@apart-tech/intelligence-core 1.12.1 → 1.13.0

package/dist/services/ocr-service.js

@@ -0,0 +1,61 @@
+/**
+ * OCR text extraction via the AIShield proxy (Mistral OCR model).
+ *
+ * Calls the OpenAI-compatible chat completions endpoint at AIShield,
+ * passing a signed GCS document URL as a `document_url` content part.
+ * The model returns extracted text as structured markdown.
+ */
+const OCR_PROMPT = `Extract all text content from this document. Format the output as clean markdown:
+- Preserve document structure with headings (#, ##, ###).
+- Render tables as markdown tables.
+- Separate pages with --- (horizontal rule).
+- Do not add any commentary — output only the extracted text.`;
+export class OcrService {
+    baseUrl;
+    apiKey;
+    model;
+    constructor(baseUrl, apiKey, model = "mistral-ocr-latest") {
+        this.baseUrl = baseUrl;
+        this.apiKey = apiKey;
+        this.model = model;
+    }
+    async extractText(documentUrl, organizationId) {
+        const body = {
+            model: this.model,
+            messages: [
+                {
+                    role: "user",
+                    content: [
+                        { type: "text", text: OCR_PROMPT },
+                        { type: "document_url", document_url: documentUrl },
+                    ],
+                },
+            ],
+        };
+        const response = await fetch(`${this.baseUrl}/v1/chat/completions`, {
+            method: "POST",
+            headers: {
+                "Content-Type": "application/json",
+                Authorization: `Bearer ${this.apiKey}`,
+                Organisation: organizationId,
+            },
+            body: JSON.stringify(body),
+        });
+        if (!response.ok) {
+            const errorBody = await response.text().catch(() => "");
+            throw new Error(`OCR request failed (${response.status}): ${errorBody}`);
+        }
+        const result = await response.json();
+        const text = result.choices?.[0]?.message?.content;
+        if (text === undefined || text === null) {
+            throw new Error("OCR response missing choices[0].message.content");
+        }
+        return {
+            text,
+            tokensIn: result.usage?.prompt_tokens ?? 0,
+            tokensOut: result.usage?.completion_tokens ?? 0,
+            model: result.model ?? this.model,
+        };
+    }
+}
+//# sourceMappingURL=ocr-service.js.map

package/dist/services/ocr-service.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"ocr-service.js","sourceRoot":"","sources":["../../src/services/ocr-service.ts"],"names":[],"mappings":"AAAA;;;;;;GAMG;AASH,MAAM,UAAU,GAAG;;;;8DAI2C,CAAC;AAE/D,MAAM,OAAO,UAAU;IAEX;IACA;IACA;IAHV,YACU,OAAe,EACf,MAAc,EACd,QAAgB,oBAAoB;QAFpC,YAAO,GAAP,OAAO,CAAQ;QACf,WAAM,GAAN,MAAM,CAAQ;QACd,UAAK,GAAL,KAAK,CAA+B;IAC3C,CAAC;IAEJ,KAAK,CAAC,WAAW,CACf,WAAmB,EACnB,cAAsB;QAEtB,MAAM,IAAI,GAAG;YACX,KAAK,EAAE,IAAI,CAAC,KAAK;YACjB,QAAQ,EAAE;gBACR;oBACE,IAAI,EAAE,MAAM;oBACZ,OAAO,EAAE;wBACP,EAAE,IAAI,EAAE,MAAM,EAAE,IAAI,EAAE,UAAU,EAAE;wBAClC,EAAE,IAAI,EAAE,cAAc,EAAE,YAAY,EAAE,WAAW,EAAE;qBACpD;iBACF;aACF;SACF,CAAC;QAEF,MAAM,QAAQ,GAAG,MAAM,KAAK,CAAC,GAAG,IAAI,CAAC,OAAO,sBAAsB,EAAE;YAClE,MAAM,EAAE,MAAM;YACd,OAAO,EAAE;gBACP,cAAc,EAAE,kBAAkB;gBAClC,aAAa,EAAE,UAAU,IAAI,CAAC,MAAM,EAAE;gBACtC,YAAY,EAAE,cAAc;aAC7B;YACD,IAAI,EAAE,IAAI,CAAC,SAAS,CAAC,IAAI,CAAC;SAC3B,CAAC,CAAC;QAEH,IAAI,CAAC,QAAQ,CAAC,EAAE,EAAE,CAAC;YACjB,MAAM,SAAS,GAAG,MAAM,QAAQ,CAAC,IAAI,EAAE,CAAC,KAAK,CAAC,GAAG,EAAE,CAAC,EAAE,CAAC,CAAC;YACxD,MAAM,IAAI,KAAK,CACb,uBAAuB,QAAQ,CAAC,MAAM,MAAM,SAAS,EAAE,CACxD,CAAC;QACJ,CAAC;QAED,MAAM,MAAM,GAAG,MAAM,QAAQ,CAAC,IAAI,EAIjC,CAAC;QAEF,MAAM,IAAI,GAAG,MAAM,CAAC,OAAO,EAAE,CAAC,CAAC,CAAC,EAAE,OAAO,EAAE,OAAO,CAAC;QACnD,IAAI,IAAI,KAAK,SAAS,IAAI,IAAI,KAAK,IAAI,EAAE,CAAC;YACxC,MAAM,IAAI,KAAK,CAAC,iDAAiD,CAAC,CAAC;QACrE,CAAC;QAED,OAAO;YACL,IAAI;YACJ,QAAQ,EAAE,MAAM,CAAC,KAAK,EAAE,aAAa,IAAI,CAAC;YAC1C,SAAS,EAAE,MAAM,CAAC,KAAK,EAAE,iBAAiB,IAAI,CAAC;YAC/C,KAAK,EAAE,MAAM,CAAC,KAAK,IAAI,IAAI,CAAC,KAAK;SAClC,CAAC;IACJ,CAAC;CACF"}

package/dist/services/ocr-service.test.d.ts

@@ -0,0 +1,2 @@
+export {};
+//# sourceMappingURL=ocr-service.test.d.ts.map

package/dist/services/ocr-service.test.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"ocr-service.test.d.ts","sourceRoot":"","sources":["../../src/services/ocr-service.test.ts"],"names":[],"mappings":""}

package/dist/services/ocr-service.test.js

@@ -0,0 +1,66 @@
+import { describe, expect, it, vi, afterEach, beforeEach } from "vitest";
+import { OcrService } from "./ocr-service.js";
+describe("OcrService", () => {
+    const BASE_URL = "https://aishield.apart.tech";
+    const API_KEY = "test-api-key";
+    const MODEL = "mistral-ocr-latest";
+    const ORG_ID = "org-123";
+    const DOC_URL = "https://storage.googleapis.com/bucket/doc.pdf?X-Goog-Signature=...";
+    let service;
+    const mockFetch = vi.fn();
+    beforeEach(() => {
+        service = new OcrService(BASE_URL, API_KEY, MODEL);
+        mockFetch.mockReset();
+        vi.stubGlobal("fetch", mockFetch);
+    });
+    afterEach(() => {
+        vi.restoreAllMocks();
+    });
+    it("sends correct request format and parses response", async () => {
+        mockFetch.mockResolvedValueOnce(new Response(JSON.stringify({
+            choices: [{ message: { content: "# Title\n\nExtracted text" } }],
+            usage: { prompt_tokens: 100, completion_tokens: 50 },
+            model: "mistral-ocr-latest",
+        }), { status: 200 }));
+        const result = await service.extractText(DOC_URL, ORG_ID);
+        expect(result.text).toBe("# Title\n\nExtracted text");
+        expect(result.tokensIn).toBe(100);
+        expect(result.tokensOut).toBe(50);
+        expect(result.model).toBe("mistral-ocr-latest");
+        expect(mockFetch).toHaveBeenCalledOnce();
+        const [url, opts] = mockFetch.mock.calls[0];
+        expect(url).toBe(`${BASE_URL}/v1/chat/completions`);
+        expect(opts.method).toBe("POST");
+        const headers = opts.headers;
+        expect(headers["Authorization"]).toBe(`Bearer ${API_KEY}`);
+        expect(headers["Organisation"]).toBe(ORG_ID);
+        expect(headers["Content-Type"]).toBe("application/json");
+        const body = JSON.parse(opts.body);
+        expect(body.model).toBe(MODEL);
+        expect(body.messages[0].content).toHaveLength(2);
+        expect(body.messages[0].content[0].type).toBe("text");
+        expect(body.messages[0].content[1].type).toBe("document_url");
+        expect(body.messages[0].content[1].document_url).toBe(DOC_URL);
+    });
+    it("throws on non-200 response", async () => {
+        mockFetch.mockResolvedValueOnce(new Response("Rate limited", { status: 429 }));
+        await expect(service.extractText(DOC_URL, ORG_ID)).rejects.toThrow("OCR request failed (429): Rate limited");
+    });
+    it("throws when response missing content", async () => {
+        mockFetch.mockResolvedValueOnce(new Response(JSON.stringify({ choices: [{ message: {} }], usage: {} }), { status: 200 }));
+        await expect(service.extractText(DOC_URL, ORG_ID)).rejects.toThrow("OCR response missing choices[0].message.content");
+    });
+    it("handles missing usage gracefully", async () => {
+        mockFetch.mockResolvedValueOnce(new Response(JSON.stringify({
+            choices: [{ message: { content: "text" } }],
+        }), { status: 200 }));
+        const result = await service.extractText(DOC_URL, ORG_ID);
+        expect(result.tokensIn).toBe(0);
+        expect(result.tokensOut).toBe(0);
+    });
+    it("handles network error", async () => {
+        mockFetch.mockRejectedValueOnce(new Error("Network error"));
+        await expect(service.extractText(DOC_URL, ORG_ID)).rejects.toThrow("Network error");
+    });
+});
+//# sourceMappingURL=ocr-service.test.js.map

package/dist/services/ocr-service.test.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"ocr-service.test.js","sourceRoot":"","sources":["../../src/services/ocr-service.test.ts"],"names":[],"mappings":"AAAA,OAAO,EAAE,QAAQ,EAAE,MAAM,EAAE,EAAE,EAAE,EAAE,EAAE,SAAS,EAAE,UAAU,EAAE,MAAM,QAAQ,CAAC;AACzE,OAAO,EAAE,UAAU,EAAE,MAAM,kBAAkB,CAAC;AAE9C,QAAQ,CAAC,YAAY,EAAE,GAAG,EAAE;IAC1B,MAAM,QAAQ,GAAG,6BAA6B,CAAC;IAC/C,MAAM,OAAO,GAAG,cAAc,CAAC;IAC/B,MAAM,KAAK,GAAG,oBAAoB,CAAC;IACnC,MAAM,MAAM,GAAG,SAAS,CAAC;IACzB,MAAM,OAAO,GAAG,oEAAoE,CAAC;IAErF,IAAI,OAAmB,CAAC;IACxB,MAAM,SAAS,GAAG,EAAE,CAAC,EAAE,EAAE,CAAC;IAE1B,UAAU,CAAC,GAAG,EAAE;QACd,OAAO,GAAG,IAAI,UAAU,CAAC,QAAQ,EAAE,OAAO,EAAE,KAAK,CAAC,CAAC;QACnD,SAAS,CAAC,SAAS,EAAE,CAAC;QACtB,EAAE,CAAC,UAAU,CAAC,OAAO,EAAE,SAAS,CAAC,CAAC;IACpC,CAAC,CAAC,CAAC;IAEH,SAAS,CAAC,GAAG,EAAE;QACb,EAAE,CAAC,eAAe,EAAE,CAAC;IACvB,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,kDAAkD,EAAE,KAAK,IAAI,EAAE;QAChE,SAAS,CAAC,qBAAqB,CAC7B,IAAI,QAAQ,CACV,IAAI,CAAC,SAAS,CAAC;YACb,OAAO,EAAE,CAAC,EAAE,OAAO,EAAE,EAAE,OAAO,EAAE,2BAA2B,EAAE,EAAE,CAAC;YAChE,KAAK,EAAE,EAAE,aAAa,EAAE,GAAG,EAAE,iBAAiB,EAAE,EAAE,EAAE;YACpD,KAAK,EAAE,oBAAoB;SAC5B,CAAC,EACF,EAAE,MAAM,EAAE,GAAG,EAAE,CAChB,CACF,CAAC;QAEF,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,WAAW,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;QAE1D,MAAM,CAAC,MAAM,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,2BAA2B,CAAC,CAAC;QACtD,MAAM,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,IAAI,CAAC,GAAG,CAAC,CAAC;QAClC,MAAM,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,EAAE,CAAC,CAAC;QAClC,MAAM,CAAC,MAAM,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,oBAAoB,CAAC,CAAC;QAEhD,MAAM,CAAC,SAAS,CAAC,CAAC,oBAAoB,EAAE,CAAC;QACzC,MAAM,CAAC,GAAG,EAAE,IAAI,CAAC,GAAG,SAAS,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,CAA0B,CAAC;QACrE,MAAM,CAAC,GAAG,CAAC,CAAC,IAAI,CAAC,GAAG,QAAQ,sBAAsB,CAAC,CAAC;QACpD,MAAM,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAEjC,MAAM,OAAO,GAAG,IAAI,CAAC,OAAiC,CAAC;QACvD,MAAM,CAAC,OAAO,CAAC,eAAe,CAAC,CAAC,CAAC,IAAI,CAAC,UAAU,OAAO,EAAE,CAAC,CAAC;QAC3D,MAAM,CAAC,OAAO,CAAC,cAAc,CAAC,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QAC7C,MAAM,CAAC,OAAO,CAAC,cAAc,CAAC,CAAC,CAAC,IAAI,CAAC,kBAAkB,CAAC,CAAC;QAEzD,MAAM,IAAI,GAAG,IAAI,CAAC,KAAK,CAAC,IAAI,CAAC,IAAc,CAAC,CAAC;QAC7C,MAAM,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC,IAAI,CAAC,KAAK,CAAC,CAAC;QAC/B,MAAM,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,YAAY,CAAC,CAAC,CAAC,CAAC;QACjD,MAAM,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,MAAM,CAAC,CAAC;QACtD,MAAM,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,IAAI,CAAC,CAAC,IAAI,CAAC,cAAc,CAAC,CAAC;QAC9D,MAAM,CAAC,IAAI,CAAC,QAAQ,CAAC,CAAC,CAAC,CAAC,OAAO,CAAC,CAAC,CAAC,CAAC,YAAY,CAAC,CAAC,IAAI,CAAC,OAAO,CAAC,CAAC;IACjE,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,4BAA4B,EAAE,KAAK,IAAI,EAAE;QAC1C,SAAS,CAAC,qBAAqB,CAC7B,IAAI,QAAQ,CAAC,cAAc,EAAE,EAAE,MAAM,EAAE,GAAG,EAAE,CAAC,CAC9C,CAAC;QAEF,MAAM,MAAM,CAAC,OAAO,CAAC,WAAW,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAChE,wCAAwC,CACzC,CAAC;IACJ,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,sCAAsC,EAAE,KAAK,IAAI,EAAE;QACpD,SAAS,CAAC,qBAAqB,CAC7B,IAAI,QAAQ,CACV,IAAI,CAAC,SAAS,CAAC,EAAE,OAAO,EAAE,CAAC,EAAE,OAAO,EAAE,EAAE,EAAE,CAAC,EAAE,KAAK,EAAE,EAAE,EAAE,CAAC,EACzD,EAAE,MAAM,EAAE,GAAG,EAAE,CAChB,CACF,CAAC;QAEF,MAAM,MAAM,CAAC,OAAO,CAAC,WAAW,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAChE,iDAAiD,CAClD,CAAC;IACJ,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,kCAAkC,EAAE,KAAK,IAAI,EAAE;QAChD,SAAS,CAAC,qBAAqB,CAC7B,IAAI,QAAQ,CACV,IAAI,CAAC,SAAS,CAAC;YACb,OAAO,EAAE,CAAC,EAAE,OAAO,EAAE,EAAE,OAAO,EAAE,MAAM,EAAE,EAAE,CAAC;SAC5C,CAAC,EACF,EAAE,MAAM,EAAE,GAAG,EAAE,CAChB,CACF,CAAC;QAEF,MAAM,MAAM,GAAG,MAAM,OAAO,CAAC,WAAW,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC;QAC1D,MAAM,CAAC,MAAM,CAAC,QAAQ,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;QAChC,MAAM,CAAC,MAAM,CAAC,SAAS,CAAC,CAAC,IAAI,CAAC,CAAC,CAAC,CAAC;IACnC,CAAC,CAAC,CAAC;IAEH,EAAE,CAAC,uBAAuB,EAAE,KAAK,IAAI,EAAE;QACrC,SAAS,CAAC,qBAAqB,CAAC,IAAI,KAAK,CAAC,eAAe,CAAC,CAAC,CAAC;QAE5D,MAAM,MAAM,CAAC,OAAO,CAAC,WAAW,CAAC,OAAO,EAAE,MAAM,CAAC,CAAC,CAAC,OAAO,CAAC,OAAO,CAChE,eAAe,CAChB,CAAC;IACJ,CAAC,CAAC,CAAC;AACL,CAAC,CAAC,CAAC"}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@apart-tech/intelligence-core",
-  "version": "1.12.1",
+  "version": "1.13.0",
   "description": "Core library: database, services, and providers for Apart Intelligence",
   "type": "module",
   "main": "dist/index.js",

package/prisma/schema.prisma

@@ -39,6 +39,7 @@ model Organization {
   nodeTags        NodeTag[]
   nodeChunks      NodeChunk[]
   piiQueryLogs    PiiQueryLog[]
+  documents       Document[]
 
   @@map("organizations")
 }
@@ -608,3 +609,31 @@ model NodeTagAudit {
   @@index([orgId, createdAt], map: "idx_tag_audit_org_time")
   @@map("node_tag_audit")
 }
+
+// ── Document Archive ────────────────────────────────────────────────────────
+
+model Document {
+  id              String    @id @default(dbgenerated("gen_random_uuid()")) @db.Uuid
+  organizationId  String    @map("organization_id") @db.Uuid
+  fileName        String    @map("file_name") @db.VarChar(500)
+  mimeType        String    @map("mime_type") @db.VarChar(100)
+  fileSizeBytes   Int       @map("file_size_bytes")
+  storagePath     String    @map("storage_path") @db.VarChar(1000)
+  contentHash     String?   @map("content_hash") @db.VarChar(64)
+  status          String    @default("uploading") @db.VarChar(20)
+  ocrModel        String?   @map("ocr_model") @db.VarChar(100)
+  ocrTokensIn     Int?      @map("ocr_tokens_in")
+  ocrTokensOut    Int?      @map("ocr_tokens_out")
+  pageCount       Int?      @map("page_count")
+  nodeId          String?   @map("node_id") @db.Uuid
+  errorMessage    String?   @map("error_message")
+  createdBy       String    @map("created_by") @db.VarChar(255)
+  createdAt       DateTime  @default(now()) @map("created_at") @db.Timestamptz
+  updatedAt       DateTime  @updatedAt @map("updated_at") @db.Timestamptz
+
+  organization    Organization @relation(fields: [organizationId], references: [id], onDelete: Cascade)
+
+  @@index([organizationId, contentHash], map: "idx_documents_org_hash")
+  @@index([organizationId, status], map: "idx_documents_org_status")
+  @@map("documents")
+}