Fixed potential indexing issues

src/cache-manager.ts

@@ -60,12 +60,10 @@ async function getAndMapIndexedDocument(
   }
 
   // ** Image or PDF **
-  if (extractor) {
+  else if (extractor?.canFileBeExtracted(path)) {
-    if (extractor.canFileBeExtracted(path)) {
     content = await extractor.extractText(file)
   } else {
-      throw new Error('Invalid file format: ' + file.path)
+    throw new Error(`Unsupported file type: "${path}"`)
-    }
   }
 
   if (content === null || content === undefined) {
@@ -125,6 +123,10 @@ class CacheManager {
    */
   private documents: Map<string, IndexedDocument> = new Map()
 
+  /**
+   * Set or update the live cache with the content of the given file.
+   * @param path
+   */
   public async addToLiveCache(path: string): Promise<void> {
     try {
       const doc = await getAndMapIndexedDocument(path)

src/main.ts

@@ -64,11 +64,12 @@ export default class OmnisearchPlugin extends Plugin {
       },
     })
 
+    app.workspace.onLayoutReady(async () => {
       // Listeners to keep the search index up-to-date
       this.registerEvent(
-      this.app.vault.on('create', async file => {
+        this.app.vault.on('create', file => {
           if (isFileIndexable(file.path)) {
-          await cacheManager.addToLiveCache(file.path)
+            // await cacheManager.addToLiveCache(file.path)
             searchEngine.addFromPaths([file.path])
           }
         })
@@ -98,7 +99,6 @@ export default class OmnisearchPlugin extends Plugin {
         })
       )
 
-    app.workspace.onLayoutReady(async () => {
       this.executeFirstLaunchTasks()
       await this.populateIndex()
     })
@@ -145,31 +145,39 @@ export default class OmnisearchPlugin extends Plugin {
     indexingStep.set(IndexingStepType.ReadingFiles)
     const files = app.vault.getFiles().filter(f => isFileIndexable(f.path))
     console.log(`Omnisearch - ${files.length} files total`)
-
+    console.log(
+      `Omnisearch - Cache is ${isCacheEnabled() ? 'enabled' : 'disabled'}`
+    )
     // Map documents in the background
     // Promise.all(files.map(f => cacheManager.addToLiveCache(f.path)))
 
     if (isCacheEnabled()) {
       console.time('Omnisearch - Loading index from cache')
       indexingStep.set(IndexingStepType.LoadingCache)
-      await searchEngine.loadCache()
+      const hasCache = await searchEngine.loadCache()
+      if (hasCache) {
         console.timeEnd('Omnisearch - Loading index from cache')
       }
+    }
 
     const diff = searchEngine.getDiff(
       files.map(f => ({ path: f.path, mtime: f.stat.mtime }))
     )
 
+    if (isCacheEnabled()) {
       if (diff.toAdd.length) {
         console.log(
-        'Omnisearch - Total number of files to add/update: ' + diff.toAdd.length
+          'Omnisearch - Total number of files to add/update: ' +
+            diff.toAdd.length
         )
       }
       if (diff.toRemove.length) {
         console.log(
-        'Omnisearch - Total number of files to remove: ' + diff.toRemove.length
+          'Omnisearch - Total number of files to remove: ' +
+            diff.toRemove.length
         )
       }
+    }
 
     if (diff.toAdd.length >= 1000 && isCacheEnabled()) {
       new Notice(

src/search/omnisearch.ts

@@ -16,6 +16,7 @@ import {
 import { Notice, Platform } from 'obsidian'
 import type { Query } from './query'
 import { cacheManager } from '../cache-manager'
+import { sortBy } from 'lodash-es'
 
 const tokenize = (text: string): string[] => {
   const tokens = text.split(SPACE_OR_PUNCTUATION)
@@ -59,12 +60,19 @@ export class Omnisearch {
     this.minisearch = new MiniSearch(Omnisearch.options)
   }
 
-  async loadCache(): Promise<void> {
+  /**
+   * Return true if the cache is valid
+   */
+  async loadCache(): Promise<boolean> {
     const cache = await cacheManager.getMinisearchCache()
     if (cache) {
+      // console.log('Omnisearch - Cache', cache)
       this.minisearch = MiniSearch.loadJS(cache.data, Omnisearch.options)
       this.indexedDocuments = new Map(cache.paths.map(o => [o.path, o.mtime]))
+      return true
     }
+    console.log('Omnisearch - No cache found')
+    return false
   }
 
   /**
@@ -77,11 +85,13 @@ export class Omnisearch {
   } {
     const docsMap = new Map(docs.map(d => [d.path, d.mtime]))
 
+    // console.log(this.indexedDocuments)
     const toAdd = docs.filter(
       d =>
         !this.indexedDocuments.has(d.path) ||
         this.indexedDocuments.get(d.path) !== d.mtime
     )
+    // console.log(toAdd)
     const toRemove = [...this.indexedDocuments]
       .filter(
         ([path, mtime]) => !docsMap.has(path) || docsMap.get(path) !== mtime
@@ -100,6 +110,8 @@ export class Omnisearch {
         paths.map(async path => await cacheManager.getDocument(path))
       )
     ).filter(d => !!d?.path)
+    // Index markdown files first
+    documents = sortBy(documents, d => (d.path.endsWith('.md') ? 0 : 1))
 
     // If a document is already added, discard it
     this.removeFromPaths(

src/tools/utils.ts

@@ -237,8 +237,9 @@ export function getCtrlKeyLabel(): 'ctrl' | '⌘' {
 }
 
 export function isFileIndexable(path: string): boolean {
-  const canIndexPDF = !!getTextExtractor() && settings.PDFIndexing
+  const hasTextExtractor = !!getTextExtractor()
-  const canIndexImages = !!getTextExtractor() && settings.imagesIndexing
+  const canIndexPDF = hasTextExtractor && settings.PDFIndexing
+  const canIndexImages = hasTextExtractor && settings.imagesIndexing
   return (
     isFilePlaintext(path) ||
     isFileCanvas(path) ||