tesseract.js-utils - npm Package Compare versions

Comparing version 1.0.0-beta.1 to 1.0.0-beta.2

package.json

 {
   "name": "tesseract.js-utils",
-  "version": "1.0.0-beta.1",
+  "version": "1.0.0-beta.2",
   "description": "Utilities for tesseract.js",
@@ -45,3 +45,3 @@ "main": "src/index.node.js",
     "mocha": "^5.2.0",
-    "tesseract.js-core": "^2.0.0-beta.4",
+    "tesseract.js-core": "^2.0.0-beta.5",
     "testcafe": "^0.23.2",
@@ -52,2 +52,4 @@ "webpack": "^4.25.1",
   "dependencies": {
+    "bmp-js": "^0.1.0",
+    "file-type": "^10.5.0",
     "idb-keyval": "^3.1.0",
@@ -54,0 +56,0 @@ "node-fetch": "^2.3.0",

src/common/browser/cache.js

@@ -1,2 +0,2 @@
-const { set, get } = require('idb-keyval');
+const { set, get, del } = require('idb-keyval');
 
@@ -6,2 +6,7 @@ module.exports = {
   writeCache: set,
+  deleteCache: del,
+  checkCache: path => (
+    get(path)
+      .then(v => typeof v !== 'undefined')
+  ),
 };

src/common/node/cache.js

 const util = require('util');
 const fs = require('fs');
 
-const readFile = util.promisify(fs.readFile);
-const writeFile = util.promisify(fs.writeFile);
-
 module.exports = {
-  readCache: uri => readFile(uri),
-  writeCache: (uri, data) => writeFile(uri, data),
+  readCache: util.promisify(fs.readFile),
+  writeCache: util.promisify(fs.writeFile),
+  deleteCache: path => (
+    util.promisify(fs.unlink)(path)
+      .catch(() => {})
+  ),
+  checkCache: path => (
+    util.promisify(fs.access)(path, fs.F_OK)
+      .then(err => (err === null))
+      .catch(() => false)
+  ),
 };

src/index.browser.js

@@ -0,1 +1,3 @@
+const cache = require('./common/browser/cache');
+
 module.exports = {
@@ -5,5 +7,6 @@ loadLang: require('./loadLang')({
     gunzip: require('./common/browser/gunzip'),
-    ...require('./common/browser/cache'),
+    ...cache,
   }),
   readImage: require('./readImage'),
+  cache,
 };

src/index.node.js

@@ -0,1 +1,3 @@
+const cache = require('./common/node/cache');
+
 module.exports = {
@@ -5,5 +7,6 @@ loadLang: require('./loadLang')({
     gunzip: require('./common/node/gunzip'),
-    ...require('./common/node/cache'),
+    ...cache,
   }),
   readImage: require('./readImage'),
+  cache,
 };

src/loadLang.js

@@ -5,3 +5,3 @@ const handleLang = modules => ({
   cachePath,
-  cache,
+  cacheMethod,
   lang,
@@ -14,3 +14,3 @@ }) => (data) => {
       } catch (err) {
-        // Do nothing
+        // TODO: Do some error handling here.
       }
@@ -20,5 +20,7 @@ }
   }
-  if (cache) {
-    modules.writeCache(`${cachePath || '.'}/${lang}.traineddata`, data);
+  if (['write', 'refresh', undefined].includes(cacheMethod)) {
+    return modules.writeCache(`${cachePath || '.'}/${lang}.traineddata`, data)
+      .then(() => data);
   }
+
   return data;
@@ -30,5 +32,11 @@ };
   cachePath,
+  cacheMethod,
   ...options
-}) => lang => (
-  modules.readCache(`${cachePath || '.'}/${lang}.traineddata`)
+}) => (lang) => {
+  let { readCache } = modules;
+  if (['refresh', 'none'].includes(cacheMethod)) {
+    readCache = () => Promise.resolve();
+  }
+
+  return readCache(`${cachePath || '.'}/${lang}.traineddata`)
     .then((data) => {
@@ -38,3 +46,5 @@ if (typeof data === 'undefined') {
       }
-      return handleLang(modules)({ cachePath, lang, ...options })(data);
+      return handleLang(modules)({
+        cachePath, cacheMethod, lang, ...options,
+      })(data);
     })
@@ -46,5 +56,7 @@ .catch(() => (
         .then(buf => modules.gunzip(new Uint8Array(buf)))
-        .then(handleLang(modules)({ cachePath, lang, ...options }))
-    ))
-);
+        .then(handleLang(modules)({
+          cachePath, cacheMethod, lang, ...options,
+        }))
+    ));
+};
 
@@ -66,3 +78,7 @@ /**
  * @param {string} dataPath - path to store data in mem
- * @param {boolean} cache - true for caching
+ * @param {string} cacheMethod - method of cache invaliation, should one of following options:
+ *  write: read cache and write back (default option)
+ *  readOnly: read cache and not to write back (if cache does not exist, it is the same as none)
+ *  refresh: not to read cache and write back
+ *  none: not to read cache and not to write back
  *
@@ -69,0 +85,0 @@ */

src/readImage.js

@@ -1,29 +0,49 @@
-module.exports = (TessModule, buf) => {
-  const ptr = TessModule._malloc(buf.length * Uint8Array.BYTES_PER_ELEMENT);
-  TessModule.HEAPU8.set(buf, ptr);
-  const pix = TessModule._pixReadMem(ptr, buf.length);
-  if (TessModule.getValue(pix + (7 * 4), 'i32') === 0) {
-    /*
-     * Set a yres default value to prevent warning from tesseract
-     * See kMinCredibleResolution in tesseract/src/ccstruct/publictypes.h
-     */
-    TessModule.setValue(pix + (7 * 4), 70, 'i32');
+const bmp = require('bmp-js');
+const fileType = require('file-type');
+
+module.exports = (TessModule, iBuf) => {
+  const buf = Buffer.from(iBuf);
+  const type = fileType(buf);
+  let bytesPerPixel = 0;
+  let data = null;
+  let pix = null;
+  let w = 0;
+  let h = 0;
+
+  /*
+   * Although leptonica should support reading bmp, there is a bug of "compressed BMP files".
+   * As there is no solution, we need to use bmp-js for now.
+   * @see https://groups.google.com/forum/#!topic/tesseract-ocr/4mPD9zTxdxE
+   */
+  if (type && type.mime === 'image/bmp') {
+    const bmpBuf = bmp.decode(buf);
+    data = TessModule._malloc(bmpBuf.data.length * Uint8Array.BYTES_PER_ELEMENT);
+    TessModule.HEAPU8.set(bmpBuf.data, data);
+    w = bmpBuf.width;
+    h = bmpBuf.height;
+    bytesPerPixel = 4;
+  } else {
+    const ptr = TessModule._malloc(buf.length * Uint8Array.BYTES_PER_ELEMENT);
+    TessModule.HEAPU8.set(buf, ptr);
+    pix = TessModule._pixReadMem(ptr, buf.length);
+    if (TessModule.getValue(pix + (7 * 4), 'i32') === 0) {
+      /*
+       * Set a yres default value to prevent warning from tesseract
+       * See kMinCredibleResolution in tesseract/src/ccstruct/publictypes.h
+       */
+      TessModule.setValue(pix + (7 * 4), 70, 'i32');
+    }
+    [w, h] = Array(2).fill(0)
+      .map((v, idx) => (
+        TessModule.getValue(pix + (idx * 4), 'i32')
+      ));
   }
-  const [w, h, d, spp, wpl, refcount, xres, yres, informat, special] = Array(10).fill(0)
-    .map((v, idx) => (
-      TessModule.getValue(pix + (idx * 4), 'i32')
-    ));
+
   return {
     w,
     h,
-    d,
-    spp,
-    wpl,
-    refcount,
-    xres,
-    yres,
-    informat,
-    special,
-    data: pix,
+    bytesPerPixel,
+    data,
+    pix,
   };
 };

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

Total package byte prevSize

6806

increased by31.26%

Lines of code

177

increased by36.15%

Worsened metrics

Dependency count

6

increased by50%

Dependency changes

• + Addedbmp-js@^0.1.0

• + Addedfile-type@^10.5.0

• + Addedbmp-js@0.1.0(transitive)

• + Addedfile-type@10.11.0(transitive)