PDF extracting looks ok

src/lib.rs

@@ -2,16 +2,14 @@ use js_sys::Uint8Array;
 use pdf_extract::extract_text_from_mem;
 use wasm_bindgen::prelude::*;
 
-mod obsidian;
+// mod obsidian;
 
 #[wasm_bindgen]
-pub fn extract_pdf_text(arr: Uint8Array) -> String {
-    // FIXME: return a Result<> here, to throw in JS in case of an error
-    let txt = match extract_text_from_mem(&arr.to_vec()) {
-        Ok(txt) => txt,
-        Err(e) => e.to_string(),
+pub fn extract_pdf_text(arr: Uint8Array) -> Result<String, JsError> {
+    match extract_text_from_mem(&arr.to_vec()) {
+        Ok(txt) => return Ok(txt),
+        Err(e) => return Err(JsError::new(&e.to_string())),
     };
-    txt
 }
 
 // #[wasm_bindgen]

src/pdf-manager.ts

@@ -32,7 +32,7 @@ class PDFManager {
     const worker = new PDFWorker({ name: 'PDF Text Extractor' })
     return new Promise(async (resolve, reject) => {
       // @ts-ignore
-      worker.postMessage({ data })
+      worker.postMessage({ data, name: file.basename })
       worker.onmessage = (evt: any) => {
         const txt = evt.data.text
         this.updatePDFCache(hash, txt)

src/pdf-worker.ts

@@ -6,8 +6,13 @@ const decodedPlugin = decodeBase64(rustPlugin as any)
 onmessage = async evt => {
   const buffer = Uint8Array.from(decodedPlugin, c => c.charCodeAt(0))
   await plugin.default(Promise.resolve(buffer))
-  const text = plugin.extract_pdf_text(evt.data.data as Uint8Array)
-  self.postMessage({ text })
+  try {
+    const text = plugin.extract_pdf_text(evt.data.data as Uint8Array)
+    self.postMessage({ text })
+  } catch (e) {
+    console.warn('Omnisearch - Could not extract text from ' + evt.data.name)
+    self.postMessage({ text: '' })
+  }
 }
 
 function decodeBase64(data: string) {