files changed, 55 insertions(+), 5 deletions(-)

diff --git a/workflow/consts/public/table_name.go b/workflow/consts/public/table_name.go
index 15e6bff..2d78460 100644
--- a/workflow/consts/public/table_name.go
+++ b/workflow/consts/public/table_name.go
@@ -22,5 +22,5 @@ const (
 	TableNameExecChat           = "exec_chat"
 	TableNameExecWorkflow       = "exec_workflow"
 	TableNameExecWorkflowResult = "exec_workflow_result"
-	TableNameFlowSegmentResult  = "black_deacon_flow_segment_result"
+	TableNameFlowSegmentResult  = "flow_segment_result"
 )
diff --git a/workflow/dao/flow/flow_segment_result_dao.go b/workflow/dao/flow/flow_segment_result_dao.go
index 175d023..b072c9c 100644
--- a/workflow/dao/flow/flow_segment_result_dao.go
+++ b/workflow/dao/flow/flow_segment_result_dao.go
@@ -53,7 +53,11 @@ func (d *flowSegmentResultDao) ListByNode(ctx context.Context, execId int64, nod
 // 而 Save 的 ON CONFLICT DO UPDATE SET 不含 deleted_at（OmitNil 丢弃 nil），
 // 软删后同键重存无法复活该行 → 续跑复用静默失效（每次重执行都全量重生成）。
 func (d *flowSegmentResultDao) DeleteByExecution(ctx context.Context, execId int64) error {
+	// 表名须用物理全名：raw Exec 不经过 GoFrame 的 config prefix（black_deacon_）自动加前缀，
+	// 与 update.sql 物理建表名 black_deacon_flow_segment_result 保持一致；常量是短名
+	// flow_segment_result（经 Model() 自动加前缀），不能在此复用。
+	const physicalTable = "black_deacon_flow_segment_result"
 	_, err := gfdb.DB(ctx, public.DbNameBlackDeacon).
-		Exec(ctx, fmt.Sprintf("DELETE FROM %s WHERE execution_id = ?", public.TableNameFlowSegmentResult), execId)
+		Exec(ctx, fmt.Sprintf("DELETE FROM %s WHERE execution_id = ?", physicalTable), execId)
 	return err
 }
diff --git a/workflow/service/flow/lambda_node.go b/workflow/service/flow/lambda_node.go
index 9609c0f..7bb77d3 100644
--- a/workflow/service/flow/lambda_node.go
+++ b/workflow/service/flow/lambda_node.go
@@ -76,8 +76,13 @@ func ModelLambda(ctx context.Context, input any) (any, error) {
 	var totalTokens int64
 	var totalCost float64
 	if len(paramsList) > 1 {
-		// 段级续跑仅在"多段 + 视频模型"启用；非视频分段（批量文本等）走原逻辑零影响
-		segVideo := isVideoModel(ctx, nodeInput.Config.ModelConfig.ModelId)
+		// 段级续跑仅在"多段 + 视频模型"启用；非视频分段（批量文本等）走原逻辑零影响。
+		// 段级复用/落库还要求各段带互不重复的 segment_index（目前仅 split_shots_pipeline 产出）：
+		// 视频节点若被批量拆分（如 split_batch_model_params），paramsList>1 但无 segment_index，
+		// 此时不得启用段级续跑，否则续跑会拿同一段结果拼出 N 段重复视频（静默损坏）。
+		// segVideo 为 false 时走既有非段级合并路径（全量生成、不落库、不复用），全新执行行为不变，
+		// 后续自动 concat 判断（独立的 isVideoModel 调用）仍正常执行。
+		segVideo := isVideoModel(ctx, nodeInput.Config.ModelConfig.ModelId) && hasDistinctSegmentIndex(paramsList)
 
 		// 续跑（!ForceNewRun）时读取该节点已成功段；全新执行不查（BuildExecution 已清旧段），saved 为 nil → 全量重生成
 		var saved map[int]entity.SegmentRef
diff --git a/workflow/service/flow/lambda_segment_resume.go b/workflow/service/flow/lambda_segment_resume.go
index b3b12ed..52d1798 100644
--- a/workflow/service/flow/lambda_segment_resume.go
+++ b/workflow/service/flow/lambda_segment_resume.go
@@ -11,6 +11,25 @@ import (
 // segmentGenerateMaxAttempts 视频段生成最大尝试次数（失败自动重试 1 次，共 2 次尝试），参数化可调
 const segmentGenerateMaxAttempts = 2
 
+// hasDistinctSegmentIndex 判断各段参数是否都带 segment_index 且互不重复。
+// 段级续跑依赖 segment_index 作为段的稳定身份；缺失或重复（如视频节点误用批量拆分）时
+// 不得启用复用/落库，否则续跑会拿同一段结果拼出 N 段重复视频（静默损坏）。
+func hasDistinctSegmentIndex(paramsList []map[string]any) bool {
+	seen := make(map[int]bool, len(paramsList))
+	for _, params := range paramsList {
+		v, ok := params["segment_index"]
+		if !ok {
+			return false
+		}
+		idx := gconv.Int(v)
+		if seen[idx] {
+			return false
+		}
+		seen[idx] = true
+	}
+	return len(paramsList) > 0
+}
+
 // planSegmentResume 段级续跑决策：把 paramsList 各段映射到"是否需重新生成"。
 // savedMap 为该节点已成功段（段序号 → {key,url}）；段在表中缺失或地址为空则需重新生成。
 // 返回值与 paramsList 对齐。全新执行（savedMap 为 nil/空）时全部需生成。
@@ -52,4 +71,4 @@ func mergeSegmentOutputs(idxList []int, needGen []bool, newRes [][]map[string]an
 		merged = append(merged, o.recs...)
 	}
 	return merged
-}
\ No newline at end of file
+}
diff --git a/workflow/service/flow/lambda_segment_resume_test.go b/workflow/service/flow/lambda_segment_resume_test.go
index 15ffacb..a8df876 100644
--- a/workflow/service/flow/lambda_segment_resume_test.go
+++ b/workflow/service/flow/lambda_segment_resume_test.go
@@ -6,6 +6,28 @@ import (
 	"ai-agent/workflow/model/entity"
 )
 
+// hasDistinctSegmentIndex：各段都带 segment_index 且互不重复 → 允许段级续跑
+func TestHasDistinctSegmentIndex(t *testing.T) {
+	distinct := []map[string]any{{"segment_index": 1}, {"segment_index": 2}, {"segment_index": 3}}
+	if !hasDistinctSegmentIndex(distinct) {
+		t.Fatal("互不重复的 segment_index 应返回 true")
+	}
+	missing := []map[string]any{{"segment_index": 1}, {"other": 2}}
+	if hasDistinctSegmentIndex(missing) {
+		t.Fatal("缺失 segment_index 应返回 false")
+	}
+	duplicate := []map[string]any{{"segment_index": 1}, {"segment_index": 1}}
+	if hasDistinctSegmentIndex(duplicate) {
+		t.Fatal("重复 segment_index 应返回 false")
+	}
+	if hasDistinctSegmentIndex(nil) {
+		t.Fatal("空列表应返回 false")
+	}
+	if hasDistinctSegmentIndex([]map[string]any{}) {
+		t.Fatal("空列表(非 nil)应返回 false")
+	}
+}
+
 // planSegmentResume：段序号从 params 的 segment_index 读出；已成功段复用，缺失/空地址需生成
 func TestPlanSegmentResume(t *testing.T) {
 	paramsList := []map[string]any{