Fix langs TS typing (#570)

docs/image-format.md

@@ -1,17 +1,18 @@
 # Image Format
 
-Support Format: **bmp, jpg, png, pbm**
+The main Tesseract.js functions (ex. recognize, detect) take an `image` parameter.  The image formats and data types supported are listed below. 
 
-The main Tesseract.js functions (ex. recognize, detect) take an `image` parameter, which should be something that is like an image. What's considered "image-like" differs depending on whether it is being run from the browser or through NodeJS.
+Support Image Formats: **bmp, jpg, png, pbm, webp**
 
-On a browser, an image can be:
-- an `img` or `canvas` element
-- a `File` object (from a file `<input>`)
-- a `Blob` object
-- a path or URL to an accessible image
-- a base64 encoded image fits `data:image\/([a-zA-Z]*);base64,([^"]*)` regexp
+For browser and Node, supported data types are:
+ - string with base64 encoded image (fits `data:image\/([a-zA-Z]*);base64,([^"]*)` regexp)
+ - buffer
 
-In Node.js, an image can be
-- a path to a local image
-- a Buffer storing binary image 
-- a base64 encoded image fits `data:image\/([a-zA-Z]*);base64,([^"]*)` regexp
+For browser only, supported data types are:
+ - `File` or `Blob` object
+ - `img` or `canvas` element
+
+For Node only, supported data types are:
+ - string containing a path to local image
+
+Note: images must be a supported image format **and** a supported data type.  For example, a buffer containing a png image is supported.  A buffer containing raw pixel data is not supported. 

package-lock.json

@@ -1,12 +1,12 @@
 {
   "name": "tesseract.js",
-  "version": "3.0.2",
+  "version": "3.0.3",
   "lockfileVersion": 2,
   "requires": true,
   "packages": {
     "": {
       "name": "tesseract.js",
-      "version": "3.0.2",
+      "version": "3.0.3",
       "hasInstallScript": true,
       "license": "Apache-2.0",
       "dependencies": {
@@ -20,7 +20,7 @@
         "opencollective-postinstall": "^2.0.2",
         "regenerator-runtime": "^0.13.3",
         "resolve-url": "^0.2.1",
-        "tesseract.js-core": "^3.0.1",
+        "tesseract.js-core": "^3.0.2",
         "wasm-feature-detect": "^1.2.11",
         "zlibjs": "^0.3.1"
       },
@@ -8500,9 +8500,9 @@
       }
     },
     "node_modules/tesseract.js-core": {
-      "version": "3.0.1",
-      "resolved": "https://registry.npmjs.org/tesseract.js-core/-/tesseract.js-core-3.0.1.tgz",
-      "integrity": "sha512-kcEGcZG4Vl8tmdsPgLacPYoXFRo8IhG9SQxTLbK/6vG+aVTwD89voIJs1KgL3x+RcWDR7sNmd2sMVcpgcLBMMg=="
+      "version": "3.0.2",
+      "resolved": "https://registry.npmjs.org/tesseract.js-core/-/tesseract.js-core-3.0.2.tgz",
+      "integrity": "sha512-2fD76ka9nO/C616R0fq+M9Zu91DA3vEfyozp0jlxaJOBmpfeprtgRP3cqVweZh2darE1kK/DazoxZ65g7WU99Q=="
     },
     "node_modules/test-exclude": {
       "version": "6.0.0",
@@ -15989,9 +15989,9 @@
       }
     },
     "tesseract.js-core": {
-      "version": "3.0.1",
-      "resolved": "https://registry.npmjs.org/tesseract.js-core/-/tesseract.js-core-3.0.1.tgz",
-      "integrity": "sha512-kcEGcZG4Vl8tmdsPgLacPYoXFRo8IhG9SQxTLbK/6vG+aVTwD89voIJs1KgL3x+RcWDR7sNmd2sMVcpgcLBMMg=="
+      "version": "3.0.2",
+      "resolved": "https://registry.npmjs.org/tesseract.js-core/-/tesseract.js-core-3.0.2.tgz",
+      "integrity": "sha512-2fD76ka9nO/C616R0fq+M9Zu91DA3vEfyozp0jlxaJOBmpfeprtgRP3cqVweZh2darE1kK/DazoxZ65g7WU99Q=="
     },
     "test-exclude": {
       "version": "6.0.0",

package.json

@@ -1,6 +1,6 @@
 {
   "name": "tesseract.js",
-  "version": "3.0.2",
+  "version": "3.0.3",
   "description": "Pure Javascript Multilingual OCR",
   "main": "src/index.js",
   "types": "src/index.d.ts",
@@ -70,7 +70,7 @@
     "opencollective-postinstall": "^2.0.2",
     "regenerator-runtime": "^0.13.3",
     "resolve-url": "^0.2.1",
-    "tesseract.js-core": "^3.0.1",
+    "tesseract.js-core": "^3.0.2",
     "wasm-feature-detect": "^1.2.11",
     "zlibjs": "^0.3.1"
   },

src/constants/PSM.js

@@ -15,4 +15,5 @@ module.exports = {
   SINGLE_CHAR: '10',
   SPARSE_TEXT: '11',
   SPARSE_TEXT_OSD: '12',
+  RAW_LINE: '13',
 };

src/index.d.ts

@@ -19,12 +19,18 @@ declare namespace Tesseract {
     readText(path: string, jobId?: string): Promise<ConfigResult>
     removeText(path: string, jobId?: string): Promise<ConfigResult>
     FS(method: string, args: any[], jobId?: string): Promise<ConfigResult>
-    loadLanguage(langs?: string, jobId?: string): Promise<ConfigResult>
-    initialize(langs?: string, oem?: OEM, jobId?: string): Promise<ConfigResult>
+    loadLanguage(langs?: string | Lang[], jobId?: string): Promise<ConfigResult>
+    initialize(langs?: string | Lang[], oem?: OEM, jobId?: string): Promise<ConfigResult>
     setParameters(params: Partial<WorkerParams>, jobId?: string): Promise<ConfigResult>
     recognize(image: ImageLike, options?: Partial<RecognizeOptions>, jobId?: string): Promise<RecognizeResult>
     detect(image: ImageLike, jobId?: string): Promise<DetectResult>
     terminate(jobId?: string): Promise<ConfigResult>
+    getPDF(title?: string, textonly?: boolean, jobId?: string):Promise<GetPDFResult>
+  }
+
+  interface Lang {
+    code: string;
+    data: unknown;
   }
 
   interface WorkerOptions {
@@ -62,6 +68,10 @@ declare namespace Tesseract {
     jobId: string
     data: Page
   }
+  interface GetPDFResult {
+    jobId: string
+    data: number[]
+  }
   interface DetectResult {
     jobId: string
     data: DetectData
@@ -79,13 +89,13 @@ declare namespace Tesseract {
     width: number
     height: number
   }
-  const enum OEM {
+  enum OEM {
     TESSERACT_ONLY,
     LSTM_ONLY,
     TESSERACT_LSTM_COMBINED,
     DEFAULT,
   }
-  const enum PSM {
+  enum PSM {
     OSD_ONLY = '0',
     AUTO_OSD = '1',
     AUTO_ONLY = '2',
@@ -99,6 +109,7 @@ declare namespace Tesseract {
     SINGLE_CHAR = '10',
     SPARSE_TEXT = '11',
     SPARSE_TEXT_OSD = '12',
+    RAW_LINE = '13'
   }
   type ImageLike = string | HTMLImageElement | HTMLCanvasElement | HTMLVideoElement
     | CanvasRenderingContext2D | File | Blob | ImageData | Buffer;

src/worker-script/index.js

@@ -144,16 +144,8 @@ res) => {
     res.progress({ workerId, status: 'loaded language traineddata', progress: 1 });
     res.resolve(langs);
   } catch (err) {
-    if (isWebWorker && err instanceof DOMException) {
-      /*
-       * For some reason google chrome throw DOMException in loadLang,
-       * while other browser is OK, for now we ignore this exception
-       * and hopefully to find the root cause one day.
-       */
-    } else {
     res.reject(err.toString());
   }
-  }
 };
 
 const setParameters = ({ payload: { params: _params } }, res) => {
@@ -185,7 +177,10 @@ const initialize = ({
       api.End();
     }
     api = new TessModule.TessBaseAPI();
-    api.Init(null, langs, oem);
+    const status = api.Init(null, langs, oem);
+    if (status === -1) {
+      res.reject('initialization failed');
+    }
     params = defaultParams;
     setParameters({ payload: { params } });
     res.progress({