Added interface for setting 'init only' options per #613

naptha · Sep 24, 2022 · b952488 · b952488
1 parent c0298ff
commit b952488
Show file tree

Hide file tree

Showing 3 changed files with 27 additions and 5 deletions.
diff --git a/src/createWorker.js b/src/createWorker.js
@@ -113,11 +113,11 @@ module.exports = async (_options = {}) => {
     }))
   );
 
-  const initialize = (langs = 'eng', oem = defaultOEM, jobId) => (
+  const initialize = (langs = 'eng', oem = defaultOEM, config, jobId) => (
     startJob(createJob({
       id: jobId,
       action: 'initialize',
-      payload: { langs, oem },
+      payload: { langs, oem, config },
     }))
   );
 

diff --git a/src/index.d.ts b/src/index.d.ts
@@ -20,7 +20,7 @@ declare namespace Tesseract {
     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>
+    initialize(langs?: string, oem?: OEM, config?: string | Partial<InitOptions>, jobId?: string): Promise<ConfigResult>
     setParameters(params: Partial<WorkerParams>, jobId?: string): Promise<ConfigResult>
     getImage(type: imageType): string
     recognize(image: ImageLike, options?: Partial<RecognizeOptions>, output?: Partial<OutputFormats>, jobId?: string): Promise<RecognizeResult>
@@ -29,6 +29,14 @@ declare namespace Tesseract {
     getPDF(title?: string, textonly?: boolean, jobId?: string):Promise<GetPDFResult>
   }
 
+  interface InitOptions {
+    load_system_dawg: string
+    load_freq_dawg: string
+    load_unambig_dawg: string
+    load_punc_dawg: string
+    load_number_dawg: string
+    load_bigram_dawg: string
+  }
   interface WorkerOptions {
     corePath: string
     langPath: string

diff --git a/src/worker-script/index.js b/src/worker-script/index.js
@@ -176,7 +176,7 @@ const setParameters = async ({ payload: { params: _params } }, res) => {
 
 const initialize = async ({
   workerId,
-  payload: { langs: _langs, oem },
+  payload: { langs: _langs, oem, config},
 }, res) => {
   const langs = (typeof _langs === 'string')
     ? _langs
@@ -189,8 +189,22 @@ const initialize = async ({
     if (api !== null) {
       api.End();
     }
+    let configFile = undefined;
+    let configStr = undefined;
+    // config argument may either be config file text, or object with key/value pairs
+    // In the latter case we convert to config file text here
+    if (typeof config === "object") {
+      configStr = JSON.stringify(config).replace(/,/g, "\n").replace(/:/g, " ").replace(/["'{}]/g, "");
+    } else {
+      configStr = config;
+    }
+    if (typeof configStr === "string") {
+      configFile = "/config";
+      TessModule.FS.writeFile(configFile, configStr);
+    }
+
     api = new TessModule.TessBaseAPI();
-    api.Init(null, langs, oem);
+    api.Init(null, langs, oem, configFile);
     params = defaultParams;
     await setParameters({ payload: { params } });
     res.progress({