fix(server): get pending embedding docs in event handler (#13095)

maybe fix AI-309

packages/backend/server/src/models/copilot-workspace.ts

@@ -2,7 +2,7 @@ import { randomUUID } from 'node:crypto';
 
 import { Injectable } from '@nestjs/common';
 import { Transactional } from '@nestjs-cls/transactional';
-import { Prisma } from '@prisma/client';
+import { Prisma, PrismaClient } from '@prisma/client';
 
 import { PaginationInput } from '../base';
 import { BaseModel } from './base';
@@ -16,6 +16,10 @@ import type {
 
 @Injectable()
 export class CopilotWorkspaceConfigModel extends BaseModel {
+  constructor(private readonly database: PrismaClient) {
+    super();
+  }
+
   @Transactional()
   private async listIgnoredDocIds(
     workspaceId: string,
@@ -41,28 +45,25 @@ export class CopilotWorkspaceConfigModel extends BaseModel {
    * @param workspaceId id of the workspace
    * @returns docIds
    */
-  @Transactional()
   async findDocsToEmbed(workspaceId: string): Promise<string[]> {
-    const ignoredDocIds = (await this.listIgnoredDocIds(workspaceId)).map(
+    // NOTE: for unknown reason, the transaction will timeout if call from event handler
-      d => d.docId
+    // so we use an independent client here
-    );
+    const docIds = await this.database.$queryRaw<{ id: string }[]>`
+      SELECT s.guid as id
+        FROM snapshots AS s
+          LEFT JOIN ai_workspace_embeddings e
+            ON e.workspace_id = s.workspace_id
+               AND e.doc_id = s.guid
+          LEFT JOIN ai_workspace_ignored_docs id
+            ON id.workspace_id = s.workspace_id
+               AND id.doc_id = s.guid
+        WHERE s.workspace_id = ${workspaceId}
+          AND s.guid != s.workspace_id
+          AND s.guid NOT LIKE '%$%'
+          AND e.doc_id IS NULL
+          AND id.doc_id IS NULL;`;
 
-    const docIds = await this.db.snapshot
+    return docIds.map(r => r.id);
-      .findMany({
-        where: {
-          workspaceId,
-          AND: [
-            { id: { notIn: ignoredDocIds } },
-            { id: { not: workspaceId } },
-            { id: { not: { contains: '$' } } },
-          ],
-          embedding: { none: {} },
-        },
-        select: { id: true },
-      })
-      .then(r => r.map(doc => doc.id));
-
-    return docIds;
   }
 
   @Transactional()