Merge branch 'develop'

.github/ISSUE_TEMPLATE/bug_report.md

@@ -1,17 +1,12 @@
 ---
 name: Bug report
 about: Create a report to help us improve
-title: "[BUG]"
+title: "[BUG] "
 labels: ''
 assignees: ''
 
 ---
 
-<!--
-Please, if possible and before filing an issue,
-make sure that you have the latest available version of Omnisearch.
--->
-
 **Problem description:**
 
 <!-- Describe your problem in details. -->
@@ -20,8 +15,16 @@ make sure that you have the latest available version of Omnisearch.
 
 **Your environment:**
 
-- Omnisearch version:
+<!-- Please, if possible and before filing an issue, -->
-- Obsidian version:
+<!-- make sure that you have the latest available version of Omnisearch. -->
+
+- Omnisearch version: 
+- Obsidian version: 
 - Operating system: 
-- Number of notes in your vault (approx.):
+- Number of indexed documents in your vault (approx.): 
-- Other plugins that may be related to the issue:
+
+**Things to try:**
+
+- Does the problem occur when Obsidian is the only active community plugin:  
+- Does the problem occur when you _don't_ index PDFs, images, or other non-notes files:  
+- Does the problem occur after a cache reset: 

.github/workflows/release.yml

@@ -18,13 +18,13 @@ jobs:
 
       - uses: pnpm/action-setup@v2.1.0
         with:
-          version: 6.32.4
+          version: 7.17.0
           run_install: true
 
       - name: Use Node.js
         uses: actions/setup-node@v1
         with:
-          node-version: "14.x"
+          node-version: "18.x"
 
       - name: Build
         id: build

README.md

@@ -119,16 +119,6 @@ There are several CSS classes you can use to customize the appearance of Omnisea
 .omnisearch-input-field
 ```
 
-For example, if you'd like the usual yellow highlight on search matches, you can add this code inside a CSS snippet
-file:
-
-```css
-.omnisearch-highlight {
-  color: var(--text-normal);
-  background-color: var(--text-highlight-bg);
-}
-```
-
 See [styles.css](./assets/styles.css) for more information.
 
 ## Issues & Solutions

manifest.json

@@ -6,5 +6,9 @@
 	"description": "A search engine that just works",
 	"author": "Simon Cambier",
 	"authorUrl": "https://github.com/scambier/obsidian-omnisearch",
-	"isDesktopOnly": false
+	"isDesktopOnly": false,
+    "fundingUrl": {
+		"Github": "https://github.com/sponsors/scambier",
+		"Ko-fi": "https://ko-fi.com/scambier"
+	}
 }

package.json

@@ -1,13 +1,13 @@
 {
   "name": "scambier.obsidian-search",
-  "version": "1.9.1",
+  "version": "1.10.0-beta.4",
   "description": "A search engine for Obsidian",
   "main": "dist/main.js",
   "scripts": {
     "dev": "node esbuild.config.mjs",
     "build": "tsc -noEmit -skipLibCheck && node esbuild.config.mjs production",
     "check": "svelte-check --tsconfig ./tsconfig.json",
-    "version": "node version-bump.mjs && git add manifest.json versions.json package.json",
+    "version": "node version-bump.mjs",
     "test": "jest"
   },
   "keywords": [],
@@ -43,7 +43,7 @@
     "dexie": "^3.2.2",
     "lodash-es": "4.17.21",
     "minisearch": "6.0.0-beta.1",
-    "obsidian-text-extract": "1.0.4",
+    "obsidian-text-extract": "^1.0.4",
     "pure-md5": "^0.1.14"
   },
   "pnpm": {

pnpm-lock.yaml

@@ -22,7 +22,7 @@ specifiers:
   lodash-es: 4.17.21
   minisearch: 6.0.0-beta.1
   obsidian: latest
-  obsidian-text-extract: 1.0.4
+  obsidian-text-extract: ^1.0.4
   prettier: ^2.8.1
   prettier-plugin-svelte: ^2.8.1
   pure-md5: ^0.1.14
@@ -56,7 +56,7 @@ devDependencies:
   esbuild-plugin-copy: 1.3.0_esbuild@0.14.0
   esbuild-svelte: 0.7.1_wvi5wuag3veo5vm52k3h7pgaae
   jest: 27.5.1
-  obsidian: 0.16.3
+  obsidian: 1.1.1
   prettier: 2.8.1
   prettier-plugin-svelte: 2.8.1_sro2v6ld777payjtkjtiuogcxi
   svelte: 3.54.0
@@ -4138,6 +4138,7 @@ packages:
 
   /obsidian-text-extract/1.0.4:
     resolution: {integrity: sha512-lJ7HaEPGUoGVYr6iCSEU8oPrnA8xRBQgSU9KryWteq3wO+Yiw2Zue70G0rZf9Yj+DkgR92PtuR8aQZQxTy/uLA==}
+    deprecated: Package no longer supported. Contact Support at https://www.npmjs.com/support for more info.
     dependencies:
       dexie: 3.2.2
       p-limit: 4.0.0
@@ -4149,8 +4150,8 @@ packages:
       - supports-color
     dev: false
 
-  /obsidian/0.16.3:
+  /obsidian/1.1.1:
-    resolution: {integrity: sha512-hal9qk1A0GMhHSeLr2/+o3OpLmImiP+Y+sx2ewP13ds76KXsziG96n+IPFT0mSkup1zSwhEu+DeRhmbcyCCXWw==}
+    resolution: {integrity: sha512-GcxhsHNkPEkwHEjeyitfYNBcQuYGeAHFs1pEpZIv0CnzSfui8p8bPLm2YKLgcg20B764770B1sYGtxCvk9ptxg==}
     peerDependencies:
       '@codemirror/state': ^6.0.0
       '@codemirror/view': ^6.0.0

src/cache-manager.ts

@@ -1,12 +1,15 @@
 import { Notice } from 'obsidian'
-import type { DocumentRef, IndexedDocument } from './globals'
+import {
+  type DocumentRef,
+  getTextExtractor,
+  type IndexedDocument,
+} from './globals'
 import { database } from './database'
-import type { AsPlainObject } from 'minisearch'
-import type MiniSearch from 'minisearch'
 import {
   extractHeadingsFromCache,
   getAliasesFromMetadata,
   getTagsFromMetadata,
+  isFileCanvas,
   isFileImage,
   isFilePDF,
   isFilePlaintext,
@@ -14,19 +17,71 @@ import {
   removeDiacritics,
 } from './tools/utils'
 import { getImageText, getPdfText } from 'obsidian-text-extract'
+import type { CanvasData } from 'obsidian/canvas'
+import type { AsPlainObject } from 'minisearch'
+import type MiniSearch from 'minisearch'
 
-async function getIndexedDocument(path: string): Promise<IndexedDocument> {
+/**
+ * This function is responsible for extracting the text from a file and
+ * returning it as an `IndexedDocument` object.
+ * @param path
+ */
+async function getAndMapIndexedDocument(
+  path: string
+): Promise<IndexedDocument> {
   const file = app.vault.getFiles().find(f => f.path === path)
   if (!file) throw new Error(`Invalid file path: "${path}"`)
-  let content: string
+  let content: string | null = null
+
+  const extractor = getTextExtractor()
+
+  // ** Plain text **
+  // Just read the file content
   if (isFilePlaintext(path)) {
     content = await app.vault.cachedRead(file)
-  } else if (isFilePDF(path)) {
+  }
-    content = await getPdfText(file)
+
-  } else if (isFileImage(file.path)) {
+  // ** Canvas **
-    content = await getImageText(file)
+  // Extract the text fields from the json
-  } else {
+  else if (isFileCanvas(path)) {
-    throw new Error('Invalid file format: ' + file.path)
+    const canvas = JSON.parse(await app.vault.cachedRead(file)) as CanvasData
+    let texts: string[] = []
+    // Concatenate text from the canvas fields
+    for (const node of canvas.nodes) {
+      if (node.type === 'text') {
+        texts.push(node.text)
+      } else if (node.type === 'file') {
+        texts.push(node.file)
+      }
+    }
+    for (const edge of canvas.edges.filter(e => !!e.label)) {
+      texts.push(edge.label!)
+    }
+    content = texts.join('\r\n')
+  }
+
+  // a) ** Image or PDF ** with Text Extractor
+  else if (extractor) {
+    if (extractor.canFileBeExtracted(path)) {
+      content = await extractor.extractText(file)
+    } else {
+      throw new Error('Invalid file format: ' + file.path)
+    }
+  }
+  // b) ** Image or PDF ** without the text-extractor plugin
+  else {
+    if (isFilePDF(path)) {
+      content = await getPdfText(file)
+    } else if (isFileImage(file.path)) {
+      content = await getImageText(file)
+    } else {
+      throw new Error('Invalid file format: ' + file.path)
+    }
+  }
+  if (content === null || content === undefined) {
+    // This shouldn't happen
+    console.warn(`Omnisearch: ${content} content for file`, file.path)
+    content = ''
   }
   content = removeDiacritics(content)
   const metadata = app.metadataCache.getFileCache(file)
@@ -81,8 +136,12 @@ class CacheManager {
   private documents: Map<string, IndexedDocument> = new Map()
 
   public async addToLiveCache(path: string): Promise<void> {
-    const doc = await getIndexedDocument(path)
+    try {
-    this.documents.set(path, doc)
+      const doc = await getAndMapIndexedDocument(path)
+      this.documents.set(path, doc)
+    } catch (e) {
+      console.warn('Omnisearch: Error while adding to live cache', e)
+    }
   }
 
   public removeFromLiveCache(path: string): void {

src/components/InputSearch.svelte

@@ -11,6 +11,10 @@
   let elInput: HTMLInputElement
   const dispatch = createEventDispatcher()
 
+  export function setInputValue(v:string): void {
+    value = v
+  }
+
   $: {
     if (initialValue && !initialSet && !value) {
       initialSet = true
@@ -19,12 +23,15 @@
     }
   }
 
-  async function selectInput() {
+  function selectInput(_?: HTMLElement): void {
-    await tick()
+    tick()
-    elInput.focus()
+      .then(() => {
-    await tick()
+        elInput.focus()
-    elInput.select()
+        return tick()
-    await tick()
+      })
+      .then(() => {
+        elInput.select()
+      })
   }
 
   const debouncedOnInput = debounce(() => {

src/components/ModalVault.svelte

@@ -23,6 +23,7 @@
   let query: Query
   let indexingStepDesc = ''
   let searching = true
+  let refInput: InputSearch|undefined
 
   $: selectedNote = resultNotes[selectedIndex]
   $: searchQuery = searchQuery ?? previousQuery
@@ -31,6 +32,8 @@
     searching = true
     updateResults().then(() => {
       searching = false
+    }).catch((e) => {
+      console.error(e)
     })
   } else {
     searching = false
@@ -85,7 +88,8 @@
     if (++historySearchIndex >= history.length) {
       historySearchIndex = 0
     }
-    previousQuery = history[historySearchIndex]
+    searchQuery = history[historySearchIndex]
+    refInput?.setInputValue(searchQuery)
   }
 
   async function nextSearchHistory() {
@@ -93,7 +97,8 @@
     if (--historySearchIndex < 0) {
       historySearchIndex = history.length ? history.length - 1 : 0
     }
-    previousQuery = history[historySearchIndex]
+    searchQuery = history[historySearchIndex]
+    refInput?.setInputValue(searchQuery)
   }
 
   async function updateResults() {
@@ -228,6 +233,7 @@
 
 <InputSearch
   initialValue="{searchQuery}"
+  bind:this={refInput}
   on:input="{e => (searchQuery = e.detail)}"
   placeholder="Omnisearch - Vault">
   {#if settings.showCreateButton}

src/globals.ts

@@ -1,6 +1,7 @@
 import { EventBus } from './tools/event-bus'
 import { writable } from 'svelte/store'
 import { settings } from './settings'
+import type { TFile } from 'obsidian'
 
 export const regexLineSplit = /\r?\n|\r|((\.|\?|!)( |\r?\n|\r))/g
 export const regexYaml = /^---\s*\n(.*?)\n?^---\s?/ms
@@ -76,9 +77,27 @@ export function isInputComposition(): boolean {
   return inComposition
 }
 
+
+/**
+ * Plugin dependency - Chs Patch for Chinese word segmentation
+ * @returns 
+ */
 export function getChsSegmenter(): any | undefined {
   return (app as any).plugins.plugins['cm-chs-patch']
 }
 
+export type TextExtractorApi = {
+  extractText: (file: TFile) => Promise<string>
+  canFileBeExtracted: (filePath: string) => boolean
+}
+
+/**
+ * Plugin dependency - Text Extractor
+ * @returns 
+ */
+export function getTextExtractor(): TextExtractorApi | undefined {
+  return (app as any).plugins?.plugins?.['text-extractor']?.api
+}
+
 export const SPACE_OR_PUNCTUATION =
   /[|\n\r -#%-*,-/:;?@[-\]_{}\u00A0\u00A1\u00A7\u00AB\u00B6\u00B7\u00BB\u00BF\u037E\u0387\u055A-\u055F\u0589\u058A\u05BE\u05C0\u05C3\u05C6\u05F3\u05F4\u0609\u060A\u060C\u060D\u061B\u061E\u061F\u066A-\u066D\u06D4\u0700-\u070D\u07F7-\u07F9\u0830-\u083E\u085E\u0964\u0965\u0970\u09FD\u0A76\u0AF0\u0C77\u0C84\u0DF4\u0E4F\u0E5A\u0E5B\u0F04-\u0F12\u0F14\u0F3A-\u0F3D\u0F85\u0FD0-\u0FD4\u0FD9\u0FDA\u104A-\u104F\u10FB\u1360-\u1368\u1400\u166E\u1680\u169B\u169C\u16EB-\u16ED\u1735\u1736\u17D4-\u17D6\u17D8-\u17DA\u1800-\u180A\u1944\u1945\u1A1E\u1A1F\u1AA0-\u1AA6\u1AA8-\u1AAD\u1B5A-\u1B60\u1BFC-\u1BFF\u1C3B-\u1C3F\u1C7E\u1C7F\u1CC0-\u1CC7\u1CD3\u2000-\u200A\u2010-\u2029\u202F-\u2043\u2045-\u2051\u2053-\u205F\u207D\u207E\u208D\u208E\u2308-\u230B\u2329\u232A\u2768-\u2775\u27C5\u27C6\u27E6-\u27EF\u2983-\u2998\u29D8-\u29DB\u29FC\u29FD\u2CF9-\u2CFC\u2CFE\u2CFF\u2D70\u2E00-\u2E2E\u2E30-\u2E4F\u3000-\u3003\u3008-\u3011\u3014-\u301F\u3030\u303D\u30A0\u30FB\uA4FE\uA4FF\uA60D-\uA60F\uA673\uA67E\uA6F2-\uA6F7\uA874-\uA877\uA8CE\uA8CF\uA8F8-\uA8FA\uA8FC\uA92E\uA92F\uA95F\uA9C1-\uA9CD\uA9DE\uA9DF\uAA5C-\uAA5F\uAADE\uAADF\uAAF0\uAAF1\uABEB\uFD3E\uFD3F\uFE10-\uFE19\uFE30-\uFE52\uFE54-\uFE61\uFE63\uFE68\uFE6A\uFE6B\uFF01-\uFF03\uFF05-\uFF0A\uFF0C-\uFF0F\uFF1A\uFF1B\uFF1F\uFF20\uFF3B-\uFF3D\uFF3F\uFF5B\uFF5D\uFF5F-\uFF65]+/u

src/main.ts

@@ -46,7 +46,9 @@ export default class OmnisearchPlugin extends Plugin {
       id: 'show-modal-infile',
       name: 'In-file search',
       editorCallback: (_editor, view) => {
-        new OmnisearchInFileModal(app, view.file).open()
+        if (view.file) {
+          new OmnisearchInFileModal(app, view.file).open()
+        }
       },
     })
 
@@ -190,15 +192,13 @@ async function cleanOldCacheFiles() {
 }
 
 function executeFirstLaunchTasks(plugin: Plugin) {
-  const code = '1.8.0-beta.3'
+  const code = '1.10.0-beta.1'
   if (settings.welcomeMessage !== code) {
     const welcome = new DocumentFragment()
     welcome.createSpan({}, span => {
-      span.innerHTML = `<strong>Omnisearch has been updated</strong>
+      span.innerHTML = `🔎 Omnisearch will soon require the <strong>Text Extractor</strong> plugin to index PDF and images. See Omnisearch settings for more information.`
-You can now enable "Images Indexing" to use Optical Character Recognition on your scanned documents
-🔎🖼`
     })
-    new Notice(welcome, 30000)
+    new Notice(welcome, 20_000)
   }
   settings.welcomeMessage = code
 

src/search/omnisearch.ts

@@ -104,8 +104,7 @@ export class Omnisearch {
       documents.filter(d => this.indexedDocuments.has(d.path)).map(d => d.path)
     )
 
-    // Split the documents in smaller chunks to regularly save the cache.
+    // Split the documents in smaller chunks to add them to minisearch
-    // If the user shuts off Obsidian mid-indexing, we at least saved some
     const chunkedDocs = chunkArray(documents, 500)
     for (const docs of chunkedDocs) {
       // Update the list of indexed docs
@@ -146,7 +145,10 @@ export class Omnisearch {
 
     let results = this.minisearch.search(query.segmentsToStr(), {
       prefix: term => term.length >= options.prefixLength,
-      fuzzy: 0.2,
+      // length <= 3: no fuzziness
+      // length <= 5: fuzziness of 10%
+      // length > 5: fuzziness of 20%
+      fuzzy: term => (term.length <= 3 ? 0 : term.length <= 5 ? 0.1 : 0.2),
       combineWith: 'AND',
       boost: {
         basename: settings.weightBasename,

src/settings.ts

@@ -8,6 +8,7 @@ import {
 } from 'obsidian'
 import { writable } from 'svelte/store'
 import { database } from './database'
+import { getTextExtractor } from './globals'
 import type OmnisearchPlugin from './main'
 
 interface WeightingSettings {
@@ -83,16 +84,30 @@ export class SettingsTab extends PluginSettingTab {
 
     new Setting(containerEl).setName('Indexing').setHeading()
 
+    if (getTextExtractor()) {
+      const desc = new DocumentFragment()
+      desc.createSpan({}, span => {
+        span.innerHTML = `👍 You have installed <a href="https://github.com/scambier/obsidian-text-extractor">Text Extractor</a>, Omnisearch will use it to index PDFs and images.
+            <br />Text extraction only works on desktop, but the cache can be synchronized with your mobile device.`
+      })
+      new Setting(containerEl).setDesc(desc)
+    } else {
+      const label = new DocumentFragment()
+      label.createSpan({}, span => {
+        span.innerHTML = `⚠️ Omnisearch will soon require <a href="https://github.com/scambier/obsidian-text-extractor">Text Extractor</a> to index PDFs and images.
+        You can already install it to get a head start.`
+      })
+      new Setting(containerEl).setDesc(label)
+    }
+
     // PDF Indexing
-    if (!Platform.isMobileApp) {
+    if (!Platform.isMobileApp || getTextExtractor()) {
       const indexPDFsDesc = new DocumentFragment()
       indexPDFsDesc.createSpan({}, span => {
-        span.innerHTML = `Omnisearch will include PDFs in search results.<br>
+        span.innerHTML = `Include PDFs in search results - Will soon depend on Text Extractor.`
-        ⚠️ PDFs first need to be processed. This can take anywhere from a few seconds to 2 minutes, then the resulting text is cached.</li>
-        <strong style="color: var(--text-accent)">Needs a restart to fully take effect.</strong>`
       })
       new Setting(containerEl)
-        .setName('PDF Indexing')
+        .setName(`PDFs Indexing`)
         .setDesc(indexPDFsDesc)
         .addToggle(toggle =>
           toggle.setValue(settings.PDFIndexing).onChange(async v => {
@@ -100,21 +115,14 @@ export class SettingsTab extends PluginSettingTab {
             await saveSettings(this.plugin)
           })
         )
-    }
 
-    // Images Indexing
+      // Images Indexing
-    if (!Platform.isMobileApp) {
       const indexImagesDesc = new DocumentFragment()
       indexImagesDesc.createSpan({}, span => {
-        span.innerHTML = `Omnisearch will use <a href="https://en.wikipedia.org/wiki/Tesseract_(software)">Tesseract</a> to index images from their text.
+        span.innerHTML = `Include images in search results - Will soon depend on Text Extractor.`
-        <ul>
-          <li>Only English is supported at the moment.</li>
-          <li>Not all images can be correctly read by the OCR, this feature works best with scanned documents.</li>
-        </ul>      
-        <strong style="color: var(--text-accent)">Needs a restart to fully take effect.</strong>`
       })
       new Setting(containerEl)
-        .setName('BETA - Images Indexing')
+        .setName(`Images Indexing`)
         .setDesc(indexImagesDesc)
         .addToggle(toggle =>
           toggle.setValue(settings.imagesIndexing).onChange(async v => {
@@ -123,6 +131,7 @@ export class SettingsTab extends PluginSettingTab {
           })
         )
     }
+
     // Additional files to index
     const indexedFileTypesDesc = new DocumentFragment()
     indexedFileTypesDesc.createSpan({}, span => {
@@ -169,7 +178,8 @@ export class SettingsTab extends PluginSettingTab {
     const diacriticsDesc = new DocumentFragment()
     diacriticsDesc.createSpan({}, span => {
       span.innerHTML = `Normalize diacritics in search terms. Words like "brûlée" or "žluťoučký" will be indexed as "brulee" and "zlutoucky".<br/>
-        ⚠️<span style="color: var(--text-accent)">You probably should NOT disable this.</span><br>
+        ⚠️ <span style="color: var(--text-accent)">You probably should <strong>NOT</strong> disable this.</span><br>
+        ⚠️ <span style="color: var(--text-accent)">Changing this setting will clear the cache.</span><br>
         <strong style="color: var(--text-accent)">Needs a restart to fully take effect.</strong>
         `
     })
@@ -178,6 +188,7 @@ export class SettingsTab extends PluginSettingTab {
       .setDesc(diacriticsDesc)
       .addToggle(toggle =>
         toggle.setValue(settings.ignoreDiacritics).onChange(async v => {
+          await database.clearCache()
           settings.ignoreDiacritics = v
           await saveSettings(this.plugin)
         })
@@ -349,7 +360,7 @@ export class SettingsTab extends PluginSettingTab {
   }
 
   weightSlider(cb: SliderComponent, key: keyof WeightingSettings): void {
-    cb.setLimits(1, 3, 0.1)
+    cb.setLimits(1, 5, 0.1)
       .setValue(settings[key])
       .setDynamicTooltip()
       .onChange(v => {

src/tools/api.ts

@@ -1,6 +1,7 @@
 import type { ResultNote } from '../globals'
 import { Query } from '../search/query'
 import { searchEngine } from '../search/omnisearch'
+import { makeExcerpt } from './utils'
 
 type ResultNoteApi = {
   score: number
@@ -8,6 +9,7 @@ type ResultNoteApi = {
   basename: string
   foundWords: string[]
   matches: SearchMatchApi[]
+  excerpt: string
 }
 
 export type SearchMatchApi = {
@@ -17,7 +19,10 @@ export type SearchMatchApi = {
 
 function mapResults(results: ResultNote[]): ResultNoteApi[] {
   return results.map(result => {
-    const { score, path, basename, foundWords, matches } = result
+    const { score, path, basename, foundWords, matches, content } = result
+
+    const excerpt = makeExcerpt(content, matches[0]?.offset ?? -1)
+
     return {
       score,
       path,
@@ -29,14 +34,18 @@ function mapResults(results: ResultNote[]): ResultNoteApi[] {
           offset: match.offset,
         }
       }),
+      excerpt,
     }
   })
 }
 
-async function search(q: string): Promise<ResultNoteApi[]> {
+async function search(
+  q: string,
+  options: Partial<{ excerpt: boolean }> = {}
+): Promise<ResultNoteApi[]> {
   const query = new Query(q)
   const raw = await searchEngine.getSuggestions(query)
   return mapResults(raw)
 }
 
-export default {search}
+export default { search }

src/tools/utils.ts

@@ -5,11 +5,11 @@ import {
   parseFrontMatterAliases,
   Platform,
 } from 'obsidian'
-import type { SearchMatch } from '../globals'
+import { getTextExtractor, type SearchMatch } from '../globals'
 import {
-  getChsSegmenter,
   excerptAfter,
   excerptBefore,
+  getChsSegmenter,
   highlightClass,
   isSearchMatch,
   regexLineSplit,
@@ -195,14 +195,40 @@ export function getAliasesFromMetadata(
 }
 
 export function getTagsFromMetadata(metadata: CachedMetadata | null): string[] {
-  return metadata ? getAllTags(metadata) ?? [] : []
+  let tags = metadata ? getAllTags(metadata) ?? [] : []
+  // This will "un-nest" tags that are in the form of "#tag/subtag"
+  // A tag like "#tag/subtag" will be split into 3 tags: '#tag/subtag", "#tag" and "#subtag"
+  // https://github.com/scambier/obsidian-omnisearch/issues/146
+  tags = [
+    ...new Set(
+      tags.reduce((acc, tag) => {
+        return [
+          ...acc,
+          ...tag
+            .split('/')
+            .filter(t => t)
+            .map(t => (t.startsWith('#') ? t : `#${t}`)),
+          tag,
+        ]
+      }, [] as string[])
+    ),
+  ]
+  return tags
 }
 
 /**
  * https://stackoverflow.com/a/37511463
  */
 export function removeDiacritics(str: string): string {
-  return str.normalize('NFD').replace(/\p{Diacritic}/gu, '')
+  if (str === null || str === undefined) {
+    return ''
+  }
+  // Keep backticks for code blocks, because otherwise they are removed by the .normalize() function
+  // https://stackoverflow.com/a/36100275
+  str = str.replaceAll('`', '[__omnisearch__backtick__]')
+  str = str.normalize('NFD').replace(/\p{Diacritic}/gu, '')
+  str = str.replaceAll('[__omnisearch__backtick__]', '`')
+  return str
 }
 
 export function getCtrlKeyLabel(): 'ctrl' | '⌘' {
@@ -210,10 +236,13 @@ export function getCtrlKeyLabel(): 'ctrl' | '⌘' {
 }
 
 export function isFileIndexable(path: string): boolean {
+  const canIndexPDF = (!Platform.isMobileApp || !!getTextExtractor()) && settings.PDFIndexing
+  const canIndexImages = (!Platform.isMobileApp || !!getTextExtractor()) && settings.imagesIndexing
   return (
     isFilePlaintext(path) ||
-    (!Platform.isMobileApp && settings.PDFIndexing && isFilePDF(path)) ||
+    isFileCanvas(path) ||
-    (!Platform.isMobileApp && settings.imagesIndexing && isFileImage(path))
+    (canIndexPDF && isFilePDF(path)) ||
+    (canIndexImages && isFileImage(path))
   )
 }
 
@@ -228,11 +257,11 @@ export function isFilePDF(path: string): boolean {
 }
 
 export function isFilePlaintext(path: string): boolean {
-  return getPlaintextExtensions().some(t => path.endsWith(`.${t}`))
+  return [...settings.indexedFileTypes, 'md'].some(t => path.endsWith(`.${t}`))
 }
 
-export function getPlaintextExtensions(): string[] {
+export function isFileCanvas(path: string): boolean {
-  return [...settings.indexedFileTypes, 'md']
+  return path.endsWith('.canvas')
 }
 
 export function getExtension(path: string): string {

version-bump.mjs

@@ -1,6 +1,7 @@
 import { readFileSync, writeFileSync } from 'fs'
 
 const targetVersion = process.env.npm_package_version
+console.log(`Bumping version to ${targetVersion}`)
 
 // read minAppVersion from manifest.json and bump version to target version
 const manifest = JSON.parse(readFileSync('manifest-beta.json', 'utf8'))