mindee 5.0.0-rc2 → 5.0.0

package/CHANGELOG.md

@@ -1,19 +1,38 @@
 # CHANGELOG
 
-## 5.0.0-rc2 - 2026-01-18
+## v5.0.0 - 2026-02-19
+### ¡Breaking Changes!
+* :boom: :sparkles: move PDF and image libraries to optional dependencies (resolves #388)
+* :recycle: :boom: rework of directory structure and exports (resolves #276)
+* :boom: drop support for node.js 18
+* :boom: :recycle: separate polling options
+* :coffin: remove support for API builder (no longer maintained, use v2)
+* :coffin: remove support for v1 products available in v2
+* :coffin: remove support for unused products
+### Changes
+* :sparkles: add support for all utility products
+* :sparkles: add V2 CLI
+* :sparkles: add Job to inference responses
+* :sparkles: handle abort signal stream inputs
+* :recycle: use undici (compatible with v6 or v7) for HTTP requests and mocking
+* :sparkles: add ability for a user to specify their own undici `Dispatcher` instance
+* :recycle: migrate to ES modules
+
+
+## v5.0.0-rc2 - 2026-02-18
 ### Changes
 * :sparkles: add Job to inference responses
 ### Fixes
 * :bug: fix for bin scripts
 
 
-## 5.0.0-rc1 - 2026-01-17
+## v5.0.0-rc1 - 2026-02-17
 ### Changes
 * :boom: drop support for node.js 18
 * :boom: :recycle: separate polling options
 
 
-## 5.0.0-alpha2 - 2026-01-16
+## v5.0.0-alpha2 - 2026-02-16
 ### Changes
 * :recycle: use node test methods
 * :recycle: v1 ocr word should be a class
@@ -21,7 +40,7 @@
 * :bug: fix for optional node-poppler
 
 
-## 5.0.0-alpha1 - 2026-01-13
+## v5.0.0-alpha1 - 2026-02-13
 ### :boom: :recycle: base for v5
 * :arrow_up: update file-type
 * :recycle: migrate to ES modules (node18)

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "mindee",
-    "version": "5.0.0-rc2",
+    "version": "5.0.0",
     "description": "Mindee Client Library for Node.js",
     "author": {
         "name": "Mindee",
@@ -47,16 +47,16 @@
         "node": ">= 20.1"
     },
     "dependencies": {
-        "commander": "~9.4.1",
-        "file-type": "^19.6.0",
+        "commander": "^14.0.3",
+        "file-type": "^21.3.0",
         "tmp": "^0.2.3",
         "tslib": "^2.8.1",
-        "undici": "^6.23.0"
+        "undici": ">=6.23.0 <8.0.0"
     },
     "optionalDependencies": {
-        "@cantoo/pdf-lib": "^2.3.2",
-        "node-poppler": "^7.2.4",
-        "pdf.js-extract": "^0.2.1",
+        "@cantoo/pdf-lib": "^2.5.3",
+        "node-poppler": "^9.1.1",
+        "pdf.js-extract": "~0.2.1",
         "sharp": "~0.34.5"
     },
     "devDependencies": {

package/src/geometry/polygon.d.ts

@@ -23,4 +23,5 @@ export declare class Polygon extends Array<Point> {
      * Determine if a Point is within the Polygon's Y axis (same line).
      */
     isPointInY(point: Point): boolean;
+    toString(): string;
 }

package/src/geometry/polygon.js

@@ -36,4 +36,7 @@ export class Polygon extends Array {
         const yCoords = this.getMinMaxY();
         return isPointInY(point, yCoords.min, yCoords.max);
     }
+    toString() {
+        return this.map((point) => `(${point})`).join(", ");
+    }
 }

package/src/http/apiCore.d.ts

@@ -1,13 +1,13 @@
-import { Dispatcher } from "undici";
+import { Dispatcher, FormData } from "undici";
 import { InputSource, PageOptions } from "../input/index.js";
-export declare const TIMEOUT_DEFAULT: number;
+export declare const TIMEOUT_SECS_DEFAULT: number;
 export interface RequestOptions {
     hostname: string;
     path: string;
     method: any;
-    timeout: number;
+    timeoutSecs: number;
     headers: any;
-    body?: any;
+    body?: FormData | string;
 }
 export interface BaseHttpResponse {
     messageObj: any;

package/src/http/apiCore.js

@@ -1,7 +1,7 @@
 import { logger } from "../logger.js";
 import { request } from "undici";
 import { LocalInputSource } from "../input/index.js";
-export const TIMEOUT_DEFAULT = 120;
+export const TIMEOUT_SECS_DEFAULT = 120;
 /**
  * Cuts a document's pages according to the given options.
  * @param inputDoc input document.
@@ -24,10 +24,9 @@ export async function sendRequestAndReadResponse(dispatcher, options) {
     const response = await request(url, {
         method: options.method,
         headers: options.headers,
-        bodyTimeout: options.timeout,
+        headersTimeout: options.timeoutSecs * 1000,
         body: options.body,
-        throwOnError: false,
-        dispatcher: dispatcher
+        dispatcher: dispatcher,
     });
     logger.debug("Parsing the response ...");
     let responseBody = await response.body.text();

package/src/http/baseSettings.d.ts

@@ -6,7 +6,7 @@ export interface MindeeApiConstructorProps {
 export declare abstract class BaseSettings {
     apiKey: string;
     hostname: string;
-    timeout: number;
+    timeoutSecs: number;
     dispatcher: Dispatcher;
     protected constructor(apiKey?: string, dispatcher?: Dispatcher);
     protected getUserAgent(): string;

package/src/http/baseSettings.js

@@ -1,7 +1,7 @@
 import { getGlobalDispatcher } from "undici";
 import packageJson from "../../package.json" with { type: "json" };
 import * as os from "os";
-import { TIMEOUT_DEFAULT } from "./apiCore.js";
+import { TIMEOUT_SECS_DEFAULT } from "./apiCore.js";
 export class BaseSettings {
     constructor(apiKey, dispatcher) {
         if (apiKey === undefined || !apiKey || apiKey.length === 0) {
@@ -12,7 +12,9 @@ export class BaseSettings {
         }
         this.dispatcher = dispatcher ?? getGlobalDispatcher();
         this.hostname = this.hostnameFromEnv();
-        this.timeout = process.env.MINDEE_REQUEST_TIMEOUT ? parseInt(process.env.MINDEE_REQUEST_TIMEOUT) : TIMEOUT_DEFAULT;
+        this.timeoutSecs = process.env.MINDEE_REQUEST_TIMEOUT
+            ? parseInt(process.env.MINDEE_REQUEST_TIMEOUT)
+            : TIMEOUT_SECS_DEFAULT;
     }
     getUserAgent() {
         let platform = os.type().toLowerCase();

package/src/v1/http/endpoint.js

@@ -5,8 +5,9 @@ var __classPrivateFieldGet = (this && this.__classPrivateFieldGet) || function (
 };
 var _Endpoint_instances, _Endpoint_predictReqPost, _Endpoint_predictAsyncReqPost, _Endpoint_documentQueueReqGet, _Endpoint_documentGetReq, _Endpoint_documentFeedbackPutReq;
 import { URLSearchParams } from "url";
+import { FormData } from "undici";
 import { LocalInputSource } from "../../input/index.js";
-import { cutDocPages, sendRequestAndReadResponse } from "../../http/apiCore.js";
+import { cutDocPages, sendRequestAndReadResponse, } from "../../http/apiCore.js";
 import { handleError } from "./errors.js";
 import { isValidAsyncResponse, isValidSyncResponse } from "./responseValidation.js";
 /**
@@ -162,7 +163,7 @@ export class Endpoint {
             headers: this.settings.baseHeaders,
             hostname: this.settings.hostname,
             path: path,
-            timeout: this.settings.timeout,
+            timeoutSecs: this.settings.timeoutSecs,
             body: form,
         };
         return await sendRequestAndReadResponse(this.settings.dispatcher, options);
@@ -193,7 +194,7 @@ async function _Endpoint_documentQueueReqGet(queueId) {
         headers: this.settings.baseHeaders,
         hostname: this.settings.hostname,
         path: `${this.urlRoot}/documents/queue/${queueId}`,
-        timeout: this.settings.timeout,
+        timeoutSecs: this.settings.timeoutSecs,
     };
     return await sendRequestAndReadResponse(this.settings.dispatcher, options);
 }, _Endpoint_documentGetReq = 
@@ -207,7 +208,7 @@ async function _Endpoint_documentGetReq(documentId) {
         headers: this.settings.baseHeaders,
         hostname: this.settings.hostname,
         path: `${this.urlRoot}/documents/${documentId}`,
-        timeout: this.settings.timeout,
+        timeoutSecs: this.settings.timeoutSecs,
     };
     return await sendRequestAndReadResponse(this.settings.dispatcher, options);
 }, _Endpoint_documentFeedbackPutReq = 
@@ -223,7 +224,7 @@ async function _Endpoint_documentFeedbackPutReq(documentId, feedback) {
         hostname: this.settings.hostname,
         path: `/v1/documents/${documentId}/feedback`,
         body: JSON.stringify(feedback),
-        timeout: this.settings.timeout,
+        timeoutSecs: this.settings.timeoutSecs,
     };
     return await sendRequestAndReadResponse(this.settings.dispatcher, options);
 };

package/src/v1/http/workflowEndpoint.js

@@ -5,8 +5,9 @@ var __classPrivateFieldGet = (this && this.__classPrivateFieldGet) || function (
 };
 var _WorkflowEndpoint_instances, _WorkflowEndpoint_workflowReqPost;
 import { URLSearchParams } from "url";
+import { FormData } from "undici";
 import { LocalInputSource } from "../../input/index.js";
-import { cutDocPages, sendRequestAndReadResponse } from "../../http/apiCore.js";
+import { cutDocPages, sendRequestAndReadResponse, } from "../../http/apiCore.js";
 import { handleError } from "./errors.js";
 import { isValidSyncResponse } from "./responseValidation.js";
 /**
@@ -78,7 +79,7 @@ export class WorkflowEndpoint {
             headers: this.settings.baseHeaders,
             hostname: this.settings.hostname,
             path: path,
-            timeout: this.settings.timeout,
+            timeoutSecs: this.settings.timeoutSecs,
             body: form,
         };
         return await sendRequestAndReadResponse(this.settings.dispatcher, options);

package/src/v2/cli.js

@@ -16,13 +16,10 @@ function initClient(options) {
 async function enqueueAndGetInference(product, inputPath, options) {
     const mindeeClient = initClient(options);
     const inputSource = new PathInput({ inputPath: inputPath });
-    const response = await mindeeClient.enqueueAndGetResult(product, inputSource, {
-        modelId: options.model,
-        pollingOptions: {
-            initialDelaySec: 2,
-            delaySec: 1.5,
-            maxRetries: 80,
-        }
+    const response = await mindeeClient.enqueueAndGetResult(product, inputSource, { modelId: options.model }, {
+        initialDelaySec: 2,
+        delaySec: 1.5,
+        maxRetries: 80,
     });
     if (!response.inference) {
         throw Error("Inference could not be retrieved");
@@ -47,11 +44,11 @@ export function cli() {
         .option("-d, --debug", "high verbosity mode")
         .option("-k, --api-key <api_key>", "your Mindee API key");
     const inferenceTypes = [
-        { name: "extraction", description: "Extract data from a document.", product: Extraction },
-        { name: "crop", description: "Crop a document.", product: Crop },
-        { name: "split", description: "Split a document into pages.", product: Split },
-        { name: "ocr", description: "Read text from a document.", product: Ocr },
-        { name: "classification", description: "Classify a document.", product: Classification },
+        { name: "extraction", description: "Extract data from a document.", productClass: Extraction },
+        { name: "crop", description: "Crop a document.", productClass: Crop },
+        { name: "split", description: "Split a document into pages.", productClass: Split },
+        { name: "ocr", description: "Read text from a document.", productClass: Ocr },
+        { name: "classification", description: "Classify a document.", productClass: Classification },
     ];
     for (const inference of inferenceTypes) {
         const inferenceCmd = program.command(inference.name)
@@ -59,7 +56,7 @@ export function cli() {
         addMainOptions(inferenceCmd);
         inferenceCmd.action(function (inputPath, options) {
             const allOptions = { ...program.opts(), ...options };
-            return enqueueAndGetInference(inference.product, inputPath, allOptions);
+            return enqueueAndGetInference(inference.productClass, inputPath, allOptions);
         });
     }
     program.parse(process.argv);

package/src/v2/client/baseParameters.d.ts

@@ -1,3 +1,4 @@
+import { FormData } from "undici";
 /**
  * Constructor parameters for BaseParameters and its subclasses.
  */

package/src/v2/client/baseParameters.js

@@ -1,3 +1,4 @@
+import { FormData } from "undici";
 import { MindeeConfigurationError } from "../../errors/index.js";
 /**
  * Parameters accepted by all v2 products.

package/src/v2/http/mindeeApiV2.js

@@ -100,7 +100,7 @@ async function _MindeeApiV2_reqPostProductEnqueue(product, inputSource, params)
         hostname: this.settings.hostname,
         path: path,
         body: form,
-        timeout: this.settings.timeout,
+        timeoutSecs: this.settings.timeoutSecs,
     };
     return await sendRequestAndReadResponse(this.settings.dispatcher, options);
 }, _MindeeApiV2_reqGetJob = async function _MindeeApiV2_reqGetJob(jobId) {
@@ -109,7 +109,7 @@ async function _MindeeApiV2_reqPostProductEnqueue(product, inputSource, params)
         headers: this.settings.baseHeaders,
         hostname: this.settings.hostname,
         path: `/v2/jobs/${jobId}`,
-        timeout: this.settings.timeout,
+        timeoutSecs: this.settings.timeoutSecs,
     };
     return await sendRequestAndReadResponse(this.settings.dispatcher, options);
 }, _MindeeApiV2_reqGetProductResult = 
@@ -126,7 +126,7 @@ async function _MindeeApiV2_reqGetProductResult(inferenceId, slug) {
         headers: this.settings.baseHeaders,
         hostname: this.settings.hostname,
         path: `/v2/products/${slug}/results/${inferenceId}`,
-        timeout: this.settings.timeout,
+        timeoutSecs: this.settings.timeoutSecs,
     };
     return await sendRequestAndReadResponse(this.settings.dispatcher, options);
 };

package/src/v2/parsing/inference/baseInference.js

@@ -11,6 +11,7 @@ export class BaseInference {
     toString() {
         return ("Inference\n" +
             "#########\n" +
+            this.job.toString() + "\n" +
             this.model.toString() + "\n" +
             this.file.toString() + "\n");
     }

package/src/v2/product/extraction/params/extractionParameters.d.ts

@@ -1,3 +1,4 @@
+import { FormData } from "undici";
 import { StringDict } from "../../../../parsing/stringDict.js";
 import { DataSchema } from "./dataSchema.js";
 import { BaseParameters, BaseParametersConstructor } from "../../../../v2/client/baseParameters.js";

package/src/v2/product/extraction/params/extractionParameters.js

@@ -1,3 +1,4 @@
+import { FormData } from "undici";
 import { DataSchema } from "./dataSchema.js";
 import { BaseParameters } from "../../../../v2/client/baseParameters.js";
 import { logger } from "../../../../logger.js";