scribe.js-ocr 0.4.0 → 0.4.1

package/docs/API.md

@@ -41,6 +41,7 @@ Initialize the program and optionally pre-load resources.
 
 Function for extracting text from image and PDF files with a single function call.
 By default, existing text content is extracted for text-native PDF files; otherwise text is extracted using OCR.
+To control how text from PDF files is handled, set the options in the `opt.usePDFText` object.
 For more control, use `init`, `importFiles`, `recognize`, and `exportData` separately.
 
 ### Parameters
@@ -48,10 +49,7 @@ For more control, use `init`, `importFiles`, `recognize`, and `exportData` separ
 *   `files`  
 *   `langs` **[Array][23]<[string][24]>**  (optional, default `['eng']`)
 *   `outputFormat`   (optional, default `'txt'`)
-*   `options` **[Object][21]?**  (optional, default `{}`)
-
-    *   `options.skipRecPDFTextNative` **[boolean][22]** If the input is a text-native PDF, skip recognition and return the existing text. (optional, default `true`)
-    *   `options.skipRecPDFTextOCR` **[boolean][22]** If the input is an image-native PDF with existing OCR layer, skip recognition and return the existing text. (optional, default `false`)
+*   `options`   (optional, default `{}`)
 
 ## writeDebugImages
 
@@ -115,12 +113,6 @@ Alternatively, for `File` objects (browser) and file paths (Node.js), a single a
 ### Parameters
 
 *   `files` **([Array][23]\<File> | FileList | [Array][23]<[string][24]> | [SortedInputFiles][13])**&#x20;
-*   `options` **[Object][21]?**  (optional, default `{}`)
-
-    *   `options.extractPDFTextNative` **[boolean][22]** Extract text from text-native PDF documents. (optional, default `false`)
-    *   `options.extractPDFTextOCR` **[boolean][22]** Extract text from image-native PDF documents with existing OCR text layers. (optional, default `false`)
-    *   `options.extractPDFTextImage` **[boolean][22]** Extract text from image-native PDF documents with no existing OCR layer.
-        This option exists because documents may still contain some text even if they are determined to be image-native (for example, scanned documents with a text-native header). (optional, default `false`)
 
 ## recognizePage
 

package/js/containers/app.js

@@ -62,9 +62,23 @@ export class opt {
 
   static calcSuppFontInfo = false;
 
-  static usePDFTextSupp = true;
-
-  static usePDFTextMain = true;
+  /**
+   * How to use PDF text data extracted from input PDFs (if any).
+   * The `native` option controls how native text data is used (i.e. visible text rendered by the PDF viewer),
+   * while the `ocr` option controls how OCR text data is used (i.e. invisible text printed over an image).
+   * If `main` is true, then the data will be used as the primary data source.
+   * If `supp` is true, then the data will be used as a supplemental data source (may be used to correct errors in the primary data source).
+   */
+  static usePDFText = {
+    native: {
+      supp: true,
+      main: true,
+    },
+    ocr: {
+      supp: true,
+      main: false,
+    },
+  };
 
   /**
    * Number of workers to use. Must be set prior to initialization.

package/js/extractPDFText.js

@@ -1,4 +1,4 @@
-import { inputData } from './containers/app.js';
+import { inputData, opt } from './containers/app.js';
 import { ocrAll, ocrAllRaw } from './containers/dataContainer.js';
 import { ImageCache } from './containers/imageContainer.js';
 import { convertOCR } from './recognizeConvert.js';
@@ -68,19 +68,11 @@ const extractInternalPDFTextRaw = async () => {
 
 /**
  * Extract and parse text from currently loaded PDF.
- * @param {Object} [options]
- * @param {boolean} [options.extractPDFTextNative=true] - Extract text from text-native PDF documents.
- * @param {boolean} [options.extractPDFTextOCR=false] - Extract text from image-native PDF documents with existing OCR text layers.
- * @param {boolean} [options.extractPDFTextImage=false] - Extract text from image-native PDF documents with no existing OCR layer.
- *   This option exists because documents may still contain some text even if they are determined to be image-native (for example, scanned documents with a text-native header).
- * @param {boolean} [options.setActive=false] - Set the active OCR data to the extracted text.
  */
-export const extractInternalPDFText = async (options = {}) => {
-  const extractPDFTextNative = options?.extractPDFTextNative ?? true;
-  const extractPDFTextOCR = options?.extractPDFTextOCR ?? false;
-  const extractPDFTextImage = options?.extractPDFTextImage ?? false;
-
-  const setActive = options?.setActive ?? false;
+export const extractInternalPDFText = async () => {
+  const extractPDFTextNative = opt.usePDFText.native.main || opt.usePDFText.native.supp;
+  const extractPDFTextOCR = opt.usePDFText.ocr.main || opt.usePDFText.ocr.supp;
+  const extractPDFTextImage = false;
 
   const res = await extractInternalPDFTextRaw();
 
@@ -95,7 +87,7 @@ export const extractInternalPDFText = async (options = {}) => {
 
   ocrAll.pdf = Array(ImageCache.pageCount);
 
-  if (setActive) {
+  if (inputData.pdfType === 'text' && opt.usePDFText.native.main || inputData.pdfType === 'ocr' && opt.usePDFText.ocr.main) {
     ocrAllRaw.active = ocrAllRaw.pdf;
     ocrAll.active = ocrAll.pdf;
   }

package/js/import/import.js

@@ -192,21 +192,11 @@ export function sortInputFiles(files) {
  * Alternatively, for `File` objects (browser) and file paths (Node.js), a single array can be provided, which is sorted based on extension.
  * @public
  * @param {Array<File>|FileList|Array<string>|SortedInputFiles} files
- * @param {Object} [options]
- * @param {boolean} [options.extractPDFTextNative=false] - Extract text from text-native PDF documents.
- * @param {boolean} [options.extractPDFTextOCR=false] - Extract text from image-native PDF documents with existing OCR text layers.
- * @param {boolean} [options.extractPDFTextImage=false] - Extract text from image-native PDF documents with no existing OCR layer.
- *   This option exists because documents may still contain some text even if they are determined to be image-native (for example, scanned documents with a text-native header).
- * @returns
  */
-export async function importFiles(files, options = {}) {
+export async function importFiles(files) {
   clearData();
   gs.getGeneralScheduler();
 
-  const extractPDFTextNative = options?.extractPDFTextNative ?? false;
-  const extractPDFTextOCR = options?.extractPDFTextOCR ?? false;
-  const extractPDFTextImage = options?.extractPDFTextImage ?? false;
-
   /** @type {Array<File|FileNode|ArrayBuffer>} */
   let pdfFiles = [];
   /** @type {Array<File|FileNode|ArrayBuffer>} */
@@ -445,11 +435,9 @@ export async function importFiles(files, options = {}) {
         await runFontOptimization(ocrAll.active);
       }
     });
-  } else if (inputData.pdfMode && (extractPDFTextNative || extractPDFTextOCR)) {
-    await extractInternalPDFText({
-      setActive: opt.usePDFTextMain, extractPDFTextNative, extractPDFTextOCR, extractPDFTextImage,
-    });
-    if (opt.usePDFTextMain) {
+  } else if (inputData.pdfMode && (opt.usePDFText.native.main || opt.usePDFText.native.supp || opt.usePDFText.ocr.main || opt.usePDFText.ocr.supp)) {
+    await extractInternalPDFText();
+    if (inputData.pdfType === 'text' && opt.usePDFText.native.main || inputData.pdfType === 'ocr' && opt.usePDFText.ocr.main) {
       if (inputData.pdfType === 'text') FontCont.enableCleanToNimbusMono = true;
       if (opt.calcSuppFontInfo) await calcSuppFontInfo(ocrAll.pdf);
     }

package/js/recognizeConvert.js

@@ -533,7 +533,7 @@ export async function recognize(options = {}) {
   let existingOCR;
   if (ocrAll['User Upload']) {
     existingOCR = ocrAll['User Upload'];
-  } else if (opt.usePDFTextSupp && ocrAll.pdf) {
+  } else if (ocrAll.pdf && (inputData.pdfType === 'text' && opt.usePDFText.native.supp || inputData.pdfType === 'ocr' && opt.usePDFText.ocr.supp)) {
     existingOCR = ocrAll.pdf;
     // If the PDF text is not the active data, it is assumed to be for supplemental purposes only.
     forceMainData = ocrAll.pdf !== ocrAll.active;

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "scribe.js-ocr",
-  "version": "0.4.0",
+  "version": "0.4.1",
   "description": "High-quality OCR and text extraction for images and PDFs.",
   "main": "scribe.js",
   "directories": {

package/scribe.js

@@ -79,20 +79,18 @@ const init = async (params) => {
 /**
  * Function for extracting text from image and PDF files with a single function call.
  * By default, existing text content is extracted for text-native PDF files; otherwise text is extracted using OCR.
+ * To control how text from PDF files is handled, set the options in the `opt.usePDFText` object.
  * For more control, use `init`, `importFiles`, `recognize`, and `exportData` separately.
  * @public
  * @param {Parameters<typeof importFiles>[0]} files
  * @param {Array<string>} [langs=['eng']]
  * @param {Parameters<typeof exportData>[0]} [outputFormat='txt']
- * @param {Object} [options]
- * @param {boolean} [options.skipRecPDFTextNative=true] - If the input is a text-native PDF, skip recognition and return the existing text.
- * @param {boolean} [options.skipRecPDFTextOCR=false] - If the input is an image-native PDF with existing OCR layer, skip recognition and return the existing text.
  */
 const extractText = async (files, langs = ['eng'], outputFormat = 'txt', options = {}) => {
   const skipRecPDFTextNative = options?.skipRecPDFTextNative ?? true;
   const skipRecPDFTextOCR = options?.skipRecPDFTextOCR ?? false;
   init({ ocr: true, font: true });
-  await importFiles(files, { extractPDFTextNative: skipRecPDFTextNative, extractPDFTextOCR: skipRecPDFTextOCR });
+  await importFiles(files);
   if (!inputData.xmlMode[0] && !inputData.imageMode && !inputData.pdfMode) throw new Error('No relevant files to process.');
   const skipRecPDF = inputData.pdfMode && (inputData.pdfType === 'text' && skipRecPDFTextNative || inputData.pdfType === 'ocr' && skipRecPDFTextOCR);
   const skipRecOCR = inputData.xmlMode[0] && !inputData.imageMode && !inputData.pdfMode;