tesseract.js - npm Package Compare versions

Comparing version 2.0.0-alpha.4 to 2.0.0-alpha.6

docs/examples.md

@@ -59,3 +59,3 @@ # Tesseract.js Examples
 
-### with whitelist chars (^2.0.0-alpha.4)
+### with whitelist char (^2.0.0-alpha.5)
 
@@ -75,3 +75,3 @@ Sadly, whitelist chars is not supported in tesseract.js v4, so in tesseract.js we need to switch to tesseract v3 mode to make it work.
     {
-      'init_oem': OEM.TESSERACT_ONLY,
+      'tessedit_ocr_engine_mode': OEM.TESSERACT_ONLY,
       'tessedit_char_whitelist': '0123456789-.',
@@ -87,1 +87,51 @@ }
 ```
+
+### with different pageseg mode (^2.0.0-alpha.5)
+
+Check here for more details of pageseg mode: https://github.com/tesseract-ocr/tesseract/blob/4.0.0/src/ccstruct/publictypes.h#L163
+
+```javascript
+import Tesseract from 'tesseract.js';
+
+const { TesseractWorker, PSM } = Tesseract;
+const worker = new TesseractWorker();
+
+worker
+  .recognize(
+    'http://jeroen.github.io/images/testocr.png',
+    'eng',
+    {
+      'tessedit_pageseg_mode': PSM.SINGLE_BLOCK,
+    }
+  )
+  .progress((p) => {
+    console.log('progress', p);
+  })
+  .then((result) => {
+    console.log(result);
+  });
+```
+
+### with pdf output (^2.0.0-alpha.5)
+
+```javascript
+import Tesseract from 'tesseract.js';
+
+const { TesseractWorker } = Tesseract;
+const worker = new TesseractWorker();
+
+worker
+  .recognize(
+    'http://jeroen.github.io/images/testocr.png',
+    'eng',
+    {
+      'tessedit_create_pdf': '1',
+    }
+  )
+  .progress((p) => {
+    console.log('progress', p);
+  })
+  .then((result) => {
+    console.log(result);
+  });
+```

docs/local-installation.md

@@ -13,5 +13,5 @@ ## Local Installation
 const worker = Tesseract.TesseractWorker({
-  workerPath: 'https://unpkg.com/tesseract.js@v2.0.0-alpha.4/dist/worker.min.js',
+  workerPath: 'https://unpkg.com/tesseract.js@v2.0.0-alpha.6/dist/worker.min.js',
   langPath: 'https://tessdata.projectnaptha.com/4.0.0',
-  corePath: 'https://unpkg.com/tesseract.js-core@v2.0.0-beta.8/tesseract-core.wasm.js',
+  corePath: 'https://unpkg.com/tesseract.js-core@v2.0.0-beta.10/tesseract-core.wasm.js',
 });
@@ -18,0 +18,0 @@ ```

package.json

 {
   "name": "tesseract.js",
-  "version": "2.0.0-alpha.4",
+  "version": "2.0.0-alpha.6",
   "description": "Pure Javascript Multilingual OCR",
@@ -50,3 +50,3 @@ "main": "src/index.js",
     "resolve-url": "^0.2.1",
-    "tesseract.js-core": "^2.0.0-beta.8",
+    "tesseract.js-core": "^2.0.0-beta.10",
     "tesseract.js-utils": "^1.0.0-beta.5"
@@ -53,0 +53,0 @@ },

README.md

@@ -42,3 +42,3 @@ # [Tesseract.js](http://tesseract.projectnaptha.com/)
 ```html
-<script src='https://unpkg.com/tesseract.js@v2.0.0-alpha.4/dist/tesseract.min.js'></script>
+<script src='https://unpkg.com/tesseract.js@v2.0.0-alpha.6/dist/tesseract.min.js'></script>
 ```
@@ -45,0 +45,0 @@

src/browser/index.js

@@ -74,2 +74,21 @@ /**
 
+const downloadFile = (path, blob) => {
+  if (navigator.msSaveBlob) {
+    // IE 10+
+    navigator.msSaveBlob(blob, path);
+  } else {
+    const link = document.createElement('a');
+    // Browsers that support HTML5 download attribute
+    if (link.download !== undefined) {
+      const url = URL.createObjectURL(blob);
+      link.setAttribute('href', url);
+      link.setAttribute('download', path);
+      link.style.visibility = 'hidden';
+      document.body.appendChild(link);
+      link.click();
+      document.body.removeChild(link);
+    }
+  }
+}
+
 /*
@@ -87,3 +106,3 @@ * Default options for browser worker
    */
-  corePath: `https://unpkg.com/tesseract.js-core@v2.0.0-beta.8/tesseract-core.${typeof WebAssembly === 'object' ? 'wasm' : 'asm'}.js`,
+  corePath: `https://unpkg.com/tesseract.js-core@v2.0.0-beta.10/tesseract-core.${typeof WebAssembly === 'object' ? 'wasm' : 'asm'}.js`,
 };
@@ -113,3 +132,8 @@
   worker.onmessage = ({ data }) => {
-    instance.recv(data);
+    if (data.jobId.startsWith('Job')) {
+      instance.recv(data);
+    } else if (data.jobId.startsWith('Download')) {
+      const { path, blob } = data;
+      downloadFile(path, blob);
+    }
   };
@@ -116,0 +140,0 @@

src/browser/worker.js

@@ -45,2 +45,11 @@ /**
   },
+  b64toU8Array: s => new Uint8Array(atob(s).split('').map(c => c.charCodeAt(0))),
+  writeFile: (path, data, type) => {
+    const blob = new Blob([data], { type });
+    self.postMessage({
+      jobId: 'Download',
+      path,
+      blob,
+    });
+  },
 });

src/common/options.js

@@ -0,1 +1,3 @@
+const { OEM, PSM } = require('./types');
+
 module.exports = {
@@ -10,2 +12,14 @@ defaultOptions: {
   },
+  /*
+   * default params for recognize()
+   */
+  defaultParams: {
+    tessedit_ocr_engine_mode: OEM.TESSERACT_LSTM_COMBINED,
+    tessedit_pageseg_mode: PSM.SINGLE_BLOCK,
+    tessedit_char_whiltelist: '',
+    tessedit_create_pdf: '0',
+    textonly_pdf: '0',
+    pdf_name: 'tesseract.js-ocr-result',
+    pdf_title: 'Tesseract.js OCR Result',
+  },
 };

src/common/types.js

@@ -5,3 +5,3 @@ module.exports = {
    *
-   * By default tesseract.js uses DEFAULT mode, which uses LSTM when possible.
+   * By default tesseract.js uses TESSERACT_LSTM_COMBINED mode, which uses LSTM when possible.
    * If you need to use some tesseract v3 features (like tessedit_char_whitelist),
@@ -18,2 +18,21 @@ * you need to use TESSERACT_ONLY mode.
   },
+  /*
+   * PSM = Page Segmentation Mode
+   */
+  PSM: {
+    OSD_ONLY: '0',
+    AUTO_OSD: '1',
+    AUTO_ONLY: '2',
+    AUTO: '3',
+    SINGLE_COLUMN: '4',
+    SINGLE_BLOCK_VERT_TEXT: '5',
+    SINGLE_BLOCK: '6',
+    SINGLE_LINE: '7',
+    SINGLE_WORD: '8',
+    SINGLE_CHAR: '9',
+    SPARSE_TEXT: '10',
+    SPARSE_TEXT_OSD: '11',
+    RAW_LINE: '12',
+    COUNT: '13',
+  },
 };

src/common/workerUtils.js

@@ -12,3 +12,5 @@ /**
 const check = require('check-types');
+const pdfTTF = require('./pdf-ttf');
 const dump = require('./dump');
+const { defaultParams } = require('./options');
 
@@ -56,2 +58,54 @@ /*
 /**
+ * handleParams
+ *
+ * @name handleParams
+ * @function hanlde params from users
+ * @access private
+ * @param {string} lang - lang string for Init() 
+ * @param {object} customParams - an object of params
+ */
+const handleParams = (lang, customParams) => {
+  const {
+    tessedit_ocr_engine_mode,
+    ...params
+  } = {
+    ...defaultParams,
+    ...customParams,
+  };
+  api.Init(null, lang, tessedit_ocr_engine_mode);
+  Object.keys(params).forEach((key) => {
+    api.SetVariable(key, params[key]);
+  });
+};
+
+/**
+ * handleOutput
+ *
+ * @name handleOutput
+ * @function handle file output
+ * @access private
+ * @param {object} customParams - an object of params
+ */
+const handleOutput = (customParams) => {
+  const {
+    tessedit_create_pdf,
+    textonly_pdf,
+    pdf_name,
+    pdf_title,
+  } = {
+    ...defaultParams,
+    ...customParams,
+  };
+
+  if (tessedit_create_pdf === '1') {
+    const pdfRenderer = new TessModule.TessPDFRenderer(pdf_name, '/', textonly_pdf === '1');
+    pdfRenderer.BeginDocument(pdf_title);
+    pdfRenderer.AddImage(api);
+    pdfRenderer.EndDocument();
+    adapter.writeFile(`${pdf_name}.pdf`, TessModule.FS.readFile(`/${pdf_name}.pdf`), 'application/pdf');
+    TessModule._free(pdfRenderer);
+  }
+}
+
+/**
  * handleInit
@@ -80,2 +134,3 @@ *
         TessModule = tessModule;
+        TessModule.FS.writeFile('/pdf.ttf', adapter.b64toU8Array(pdfTTF));
         api = new TessModule.TessBaseAPI();
@@ -129,5 +184,2 @@ res.progress({ status: 'initialized tesseract', progress: 1 });
         .then(() => {
-          const OEM = check.undefined(params['init_oem'])
-            ? TessModule.OEM_DEFAULT
-            : params['init_oem'];
           const progressUpdate = (progress) => {
@@ -137,11 +189,8 @@ res.progress({ status: 'initializing api', progress });
           progressUpdate(0);
-          api.Init(null, lang, OEM);
-          progressUpdate(0.3);
-          Object.keys(params).filter(key => !key.startsWith('init_')).forEach((key) => {
-            api.SetVariable(key, params[key]);
-          });
-          progressUpdate(0.6);
+          handleParams(lang, params);
+          progressUpdate(0.5);
           const ptr = setImage(image);
           progressUpdate(1);
           api.Recognize(null);
+          handleOutput(params);
           const result = dump(TessModule, api);
@@ -148,0 +197,0 @@ api.End();

src/index.js

@@ -12,3 +12,3 @@ /**
 const TesseractWorker = require('./common/TesseractWorker');
-const { OEM } = require('./common/types');
+const types = require('./common/types');
 
@@ -21,3 +21,3 @@ module.exports = {
   /** Check ./common/types for more details */
-  OEM,
+  ...types,
 };

src/node/worker.js

@@ -36,2 +36,9 @@ /**
   },
+  b64toU8Array: s => Buffer.from(s, 'base64'),
+  writeFile: (path, data) => {
+    const fs = require('fs');
+    fs.writeFile(path, data, () => {
+      console.log('File Write Succeeded!');
+    });
+  },
 });

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Lines of code

2066

increased by6.88%

Worsened metrics

Total package byte prevSize

1195339

decreased by-11.26%

Number of package files

44

decreased by-13.73%

Number of low supply chain risk alerts

3

increased by50%

Dependency changes

• Updatedtesseract.js-core@^2.0.0-beta.10