tesseract.js - npm Package Compare versions

Comparing version 5.0.3 to 5.0.4

docs/api.md

@@ -54,4 +54,8 @@ # API
   - `errorHandler` a function to handle worker errors, a quick example is `err => console.error(err)`
+- `config` an object of customized options which are set prior to initialization
+  - This argument allows for setting "init only" Tesseract parameters
+	  - Most Tesseract parameters can be set after a worker is initialized, using either `worker.setParameters` or the `options` argument of `worker.recognize`.  
+	  - A handful of Tesseract parameters, referred to as "init only" parameters in Tesseract documentation, cannot be modified after Tesseract is initialized--these can only be set using this argument
+		  - Examples include `load_system_dawg`, `load_number_dawg`, and `load_punc_dawg`
 
-
 **Examples:**
@@ -61,3 +65,3 @@
 const { createWorker } = Tesseract;
-const worker = await createWorker({
+const worker = await createWorker('eng', 1, {
   langPath: '...',
@@ -69,3 +73,3 @@ logger: m => console.log(m),
 <a name="worker-recognize"></a>
-### Worker.recognize(image, options, jobId): Promise
+### Worker.recognize(image, options, output, jobId): Promise
 
@@ -154,2 +158,3 @@ Worker.recognize() provides core function of Tesseract.js as it executes OCR
 - `oem` a enum to indicate the OCR Engine Mode you use
+- `config` an object of customized options which are set prior to initialization (see details above)
 - `jobId` Please see details above
@@ -156,0 +161,0 @@

docs/local-installation.md

@@ -12,3 +12,3 @@ ## Local Installation
 ```javascript
-const worker = await createWorker({
+const worker = await createWorker('eng', 1, {
   workerPath: 'https://cdn.jsdelivr.net/npm/tesseract.js@v5.0.0/dist/worker.min.js',
@@ -15,0 +15,0 @@ langPath: 'https://tessdata.projectnaptha.com/4.0.0',

docs/workers_vs_schedulers.md

@@ -7,3 +7,3 @@ # Overview
 
-```
+```javascript
 (async () => {
@@ -26,3 +26,3 @@ const worker = await Tesseract.createWorker('eng');
  
-```
+```javascript
 const scheduler = Tesseract.createScheduler();
@@ -29,0 +29,0 @@

package.json

 {
   "name": "tesseract.js",
-  "version": "5.0.3",
+  "version": "5.0.4",
   "description": "Pure Javascript Multilingual OCR",
@@ -5,0 +5,0 @@ "main": "src/index.js",

src/worker-script/index.js

@@ -252,5 +252,5 @@ /**
     // In the latter case we convert to config file text here
-    if (typeof config === 'object') {
+    if (config && typeof config === 'object' && Object.keys(config).length > 0) {
       configStr = JSON.stringify(config).replace(/,/g, '\n').replace(/:/g, ' ').replace(/["'{}]/g, '');
-    } else {
+    } else if (config && typeof config === 'string') {
       configStr = config;
@@ -264,3 +264,3 @@ }
     api = new TessModule.TessBaseAPI();
-    let status = api.Init(null, langs, oem);
+    let status = api.Init(null, langs, oem, configFile);
     if (status === -1) {
@@ -294,3 +294,3 @@ // Cache is deleted if initialization fails to avoid keeping bad data in cache
           await loadLanguage({ workerId, payload: { langs: loadLanguageLangsWorker, options: loadLanguageOptionsWorker } }); // eslint-disable-line max-len
-          status = api.Init(null, langs, oem);
+          status = api.Init(null, langs, oem, configFile);
           if (status === -1) {
@@ -297,0 +297,0 @@ log('Language data refresh failed.');

No alert changes

Improved metrics

Total package byte prevSize

1451995

increased by0.08%

No dependency changes