mindee 4.14.1 → 4.16.0

package/CHANGELOG.md

@@ -1,5 +1,29 @@
 # CHANGELOG
 
+# v4.16.0 - 2024-09-17
+### Changes
+* :sparkles: add support for US Mail V2
+* :sparkles: add support for Bill of Lading V1
+* :sparkles: add support for FR Energy Bill V1
+* :sparkles: add support for FR Payslips V1
+* :sparkles: add support for Nutrition Facts Label V1
+* :recycle: refactor internals to accommodate for new changes
+
+### Fixes
+* :bug: fix float numbers not containing more 3 decimals in some instances
+* :bug: fix boolean field display issues
+* :bug: fix values being extracted as `undefined` instead of `null`, as is present in the initial prediction.
+
+
+## v4.15.0 - 2024-09-12
+### Changes
+* :sparkles: add support for full text OCR extraction
+* :wrench: add stricter linting rules
+* :recycle: add integration tests
+### Fixes
+* :bug: fix page operations not taking minimum threshold into account.
+
+
 ## v4.14.1 - 2024-09-03
 ### Changes
 * :memo: add classification detail to guide documentation

package/package.json

@@ -1,6 +1,6 @@
 {
     "name": "mindee",
-    "version": "4.14.1",
+    "version": "4.16.0",
     "description": "Mindee Client Library for Node.js",
     "main": "src/index.js",
     "bin": "bin/mindee.js",
@@ -10,6 +10,7 @@
         "build-for-dist": "tsc --build && cp LICENSE README.md CHANGELOG.md ./dist",
         "clean": "rm -rf ./dist ./docs/_build",
         "test": "mocha 'tests/**/*.spec.ts'",
+        "test-integration": "mocha 'tests/**/*.integration.ts'",
         "lint": "eslint  './src/**/*.ts' --report-unused-disable-directives && echo 'Your .ts files look good.'",
         "lint-fix": "eslint './src/**/*.ts' --fix",
         "docs": "typedoc --out docs/_build ./src/index.ts",

package/src/client.d.ts

@@ -2,7 +2,7 @@
 /// <reference types="node" />
 /// <reference types="node" />
 import { Readable } from "stream";
-import { Base64Input, BytesInput, InputSource, PathInput, StreamInput, PageOptions, UrlInput, BufferInput, LocalResponse } from "./input";
+import { Base64Input, BufferInput, BytesInput, InputSource, LocalResponse, PageOptions, PathInput, StreamInput, UrlInput } from "./input";
 import { Endpoint } from "./http";
 import { AsyncPredictResponse, FeedbackResponse, Inference, PredictResponse, StringDict } from "./parsing/common";
 /**
@@ -17,10 +17,12 @@ export interface PredictOptions {
      * This performs a full OCR operation on the server and will increase response time.
      */
     allWords?: boolean;
+    /**
+     * Whether to include the full ocr text. Only available on compatible APIs.
+     */
+    fullText?: boolean;
     /**
      * Whether to include cropper results for each page.
-     *
-     * This performs a cropping operation on the server and will increase response time.
      */
     cropper?: boolean;
     /**

package/src/client.js

@@ -52,6 +52,7 @@ class Client {
     async parse(productClass, inputSource, params = {
         endpoint: undefined,
         allWords: undefined,
+        fullText: undefined,
         cropper: undefined,
         pageOptions: undefined,
     }) {
@@ -62,6 +63,7 @@ class Client {
         const rawPrediction = await endpoint.predict({
             inputDoc: inputSource,
             includeWords: this.getBooleanParam(params.allWords),
+            fullText: this.getBooleanParam(params.fullText),
             pageOptions: params.pageOptions,
             cropper: this.getBooleanParam(params.cropper),
         });
@@ -83,6 +85,7 @@ class Client {
         const rawResponse = await endpoint.predictAsync({
             inputDoc: inputSource,
             includeWords: this.getBooleanParam(params.allWords),
+            fullText: this.getBooleanParam(params.fullText),
             pageOptions: params?.pageOptions,
             cropper: this.getBooleanParam(params.cropper),
         });
@@ -171,6 +174,7 @@ class Client {
     async enqueueAndParse(productClass, inputSource, asyncParams = {
         endpoint: undefined,
         allWords: undefined,
+        fullText: undefined,
         cropper: undefined,
         pageOptions: undefined,
         initialDelaySec: 4,
@@ -202,7 +206,9 @@ Job status: ${pollResults.job.status}.`);
             retryCounter++;
         }
         if (pollResults.job.status !== "completed") {
-            throw Error(`Asynchronous parsing request timed out after ${validatedAsyncParams.delaySec * retryCounter} seconds`);
+            throw Error("Asynchronous parsing request timed out after " +
+                validatedAsyncParams.delaySec * retryCounter +
+                " seconds");
         }
         return pollResults;
     }

package/src/http/endpoint.d.ts

@@ -6,6 +6,7 @@ import { StringDict } from "../parsing/common";
 export interface PredictParams {
     inputDoc: InputSource;
     includeWords: boolean;
+    fullText: boolean;
     pageOptions?: PageOptions;
     cropper: boolean;
 }
@@ -63,7 +64,8 @@ export declare class Endpoint extends BaseEndpoint {
      * @param input
      * @param predictUrl
      * @param includeWords
+     * @param fullText
      * @param cropper
      */
-    protected sendFileForPrediction(input: InputSource, predictUrl: string, includeWords?: boolean, cropper?: boolean): Promise<EndpointResponse>;
+    protected sendFileForPrediction(input: InputSource, predictUrl: string, includeWords?: boolean, fullText?: boolean, cropper?: boolean): Promise<EndpointResponse>;
 }

package/src/http/endpoint.js

@@ -40,7 +40,7 @@ class Endpoint extends baseEndpoint_1.BaseEndpoint {
         if (params.pageOptions !== undefined) {
             await __classPrivateFieldGet(this, _Endpoint_instances, "m", _Endpoint_cutDocPages).call(this, params.inputDoc, params.pageOptions);
         }
-        const response = await __classPrivateFieldGet(this, _Endpoint_instances, "m", _Endpoint_predictReqPost).call(this, params.inputDoc, params.includeWords, params.cropper);
+        const response = await __classPrivateFieldGet(this, _Endpoint_instances, "m", _Endpoint_predictReqPost).call(this, params.inputDoc, params.includeWords, params.fullText, params.cropper);
         if (!(0, responseValidation_1.isValidSyncResponse)(response)) {
             (0, error_1.handleError)(this.urlName, response, response.messageObj?.statusMessage);
         }
@@ -58,7 +58,7 @@ class Endpoint extends baseEndpoint_1.BaseEndpoint {
         if (params.pageOptions !== undefined) {
             await __classPrivateFieldGet(this, _Endpoint_instances, "m", _Endpoint_cutDocPages).call(this, params.inputDoc, params.pageOptions);
         }
-        const response = await __classPrivateFieldGet(this, _Endpoint_instances, "m", _Endpoint_predictAsyncReqPost).call(this, params.inputDoc, params.includeWords, params.cropper);
+        const response = await __classPrivateFieldGet(this, _Endpoint_instances, "m", _Endpoint_predictAsyncReqPost).call(this, params.inputDoc, params.includeWords, params.fullText, params.cropper);
         if (!(0, responseValidation_1.isValidAsyncResponse)(response)) {
             (0, error_1.handleError)(this.urlName, response, response.messageObj?.statusMessage);
         }
@@ -114,14 +114,18 @@ class Endpoint extends baseEndpoint_1.BaseEndpoint {
      * @param input
      * @param predictUrl
      * @param includeWords
+     * @param fullText
      * @param cropper
      */
-    sendFileForPrediction(input, predictUrl, includeWords = false, cropper = false) {
+    sendFileForPrediction(input, predictUrl, includeWords = false, fullText = false, cropper = false) {
         return new Promise((resolve, reject) => {
             const searchParams = new url_1.URLSearchParams();
             if (cropper) {
                 searchParams.append("cropper", "true");
             }
+            if (fullText) {
+                searchParams.append("full_text_ocr", "true");
+            }
             const form = new form_data_1.default();
             if (input instanceof base_1.LocalInputSource && input.fileObject instanceof Buffer) {
                 form.append("document", input.fileObject, {
@@ -136,7 +140,7 @@ class Endpoint extends baseEndpoint_1.BaseEndpoint {
             }
             const headers = { ...this.settings.baseHeaders, ...form.getHeaders() };
             let path = `${this.urlRoot}/${predictUrl}`;
-            if (searchParams.keys.length > 0) {
+            if (searchParams.toString().length > 0) {
                 path += `?${searchParams}`;
             }
             const options = {
@@ -163,10 +167,10 @@ async function _Endpoint_cutDocPages(inputDoc, pageOptions) {
     if (inputDoc instanceof base_1.LocalInputSource && inputDoc.isPdf()) {
         await inputDoc.cutPdf(pageOptions);
     }
-}, _Endpoint_predictReqPost = function _Endpoint_predictReqPost(input, includeWords = false, cropper = false) {
-    return this.sendFileForPrediction(input, "predict", includeWords, cropper);
-}, _Endpoint_predictAsyncReqPost = function _Endpoint_predictAsyncReqPost(input, includeWords = false, cropper = false) {
-    return this.sendFileForPrediction(input, "predict_async", includeWords, cropper);
+}, _Endpoint_predictReqPost = function _Endpoint_predictReqPost(input, includeWords = false, fullText = false, cropper = false) {
+    return this.sendFileForPrediction(input, "predict", includeWords, fullText, cropper);
+}, _Endpoint_predictAsyncReqPost = function _Endpoint_predictAsyncReqPost(input, includeWords = false, fullText = false, cropper = false) {
+    return this.sendFileForPrediction(input, "predict_async", includeWords, fullText, cropper);
 }, _Endpoint_documentQueueReqGet = function _Endpoint_documentQueueReqGet(queueId) {
     return new Promise((resolve, reject) => {
         const options = {

package/src/http/responseValidation.js

@@ -11,7 +11,9 @@ function isValidSyncResponse(response) {
     if (!response.messageObj || !response.messageObj.statusCode) {
         return false;
     }
-    return !(isNaN(response.messageObj.statusCode) || parseInt(response.messageObj.statusCode.toString()) < 200 || parseInt(response.messageObj.statusCode.toString()) > 302);
+    return !(isNaN(response.messageObj.statusCode) ||
+        parseInt(response.messageObj.statusCode.toString()) < 200 ||
+        parseInt(response.messageObj.statusCode.toString()) > 302);
 }
 exports.isValidSyncResponse = isValidSyncResponse;
 /**
@@ -37,7 +39,9 @@ function isValidAsyncResponse(response) {
     if (!response.data["job"]) {
         return false;
     }
-    return !(response.data["job"] && response.data["job"]["error"] && Object.keys(response.data["job"]["error"]).length > 0);
+    return !(response.data["job"] &&
+        response.data["job"]["error"] &&
+        Object.keys(response.data["job"]["error"]).length > 0);
 }
 exports.isValidAsyncResponse = isValidAsyncResponse;
 /**
@@ -54,10 +58,13 @@ function cleanRequestData(response) {
         return response;
     }
     if (response.data) {
-        if (response.data["api_request"] && response.data["api_request"]["status_code"] && parseInt(response.data["api_request"]["status_code"].toString()) > 399) {
+        if (response.data["api_request"] &&
+            response.data["api_request"]["status_code"] &&
+            parseInt(response.data["api_request"]["status_code"].toString()) > 399) {
             response.messageObj.statusCode = parseInt(response.data["api_request"]["status_code"].toString());
         }
-        if (response.data["job"] && response.data["job"]["error"] && Object.keys(response.data["job"]["error"]).length > 0) {
+        if (response.data["job"] && response.data["job"]["error"] &&
+            Object.keys(response.data["job"]["error"]).length > 0) {
             response.messageObj.statusCode = 500;
         }
     }

package/src/imageOperations/invoiceSplitterExtractor/invoiceSplitterExtractor.d.ts

@@ -5,7 +5,8 @@ import { ExtractedInvoiceSplitterImage } from "./extractedInvoiceSplitterImage";
  * Extracts & cuts the pages of a main document invoice according to the provided indexes.
  *
  * @param inputFile File to extract sub-invoices from.
- * @param indexes List of indexes to cut the document according to. Can be provided either as a InvoiceSplitterV1 inference, or a direct list of splits.
+ * @param indexes List of indexes to cut the document according to.
+ *  Can be provided either as a InvoiceSplitterV1 inference, or a direct list of splits.
  * @param strict If set to true, doesn't cut pages where the API isn't 100% confident.
  * @returns A promise of extracted images, as an array of ExtractedInvoiceSplitterImage.
  */

package/src/imageOperations/invoiceSplitterExtractor/invoiceSplitterExtractor.js

@@ -40,7 +40,8 @@ async function getPdfDoc(inputFile) {
  * Extracts & cuts the pages of a main document invoice according to the provided indexes.
  *
  * @param inputFile File to extract sub-invoices from.
- * @param indexes List of indexes to cut the document according to. Can be provided either as a InvoiceSplitterV1 inference, or a direct list of splits.
+ * @param indexes List of indexes to cut the document according to.
+ *  Can be provided either as a InvoiceSplitterV1 inference, or a direct list of splits.
  * @param strict If set to true, doesn't cut pages where the API isn't 100% confident.
  * @returns A promise of extracted images, as an array of ExtractedInvoiceSplitterImage.
  */

package/src/imageOperations/multiReceiptsExtractor/multiReceiptsExtractor.js

@@ -25,7 +25,9 @@ async function extractReceiptsFromPage(pdfPage, boundingBoxes, pageId) {
 async function loadPdfDoc(inputFile) {
     let pdfDoc;
     if (!["image/jpeg", "image/jpg", "image/png", "application/pdf"].includes(inputFile.mimeType)) {
-        throw new errors_1.MindeeMimeTypeError(`Unsupported file type "${inputFile.mimeType}" Currently supported types are .png, .jpg and .pdf`);
+        throw new errors_1.MindeeMimeTypeError('Unsupported file type "' +
+            inputFile.mimeType +
+            '" Currently supported types are .png, .jpg and .pdf');
     }
     else if (inputFile.isPdf()) {
         pdfDoc = await pdf_lib_1.PDFDocument.load(inputFile.fileObject);

package/src/input/localResponse.d.ts

@@ -3,6 +3,7 @@ import { StringDict } from "../parsing/common";
 import { Buffer } from "buffer";
 /**
  * Local response loaded from a file.
+ * Note: Has to be initialized through init() before use.
  */
 export declare class LocalResponse {
     private file;

package/src/input/localResponse.js

@@ -30,6 +30,7 @@ const errors_1 = require("../errors");
 const buffer_1 = require("buffer");
 /**
  * Local response loaded from a file.
+ * Note: Has to be initialized through init() before use.
  */
 class LocalResponse {
     /**

package/src/parsing/common/document.d.ts

@@ -29,4 +29,5 @@ export declare class Document<T extends Inference> {
      * Default string representation.
      */
     toString(): string;
+    private injectFullTextOcr;
 }

package/src/parsing/common/document.js

@@ -1,8 +1,8 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.Document = void 0;
-const cropperExtra_1 = require("./extras/cropperExtra");
-const extras_1 = require("./extras/extras");
+const extras_1 = require("./extras");
+const extras_2 = require("./extras/extras");
 const ocr_1 = require("./ocr");
 /**
  * Document prediction wrapper class. Holds the results of a parsed document.
@@ -19,18 +19,23 @@ class Document {
         this.filename = httpResponse?.name ?? "";
         this.ocr = httpResponse.ocr && Object.keys(httpResponse.ocr).length > 0 ? new ocr_1.Ocr(httpResponse.ocr) : undefined;
         this.inference = new inferenceClass(httpResponse["inference"]);
-        // Note: this is a convoluted but functional way of being able to implement/use Extras fields
-        // as an extension of a Map object (like having an adapted toString() method, for instance).
         if (httpResponse["extras"] &&
             Object.keys(httpResponse["extras"].length > 0)) {
             const extras = {};
             Object.entries(httpResponse["extras"]).forEach(([extraKey, extraValue]) => {
                 switch (extraKey) {
                     case "cropper":
-                        extras["cropper"] = new cropperExtra_1.CropperExtra(extraValue);
+                        extras["cropper"] = new extras_1.CropperExtra(extraValue);
+                        break;
+                    case "full_text_ocr":
+                        extras["fullTextOcr"] = new extras_1.FullTextOcrExtra(extraValue);
+                        break;
                 }
             });
-            this.extras = new extras_1.Extras(extras);
+            this.extras = new extras_2.Extras(extras);
+        }
+        if (!this.extras || !("fullTextOcr" in this.extras) || this.extras["full_text_ocr"].toString().length === 0) {
+            this.injectFullTextOcr(httpResponse);
         }
         this.nPages = httpResponse["n_pages"];
     }
@@ -44,5 +49,21 @@ class Document {
 
 ${this.inference?.toString()}`;
     }
+    injectFullTextOcr(rawPrediction) {
+        if (rawPrediction["inference"]["pages"].length < 1 ||
+            rawPrediction["inference"]["pages"][0]["extras"].length < 1 ||
+            !("full_text_ocr" in rawPrediction["inference"]["pages"][0]["extras"])) {
+            return;
+        }
+        const fullTextOcr = rawPrediction["inference"]["pages"].filter((e) => "extras" in e).map((e) => e["extras"]["full_text_ocr"]["content"]).join("\n");
+        // eslint-disable-next-line @typescript-eslint/naming-convention
+        const artificialTextObj = { "full_text_ocr": { "content": fullTextOcr.length > 0 ? fullTextOcr : "" } };
+        if (!this.extras) {
+            this.extras = new extras_2.Extras({ "fullTextOcr": new extras_1.FullTextOcrExtra(artificialTextObj) });
+        }
+        else {
+            this.extras["fullTextOcr"] = new extras_1.FullTextOcrExtra(artificialTextObj);
+        }
+    }
 }
 exports.Document = Document;

package/src/parsing/common/extras/fullTextOcrExtra.d.ts

@@ -0,0 +1,11 @@
+import { StringDict } from "../stringDict";
+import { ExtraField } from "./extras";
+export declare class FullTextOcrExtra extends ExtraField {
+    content?: string;
+    languages?: string;
+    constructor(rawPrediction: StringDict);
+    /**
+     * Default string representation.
+     */
+    toString(): string;
+}

package/src/parsing/common/extras/fullTextOcrExtra.js

@@ -0,0 +1,20 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.FullTextOcrExtra = void 0;
+const extras_1 = require("./extras");
+class FullTextOcrExtra extends extras_1.ExtraField {
+    constructor(rawPrediction) {
+        super();
+        if (rawPrediction["full_text_ocr"] !== undefined && rawPrediction["full_text_ocr"]["content"]) {
+            this.content = "content" in rawPrediction["full_text_ocr"] ? rawPrediction["full_text_ocr"]["content"] : "";
+            this.languages = "languages" in rawPrediction["full_text_ocr"] ? rawPrediction["full_text_ocr"]["languages"] : "";
+        }
+    }
+    /**
+     * Default string representation.
+     */
+    toString() {
+        return this.content !== undefined ? this.content : "";
+    }
+}
+exports.FullTextOcrExtra = FullTextOcrExtra;

package/src/parsing/common/extras/index.d.ts

@@ -0,0 +1,3 @@
+export { CropperExtra } from "./cropperExtra";
+export { Extras } from "./extras";
+export { FullTextOcrExtra } from "./fullTextOcrExtra";

package/src/parsing/common/extras/index.js

@@ -0,0 +1,9 @@
+"use strict";
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.FullTextOcrExtra = exports.Extras = exports.CropperExtra = void 0;
+var cropperExtra_1 = require("./cropperExtra");
+Object.defineProperty(exports, "CropperExtra", { enumerable: true, get: function () { return cropperExtra_1.CropperExtra; } });
+var extras_1 = require("./extras");
+Object.defineProperty(exports, "Extras", { enumerable: true, get: function () { return extras_1.Extras; } });
+var fullTextOcrExtra_1 = require("./fullTextOcrExtra");
+Object.defineProperty(exports, "FullTextOcrExtra", { enumerable: true, get: function () { return fullTextOcrExtra_1.FullTextOcrExtra; } });

package/src/parsing/common/index.d.ts

@@ -8,3 +8,5 @@ export { PredictResponse } from "./predictResponse";
 export { Prediction } from "./prediction";
 export { Page } from "./page";
 export { cleanOutString, lineSeparator } from "./summaryHelper";
+export * as extras from "./extras";
+export { floatToString, cleanSpecialChars } from "./summaryHelper";

package/src/parsing/common/index.js

@@ -1,6 +1,29 @@
 "use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || function (mod) {
+    if (mod && mod.__esModule) return mod;
+    var result = {};
+    if (mod != null) for (var k in mod) if (k !== "default" && Object.prototype.hasOwnProperty.call(mod, k)) __createBinding(result, mod, k);
+    __setModuleDefault(result, mod);
+    return result;
+};
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.lineSeparator = exports.cleanOutString = exports.Page = exports.Prediction = exports.PredictResponse = exports.AsyncPredictResponse = exports.Inference = exports.OrientationField = exports.FeedbackResponse = exports.Document = void 0;
+exports.cleanSpecialChars = exports.floatToString = exports.extras = exports.lineSeparator = exports.cleanOutString = exports.Page = exports.Prediction = exports.PredictResponse = exports.AsyncPredictResponse = exports.Inference = exports.OrientationField = exports.FeedbackResponse = exports.Document = void 0;
 var document_1 = require("./document");
 Object.defineProperty(exports, "Document", { enumerable: true, get: function () { return document_1.Document; } });
 var feedbackResponse_1 = require("./feedback/feedbackResponse");
@@ -20,3 +43,7 @@ Object.defineProperty(exports, "Page", { enumerable: true, get: function () { re
 var summaryHelper_1 = require("./summaryHelper");
 Object.defineProperty(exports, "cleanOutString", { enumerable: true, get: function () { return summaryHelper_1.cleanOutString; } });
 Object.defineProperty(exports, "lineSeparator", { enumerable: true, get: function () { return summaryHelper_1.lineSeparator; } });
+exports.extras = __importStar(require("./extras"));
+var summaryHelper_2 = require("./summaryHelper");
+Object.defineProperty(exports, "floatToString", { enumerable: true, get: function () { return summaryHelper_2.floatToString; } });
+Object.defineProperty(exports, "cleanSpecialChars", { enumerable: true, get: function () { return summaryHelper_2.cleanSpecialChars; } });

package/src/parsing/common/page.d.ts

@@ -27,4 +27,5 @@ export declare class Page<T extends Prediction> {
      * Default string representation.
      */
     toString(): string;
+    private injectFullTextOcr;
 }

package/src/parsing/common/page.js

@@ -1,8 +1,8 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
 exports.Page = void 0;
-const cropperExtra_1 = require("./extras/cropperExtra");
-const extras_1 = require("./extras/extras");
+const extras_1 = require("./extras");
+const extras_2 = require("./extras/extras");
 const orientation_1 = require("./orientation");
 /**
  * Page prediction wrapper class. Holds the results of a parsed document's page.
@@ -23,9 +23,6 @@ class Page {
                 pageId: pageId,
             });
         }
-        else {
-            orientation = undefined;
-        }
         this.id = pageId;
         this.prediction = new predictionType(rawPrediction["prediction"], pageId);
         if (rawPrediction["extras"] &&
@@ -34,10 +31,17 @@ class Page {
             Object.entries(rawPrediction["extras"]).forEach(([extraKey, extraValue]) => {
                 switch (extraKey) {
                     case "cropper":
-                        extras["cropper"] = new cropperExtra_1.CropperExtra(extraValue);
+                        extras["cropper"] = new extras_1.CropperExtra(extraValue);
+                        break;
+                    case "full_text_ocr":
+                        extras["fullTextOcr"] = new extras_1.FullTextOcrExtra(extraValue);
+                        break;
                 }
             });
-            this.extras = new extras_1.Extras(extras);
+            this.extras = new extras_2.Extras(extras);
+        }
+        if (!this.extras || !("fullTextOcr" in this.extras) || this.extras["fullTextOcr"].toString().length === 0) {
+            this.injectFullTextOcr(rawPrediction);
         }
     }
     /**
@@ -50,5 +54,24 @@ ${"-".repeat(title.length)}
 ${this.prediction.toString()}
 `;
     }
+    injectFullTextOcr(rawPrediction) {
+        if (!("extras" in rawPrediction) || !("full_text_ocr" in rawPrediction["extras"])) {
+            return;
+        }
+        const fullTextOcr = rawPrediction.map((e) => e["extras"]["full_text_ocr"]["content"]).join("\n");
+        const artificialTextObj = {
+            // eslint-disable-next-line @typescript-eslint/naming-convention
+            "full_text_ocr": {
+                "content": fullTextOcr.length > 0 ? fullTextOcr : "",
+            },
+        };
+        if (!this.extras) {
+            // eslint-disable-next-line @typescript-eslint/naming-convention
+            this.extras = new extras_2.Extras({ "full_text_ocr": new extras_1.FullTextOcrExtra(artificialTextObj) });
+        }
+        else {
+            this.extras["fullTextOcr"] = new extras_1.FullTextOcrExtra(artificialTextObj);
+        }
+    }
 }
 exports.Page = Page;

package/src/parsing/common/summaryHelper.d.ts

@@ -6,3 +6,13 @@ export declare function cleanOutString(outStr: string): string;
  * @returns A separator for table lines
  */
 export declare function lineSeparator(columnSizes: number[], separator: string): string;
+/**
+ * Replaces all special characters like \n, \r, \t, with an equivalent that can be displayed on a single line.
+ * Also trims line breaks at the end of the string.
+ * @param outStr
+ */
+export declare function cleanSpecialChars(outStr: string): string;
+/**
+ * Return a float as a string with at least 2 levels of precision.
+ */
+export declare function floatToString(value: number | null): string;

package/src/parsing/common/summaryHelper.js

@@ -1,6 +1,6 @@
 "use strict";
 Object.defineProperty(exports, "__esModule", { value: true });
-exports.lineSeparator = exports.cleanOutString = void 0;
+exports.floatToString = exports.cleanSpecialChars = exports.lineSeparator = exports.cleanOutString = void 0;
 function cleanOutString(outStr) {
     const lines = / \n/gm;
     return outStr.replace(lines, "\n");
@@ -20,3 +20,29 @@ function lineSeparator(columnSizes, separator) {
     return outStr;
 }
 exports.lineSeparator = lineSeparator;
+/**
+ * Replaces all special characters like \n, \r, \t, with an equivalent that can be displayed on a single line.
+ * Also trims line breaks at the end of the string.
+ * @param outStr
+ */
+function cleanSpecialChars(outStr) {
+    return outStr
+        .replace(/\n/g, "\\n")
+        .replace(/\r/g, "\\r")
+        .replace(/\t/g, "\\t");
+}
+exports.cleanSpecialChars = cleanSpecialChars;
+/**
+ * Return a float as a string with at least 2 levels of precision.
+ */
+function floatToString(value) {
+    if (value === null) {
+        return "";
+    }
+    return value.toLocaleString(undefined, {
+        minimumFractionDigits: 2,
+        maximumFractionDigits: 5,
+        useGrouping: false,
+    });
+}
+exports.floatToString = floatToString;

package/src/parsing/generated/generatedObject.d.ts

@@ -13,8 +13,8 @@ export declare class GeneratedObjectField {
     constructor({ prediction, pageId, }: BaseFieldConstructor);
     /**
      * ReSTructured-compliant string representation.
-       Takes into account level of indentation & displays elements as list elements.
-     * @param level Lvel of indentation. 0 by default.
+     Takes into account level of indentation & displays elements as list elements.
+     * @param level Level of indentation. 0 by default.
      */
     toStringLevel(level?: number): string;
     /**
@@ -26,7 +26,7 @@ export declare class GeneratedObjectField {
     toString(): string;
 }
 /**
- * Checks whether an field is a custom object or not.
+ * Checks whether a field is a custom object or not.
  * @param strDict input dictionary to check.
  */
 export declare function isGeneratedObject(strDict: StringDict): boolean;

package/src/parsing/generated/generatedObject.js

@@ -12,7 +12,13 @@ class GeneratedObjectField {
                 itemPageId = fieldValue;
             }
             else if (["polygon", "rectangle", "quadrangle", "bounding_box"].includes(fieldName)) {
-                Object.assign(this, { [fieldName]: new standard_1.PositionField({ prediction: { [fieldName]: fieldValue }, valueKey: fieldName, pageId: pageId }) });
+                Object.assign(this, {
+                    [fieldName]: new standard_1.PositionField({
+                        prediction: { [fieldName]: fieldValue },
+                        valueKey: fieldName,
+                        pageId: pageId,
+                    }),
+                });
                 this.printableValues.push(fieldName);
             }
             else if (fieldName === "confidence") {
@@ -26,7 +32,10 @@ class GeneratedObjectField {
                     Object.assign(this, { [fieldName]: this.toNumberString(fieldValue) });
                 }
                 else {
-                    Object.assign(this, { [fieldName]: (fieldValue !== undefined && fieldValue !== null) ? String(fieldValue) : null });
+                    Object.assign(this, {
+                        [fieldName]: (fieldValue !== undefined && fieldValue !== null) ?
+                            String(fieldValue) : null,
+                    });
                 }
                 this.printableValues.push(fieldName);
             }
@@ -35,8 +44,8 @@ class GeneratedObjectField {
     }
     /**
      * ReSTructured-compliant string representation.
-       Takes into account level of indentation & displays elements as list elements.
-     * @param level Lvel of indentation. 0 by default.
+     Takes into account level of indentation & displays elements as list elements.
+     * @param level Level of indentation. 0 by default.
      */
     toStringLevel(level = 0) {
         const indent = "  " + "  ".repeat(level);
@@ -64,7 +73,7 @@ class GeneratedObjectField {
 }
 exports.GeneratedObjectField = GeneratedObjectField;
 /**
- * Checks whether an field is a custom object or not.
+ * Checks whether a field is a custom object or not.
  * @param strDict input dictionary to check.
  */
 function isGeneratedObject(strDict) {