690528:1524 ADR-030-230 context aware #02

backend/src/modules/ai/processors/ai-batch.processor.spec.ts

@@ -4,6 +4,8 @@
 // - 2026-05-21: เพิ่มการทดสอบ sandbox-extract พร้อม mock OcrService, OllamaService และ Redis (T039).
 // - 2026-05-21: แก้ไข ESLint unexpected any และ unsafe member access โดยกำหนด type ให้ redis เป็น Record<string, jest.Mock>
 // - 2026-05-22: เพิ่ม Mock dependencies (ProjectRepository, AiAuditLogRepository, TagsService, MigrationService) เพื่อแก้ปัญหา Nest resolve dependency ใน unit test และปรับโครงสร้างฟังก์ชันไม่มีบรรทัดว่าง (Zero Blank Lines) ตามกฎเหล็ก
+// - 2026-05-27: เพิ่ม Mock สำหรับ getActive และ resolveContext ของ AiPromptsService เพื่อรองรับ Context-Aware Prompt (T017)
+// - 2026-05-28: เพิ่ม test สำหรับ EC-001 (NEW_TAG_SUGGESTED) และ EC-002 (UNRESOLVED_SENDER/RECIPIENT_UUID)
 
 import { Test, TestingModule } from '@nestjs/testing';
 import { getRepositoryToken } from '@nestjs/typeorm';
@@ -86,12 +88,34 @@ describe('AiBatchProcessor', () => {
       .mockResolvedValue([
         { id: 5, publicId: 'tag-uuid-999', tagName: 'foundation' },
       ]),
+    findOrSuggestTags: jest.fn().mockResolvedValue([
+      {
+        tag: { id: 5, publicId: 'tag-uuid-999', tagName: 'foundation' },
+        isNew: false,
+      },
+    ]),
   };
   const mockMigrationService = {
     createError: jest.fn().mockResolvedValue(undefined),
     enqueueRecord: jest.fn().mockResolvedValue(undefined),
   };
   const mockAiPromptsService = {
+    getActive: jest.fn().mockResolvedValue({
+      id: 1,
+      promptType: 'ocr_extraction',
+      versionNumber: 2,
+      template:
+        'Resolved test prompt with OCR text {{ocr_text}} and context {{master_data_context}}',
+      isActive: true,
+      contextConfig: { filter: {} },
+    }),
+    resolveContext: jest.fn().mockResolvedValue({
+      availableProjects: [],
+      availableOrganizations: [],
+      availableDisciplines: [],
+      availableCorrespondenceTypes: [],
+      availableTags: [],
+    }),
     resolveActive: jest.fn().mockResolvedValue({
       resolvedPrompt: 'Resolved test prompt with OCR text',
       versionNumber: 2,
@@ -203,6 +227,114 @@ describe('AiBatchProcessor', () => {
       expect.stringContaining('completed')
     );
   });
+  it('EC-001: ควรบันทึก aiIssues เมื่อ AI สกัด Tag ใหม่ที่ไม่มีในระบบ', async () => {
+    mockTagsService.findOrSuggestTags.mockResolvedValueOnce([
+      {
+        tag: { id: 5, publicId: 'tag-uuid-999', tagName: 'foundation' },
+        isNew: false,
+      },
+      {
+        tag: { id: 99, publicId: 'tag-uuid-new', tagName: 'newlytag' },
+        isNew: true,
+      },
+    ]);
+    const mockManager = {
+      createQueryBuilder: jest.fn().mockReturnThis(),
+      select: jest.fn().mockReturnThis(),
+      from: jest.fn().mockReturnThis(),
+      where: jest.fn().mockReturnThis(),
+      getRawOne: jest.fn().mockResolvedValue({ id: 10 }),
+    };
+    jest
+      .spyOn(
+        mockAttachmentRepo as unknown as { manager: unknown },
+        'manager',
+        'get'
+      )
+      .mockReturnValue(mockManager);
+    mockProjectRepo.findOne.mockResolvedValue({
+      id: 2,
+      publicId: 'proj-uuid-456',
+    });
+    const job = {
+      id: 'job-ec001',
+      data: {
+        jobType: 'migrate-document',
+        documentPublicId: 'doc-uuid-123',
+        projectPublicId: 'proj-uuid-456',
+        payload: { documentNumber: 'LEGACY-EC001', title: 'EC001 Title' },
+        idempotencyKey: 'idem-ec001',
+        batchId: 'batch-ec001',
+      },
+    } as unknown as Job<AiBatchJobData>;
+    await processor.process(job);
+    expect(mockMigrationService.enqueueRecord).toHaveBeenCalledWith(
+      expect.objectContaining({
+        aiIssues: expect.arrayContaining([
+          expect.objectContaining({
+            type: 'NEW_TAG_SUGGESTED',
+            tagName: 'newlytag',
+          }),
+        ]),
+      })
+    );
+  });
+  it('EC-002: ควรตั้ง isValid=false และบันทึก aiIssues เมื่อ UUID ผู้ส่งไม่พบใน Master Data', async () => {
+    mockTagsService.findOrSuggestTags.mockResolvedValueOnce([]);
+    mockOllamaService.generate.mockResolvedValueOnce(
+      JSON.stringify({
+        documentNumber: 'LEGACY-EC002',
+        subject: 'EC002 Subject',
+        discipline: 'Civil',
+        category: 'Correspondence',
+        originatorOrganizationPublicId: 'unknown-org-uuid',
+        confidence: 0.95,
+        tags: [],
+        summary: 'summary',
+      })
+    );
+    const mockManager = {
+      createQueryBuilder: jest.fn().mockReturnThis(),
+      select: jest.fn().mockReturnThis(),
+      from: jest.fn().mockReturnThis(),
+      where: jest.fn().mockReturnThis(),
+      getRawOne: jest.fn().mockResolvedValue(null),
+    };
+    jest
+      .spyOn(
+        mockAttachmentRepo as unknown as { manager: unknown },
+        'manager',
+        'get'
+      )
+      .mockReturnValue(mockManager);
+    mockProjectRepo.findOne.mockResolvedValue({
+      id: 2,
+      publicId: 'proj-uuid-456',
+    });
+    const job = {
+      id: 'job-ec002',
+      data: {
+        jobType: 'migrate-document',
+        documentPublicId: 'doc-uuid-123',
+        projectPublicId: 'proj-uuid-456',
+        payload: { documentNumber: 'LEGACY-EC002', title: 'EC002 Title' },
+        idempotencyKey: 'idem-ec002',
+        batchId: 'batch-ec002',
+      },
+    } as unknown as Job<AiBatchJobData>;
+    await processor.process(job);
+    expect(mockMigrationService.enqueueRecord).toHaveBeenCalledWith(
+      expect.objectContaining({
+        isValid: false,
+        aiIssues: expect.arrayContaining([
+          expect.objectContaining({
+            type: 'UNRESOLVED_SENDER_UUID',
+            uuid: 'unknown-org-uuid',
+          }),
+        ]),
+      })
+    );
+  });
   it('ควรประมวลผล migrate-document โดยจำลอง OCR, AI และเรียก migrationService.enqueueRecord', async () => {
     const job = {
       id: 'job-migrate',
@@ -215,6 +347,9 @@ describe('AiBatchProcessor', () => {
           title: 'Legacy Title',
           senderOrgId: 1,
           receiverOrgId: 2,
+          contextOverride: {
+            contractPublicId: 'contract-uuid-789',
+          },
         },
         idempotencyKey: 'idem-migrate-123',
         batchId: 'batch-999',
@@ -228,16 +363,21 @@ describe('AiBatchProcessor', () => {
       pdfPath: '/files/test.pdf',
     });
     expect(ollamaService.generate).toHaveBeenCalledTimes(1);
-    expect(mockTagsService.findOrCreateTags).toHaveBeenCalledTimes(1);
+    expect(mockTagsService.findOrSuggestTags).toHaveBeenCalledTimes(1);
     expect(mockMigrationService.enqueueRecord).toHaveBeenCalledWith(
       expect.objectContaining({
-        documentNumber: 'LCBP3-CIV-001',
+        documentNumber: 'LEGACY-001',
         subject: 'Foundation Inspection Report',
         category: 'Correspondence',
         isValid: true,
         confidence: 0.95,
       })
     );
+    expect(mockAiPromptsService.resolveContext).toHaveBeenCalledWith(
+      expect.anything(),
+      'proj-uuid-456',
+      'contract-uuid-789'
+    );
     expect(mockAiAuditLogRepo.create).toHaveBeenCalledTimes(1);
     expect(mockAiAuditLogRepo.save).toHaveBeenCalledTimes(1);
   });

backend/src/modules/ai/processors/ai-batch.processor.ts

@@ -8,6 +8,7 @@
 // - 2026-05-22: แก้ไข type compilation error ใน processMigrateDocument และนำช่องว่างภายในฟังก์ชันออก
 // - 2026-05-25: เพิ่ม AiPromptsService เพื่อดึง Dynamic Prompt สำหรับ OCR extraction ใน sandbox และ migration pipeline
 // - 2026-05-26: แก้ไข bug lockDuration=30000ms ทำให้ sandbox-extract job stall เมื่อ Ollama ใช้เวลา >30s — เพิ่ม lockDuration: 150000
+// - 2026-05-28: EC-001 ใช้ findOrSuggestTags เพื่อตรวจจับ Tag ใหม่และบันทึก aiIssues; EC-002 ตรวจสอบ UUID ของผู้ส่ง/ผู้รับ และ Flag เมื่อหาไม่พบ
 
 import { Processor, WorkerHost } from '@nestjs/bullmq';
 import { Logger } from '@nestjs/common';
@@ -30,14 +31,16 @@ import { MigrationErrorType } from '../../migration/entities/migration-error.ent
 import { AiPromptsService } from '../prompts/ai-prompts.service';
 
 interface MigrateDocumentMetadata extends Record<string, unknown> {
-  documentNumber?: string;
+  projectPublicId?: string;
+  correspondenceTypeCode?: string;
+  disciplineCode?: string;
+  originatorOrganizationPublicId?: string;
+  recipients?: Array<{ organizationPublicId: string; recipientType: string }>;
   subject?: string;
-  category?: string;
-  discipline?: string;
-  date?: string;
-  confidence?: number;
+  documentDate?: string;
   tags?: string[];
   summary?: string;
+  confidence?: number;
 }
 
 export type AiBatchJobType =
@@ -72,6 +75,32 @@ const toStringList = (value: unknown): string[] =>
     ? value.filter((item): item is string => typeof item === 'string')
     : [];
 
+const toRecipientsList = (
+  value: unknown
+): Array<{ organizationPublicId: string; recipientType: string }> => {
+  if (!Array.isArray(value)) {
+    return [];
+  }
+  const result: Array<{ organizationPublicId: string; recipientType: string }> =
+    [];
+  for (const item of value) {
+    if (item && typeof item === 'object') {
+      const obj = item as Record<string, unknown>;
+      const orgId = readString(obj.organizationPublicId);
+      const type = readString(obj.recipientType);
+      if (orgId && type) {
+        // Normalize 'CC ' whitespace typo to 'CC'
+        const normalizedType = type.trim() === 'CC' ? 'CC' : type.trim();
+        result.push({
+          organizationPublicId: orgId,
+          recipientType: normalizedType,
+        });
+      }
+    }
+  }
+  return result;
+};
+
 const parseMigrateDocumentMetadata = (
   cleanedResponse: string
 ): MigrateDocumentMetadata => {
@@ -81,11 +110,15 @@ const parseMigrateDocumentMetadata = (
   }
   const source = parsed as Record<string, unknown>;
   return {
-    documentNumber: readString(source.documentNumber),
+    projectPublicId: readString(source.projectPublicId),
+    correspondenceTypeCode: readString(source.correspondenceTypeCode),
+    disciplineCode: readString(source.disciplineCode),
+    originatorOrganizationPublicId: readString(
+      source.originatorOrganizationPublicId
+    ),
+    recipients: toRecipientsList(source.recipients),
     subject: readString(source.subject),
-    category: readString(source.category),
-    discipline: readString(source.discipline),
-    date: readString(source.date),
+    documentDate: readString(source.documentDate),
     confidence:
       typeof source.confidence === 'number' ? source.confidence : undefined,
     tags: toStringList(source.tags),
@@ -246,8 +279,10 @@ export class AiBatchProcessor extends WorkerHost {
 
   /** ประมวลผล sandbox OCR + Metadata extraction โดยไม่บันทึกลง database */
   private async processSandboxExtract(data: AiBatchJobData): Promise<void> {
-    const { idempotencyKey, payload } = data;
+    const { idempotencyKey, payload, projectPublicId } = data;
     const pdfPath = payload.pdfPath as string;
+    const overrideProjPublicId =
+      (payload.projectPublicId as string) || projectPublicId;
     if (!pdfPath) {
       throw new Error('pdfPath is required for sandbox-extract job');
     }
@@ -261,11 +296,26 @@ export class AiBatchProcessor extends WorkerHost {
     );
     try {
       const ocrResult = await this.ocrService.detectAndExtract({ pdfPath });
-      const { resolvedPrompt, versionNumber } =
-        await this.aiPromptsService.resolveActive(
-          'ocr_extraction',
-          ocrResult.text
+
+      const activePrompt =
+        await this.aiPromptsService.getActive('ocr_extraction');
+      if (!activePrompt) {
+        throw new Error('No active ocr_extraction prompt version found');
+      }
+
+      // ดึงบริบท Master data
+      const masterDataContext = await this.aiPromptsService.resolveContext(
+        activePrompt,
+        overrideProjPublicId
+      );
+
+      const resolvedPrompt = activePrompt.template
+        .replace('{{ocr_text}}', ocrResult.text)
+        .replace(
+          '{{master_data_context}}',
+          JSON.stringify(masterDataContext, null, 2)
         );
+
       const response = await this.ollamaService.generate(resolvedPrompt, {
         timeoutMs: 120000,
       });
@@ -286,7 +336,7 @@ export class AiBatchProcessor extends WorkerHost {
       }
       await this.aiPromptsService.saveTestResult(
         'ocr_extraction',
-        versionNumber,
+        activePrompt.versionNumber,
         extractedMetadata
       );
       await this.redis.setex(
@@ -296,7 +346,7 @@ export class AiBatchProcessor extends WorkerHost {
           requestPublicId: idempotencyKey,
           status: 'completed',
           answer: JSON.stringify(extractedMetadata, null, 2),
-          promptVersionUsed: versionNumber,
+          promptVersionUsed: activePrompt.versionNumber,
           completedAt: new Date().toISOString(),
         })
       );
@@ -323,6 +373,13 @@ export class AiBatchProcessor extends WorkerHost {
     const startTime = Date.now();
     const { documentPublicId, projectPublicId, payload, batchId } = job.data;
     const docNumber = payload.documentNumber as string;
+    const contextOverride =
+      payload.contextOverride &&
+      typeof payload.contextOverride === 'object' &&
+      !Array.isArray(payload.contextOverride)
+        ? (payload.contextOverride as Record<string, unknown>)
+        : {};
+    const contractPublicId = readString(contextOverride.contractPublicId);
     const attachment = await this.attachmentRepo.findOne({
       where: { publicId: documentPublicId },
     });
@@ -358,10 +415,27 @@ export class AiBatchProcessor extends WorkerHost {
       });
       throw err;
     }
-    const { resolvedPrompt } = await this.aiPromptsService.resolveActive(
-      'ocr_extraction',
-      ocrResult.text
+
+    const activePrompt =
+      await this.aiPromptsService.getActive('ocr_extraction');
+    if (!activePrompt) {
+      throw new Error('No active prompt found for ocr_extraction');
+    }
+
+    // ดึงบริบทอ้างอิงโครงการที่กรองแล้ว (Data Isolation)
+    const masterDataContext = await this.aiPromptsService.resolveContext(
+      activePrompt,
+      projectPublicId,
+      contractPublicId
     );
+
+    const resolvedPrompt = activePrompt.template
+      .replace('{{ocr_text}}', ocrResult.text)
+      .replace(
+        '{{master_data_context}}',
+        JSON.stringify(masterDataContext, null, 2)
+      );
+
     let aiResponse: string;
     try {
       aiResponse = await this.ollamaService.generate(resolvedPrompt, {
@@ -411,50 +485,162 @@ export class AiBatchProcessor extends WorkerHost {
       });
       throw new Error(errMsg);
     }
+
+    // 3. ตรวจสอบและค้นหา Tags Suggestion ร่วมกับ Auto-Diff (EC-001)
+    const aiIssues: Record<string, unknown>[] = [];
     let mappedTags: Record<string, string>[] = [];
     if (extractedMetadata.tags && extractedMetadata.tags.length > 0) {
-      const tags = await this.tagsService.findOrCreateTags(
+      const tagResults = await this.tagsService.findOrSuggestTags(
         project.id,
         extractedMetadata.tags,
         attachment.uploadedByUserId
       );
-      mappedTags = tags.map((t) => ({
-        publicId: t.publicId,
-        tagName: t.tagName,
+      mappedTags = tagResults.map(({ tag }) => ({
+        publicId: tag.publicId,
+        tagName: tag.tagName,
       }));
+      // บันทึก Tag ใหม่ที่ไม่มีในระบบเป็น aiIssues เพื่อให้มนุษย์ตรวจสอบ
+      for (const { tag, isNew } of tagResults) {
+        if (isNew) {
+          aiIssues.push({
+            type: 'NEW_TAG_SUGGESTED',
+            tagPublicId: tag.publicId,
+            tagName: tag.tagName,
+            message: `Tag '${tag.tagName}' ถูกสร้างใหม่โดย AI — ต้องการการตรวจสอบจากมนุษย์`,
+          });
+        }
+      }
     }
     const confidence =
       typeof extractedMetadata.confidence === 'number'
         ? extractedMetadata.confidence
         : 0.5;
-    const isValid = confidence >= 0.6 && !!extractedMetadata.documentNumber;
+
+    // 4. Resolve UUIDs of Sender/Recipient Organizations to Database IDs (ADR-019)
+    // EC-002: UUID ที่หาไม่พบใน Master Data จะถูก flag ใน aiIssues และ isValid = false
+    let senderOrgId: number | undefined = undefined;
+    if (extractedMetadata.originatorOrganizationPublicId) {
+      const foundOrg = await this.attachmentRepo.manager
+        .createQueryBuilder()
+        .select('org.id', 'id')
+        .from('organizations', 'org')
+        .where('org.uuid = :uuid', {
+          uuid: extractedMetadata.originatorOrganizationPublicId,
+        })
+        .getRawOne<{ id: number }>();
+      if (foundOrg) {
+        senderOrgId = Number(foundOrg.id);
+      } else {
+        // EC-002: UUID ของผู้ส่งไม่มีใน Master Data — flag เพื่อ human review
+        aiIssues.push({
+          type: 'UNRESOLVED_SENDER_UUID',
+          uuid: extractedMetadata.originatorOrganizationPublicId,
+          message: `UUID ผู้ส่ง '${extractedMetadata.originatorOrganizationPublicId}' ไม่พบใน Master Data — ต้องการการตรวจสอบจากมนุษย์`,
+        });
+      }
+    }
+
+    let primaryReceiverOrgId: number | undefined = undefined;
+    if (
+      extractedMetadata.recipients &&
+      extractedMetadata.recipients.length > 0
+    ) {
+      // ดึงผู้รับที่เป็นประเภท TO รายแรกเป็นผู้รับหลัก (Primary Receiver)
+      const primaryReceiverObj =
+        extractedMetadata.recipients.find((r) => r.recipientType === 'TO') ||
+        extractedMetadata.recipients[0];
+      const foundOrg = await this.attachmentRepo.manager
+        .createQueryBuilder()
+        .select('org.id', 'id')
+        .from('organizations', 'org')
+        .where('org.uuid = :uuid', {
+          uuid: primaryReceiverObj.organizationPublicId,
+        })
+        .getRawOne<{ id: number }>();
+      if (foundOrg) {
+        primaryReceiverOrgId = Number(foundOrg.id);
+      } else {
+        // EC-002: UUID ของผู้รับไม่มีใน Master Data — flag เพื่อ human review
+        aiIssues.push({
+          type: 'UNRESOLVED_RECIPIENT_UUID',
+          uuid: primaryReceiverObj.organizationPublicId,
+          message: `UUID ผู้รับ '${primaryReceiverObj.organizationPublicId}' ไม่พบใน Master Data — ต้องการการตรวจสอบจากมนุษย์`,
+        });
+      }
+    }
+
+    // 5. ดึงประเภทเอกสารโต้ตอบ (Category Type) และสาขางาน (Discipline)
+    let matchedCategory = 'Correspondence';
+    if (extractedMetadata.correspondenceTypeCode) {
+      const foundType = await this.attachmentRepo.manager
+        .createQueryBuilder()
+        .select('t.type_name', 'name')
+        .from('correspondence_types', 't')
+        .where('t.type_code = :code', {
+          code: extractedMetadata.correspondenceTypeCode,
+        })
+        .getRawOne<{ name: string }>();
+      if (foundType) {
+        matchedCategory = foundType.name;
+      }
+    }
+
+    let matchedDisciplineId: number | undefined = undefined;
+    if (extractedMetadata.disciplineCode) {
+      const foundDisp = await this.attachmentRepo.manager
+        .createQueryBuilder()
+        .select('d.id', 'id')
+        .from('disciplines', 'd')
+        .where('d.discipline_code = :code', {
+          code: extractedMetadata.disciplineCode,
+        })
+        .getRawOne<{ id: number }>();
+      if (foundDisp) {
+        matchedDisciplineId = Number(foundDisp.id);
+      }
+    }
+
+    // 6. ส่งบันทึกเข้าสู่ Review Queue พร้อมคืนค่าผู้รับ Object Array ใน JSON metadata details
+    // EC-002: หากมี UUID ที่ไม่สามารถ resolve ได้ ให้ isValid = false เพื่อส่งเข้า review เสมอ
+    const hasUnresolvedUuids = aiIssues.some(
+      (issue) =>
+        issue.type === 'UNRESOLVED_SENDER_UUID' ||
+        issue.type === 'UNRESOLVED_RECIPIENT_UUID'
+    );
+    const isValid = confidence >= 0.6 && !!docNumber && !hasUnresolvedUuids;
     const payloadTitle = readString(payload.title);
+
     await this.migrationService.enqueueRecord({
-      documentNumber: extractedMetadata.documentNumber || docNumber,
+      documentNumber: docNumber,
       subject: extractedMetadata.subject || payloadTitle,
       originalSubject: payloadTitle,
       body: extractedMetadata.summary || '',
-      category: extractedMetadata.category || 'Correspondence',
+      category: matchedCategory,
       aiSummary: extractedMetadata.summary || '',
       projectId: project.id,
-      senderOrgId: readNumberId(payload.senderOrgId),
-      receiverOrgId: readNumberId(payload.receiverOrgId),
-      issuedDate: extractedMetadata.date || undefined,
-      receivedDate: extractedMetadata.date || undefined,
+      senderOrgId: senderOrgId || readNumberId(payload.senderOrgId),
+      receiverOrgId:
+        primaryReceiverOrgId || readNumberId(payload.receiverOrgId),
+      issuedDate: extractedMetadata.documentDate || undefined,
+      receivedDate: extractedMetadata.documentDate || undefined,
       extractedTags: mappedTags,
       tempAttachmentId: attachment.id,
       isValid,
       confidence,
       aiJobId: String(job.id),
+      aiIssues: aiIssues.length > 0 ? aiIssues : undefined,
       details: {
-        discipline: extractedMetadata.discipline,
+        disciplineCode: extractedMetadata.disciplineCode,
+        disciplineId: matchedDisciplineId,
+        recipientsList: extractedMetadata.recipients, // บันทึก Object Array สกัดใหม่
       },
     });
+
     await this.saveAiAuditLog({
       documentPublicId,
       aiModel: this.ollamaService.getMainModelName(),
       status: AiAuditStatus.SUCCESS,
-      aiSuggestionJson: extractedMetadata,
+      aiSuggestionJson: extractedMetadata as unknown as Record<string, unknown>,
       confidenceScore: confidence,
       processingTimeMs: Date.now() - startTime,
     });